Skip to navigation – Site map

A dilemma regarding the nature of norms

Ricardo Caracciolo

Abstract

Most philosophical theories generally accept a set of ideas concerning the nature of norms together with a certain conception of action that can be traced back to Hume. These ideas can be expressed clearly through several theses that are notoriously plausible. On the one hand, there is the thesis that norms, and the duties constituted by norms, are objective in a certain sense, while on the other hand, there is the thesis that the concept of “norm” implies the possibility of complying (or not complying) with the relevant normative requirements. The problem is that these theses cannot be held simultaneously, thus giving rise to a genuine dilemma. This paper explores several ways of avoiding the dilemma, and argues that they are all ultimately unsuccessful. | The Spanish original of this paper was published in Doxa. Cuadernos de filosofía del derecho (2008) 31: 91–104.

Top of page

Full text

1 Introduction

1The idea that norms exist and that, at least in part, our everyday behaviour is determined by norms we unwittingly follow, is doubtless part of the way we perceive our social environment and our actions. Literally speaking, this is obvious with regard to our self-understanding. Besides, corroborating the existence of morality and law appears to rely on this manner of conceiving the world. Likewise, it is also generally accepted that norms do not refer only to actions. It is assumed that there are norms that refer to beliefs, to attitudes, and even to the way people use language to communicate with each other. As we know, this everyday perception gives rise to highly important, complex philosophical problems that remain as yet unresolved. I intend to address only one of these problems in this paper, the one which refers to those norms with the purpose of regulating behaviour.

  • 1 By “general” in this context I understand an idea or thesis that can be assumed by different alter (...)

2 The scope of the discussion needs to be circumscribed further because the term “norm” is used in a number of ways and norms are supposed to fulfil a series of functions. In one sense, norms supply assessment criteria, which means they make it possible to distinguish “correct” actions from “incorrect” actions (or justified actions from unjustified actions). In another sense, not necessarily incompatible with the former, norms provide guidelines for behaviour. Put in different words, they are tools for resolving practical issues, that is to say, issues regarding the action that ought to be chosen in contexts in which alternative actions have to be opted for. If the focus is even more closely constrained to this last function, a question arises regarding the issue of knowing how and under what conditions norms determine, or are able to determine, the choices of the agent or agents involved. This, in turn, seems to demand an elucidation, firstly, of what “determine” means in this context, and secondly, of the relationships there are – or there may be – between norms and actions. Clearly, a response to this question depends on what there is to be said about norms and about actions. Since there is no way I can present a thorough discussion on this matter in this paper, I shall here only attempt to show the following: current philosophical theories present (a) a set of general ideas regarding the nature of norms, on the one hand,1 and (b) another set of general ideas in relation to the nature of actions, that are notoriously plausible and that most people would be willing to accept as mere truisms, on the other. The problem is that they cannot be held simultaneously and that, as a result, a consistent conception of norms and actions would necessarily have to forsake one of these ideas. A genuine dilemma is involved in deciding which of them one is prepared to reject. Before showing this, I shall formulate the theses that express these conflicting ideas while adding a few remarks.

2 Five theses about norms and action

2.1 Norms determine available options of action

  • 2 Hart 1961: 6.
  • 3 Hart presents this idea in a context in which he takes the law as being made up of rules or genera (...)
  • 4 Metaethical realism generally assumes that norms, insofar as they are linguistic entities, describ (...)

3If norms actually exist, or when a norm exists, it can be said – to paraphrase Hart –2 that the behaviour of certain individuals (those addressed by the norms) becomes non-optional, or non-eligible. That is to say, it cannot be chosen. By virtue of this restriction, the opposite behaviour is, in some way as Hart claims, obligatory or what shall be considered here to be equivalent, its description constitutes the content of a duty.3 This is a minimum, perfectly acceptable nucleus of the idea of “determination” or exclusion of alternatives: norms determine the sense of a choice by constituting (or describing)4 the fact that some actions ought to be performed, or ought not to be performed. Albeit acceptable (and accepted) in general terms, this idea is complex and generates a number of perplexities that I cannot engage here. (One might suppose, for example, a certain risk of circularity, if “duty” is to be defined as “non-eligible”.) However, at least some clarification is called for:

  • 5 I am aware that some might hold that the correspondence is not one-to-one: while “duty” implies “o (...)
  • 6 These distinctions correspond to the Kantian ideas of “hypothetical duty” on the one hand, and “ca (...)

4 The term “ought to” and its correlative “duty”5 are systematically ambiguous, which leaves one of their meanings unlinked to the notion of “norm” (at least, to the notion of “norm” that I shall discuss in this paper). In a “weak” sense, an individual A ought to or is duty-bound to perform action p, so long as p is a means for achieving a goal that A wants to achieve or a purpose that A wants to satisfy.6 In a “strong” sense, A ought to do p regardless of the aims or purposes she wishes to achieve. Or, as can be assumed to be equivalent, regardless of her own options. As the content of the duty in the “weak” sense depends crucially on A’s choices, it cannot be said that her behaviour becomes “non-optional” in this assumption. On this account, it appears to be the “strong” sense that has to be elucidated to account for the idea according to which the actions become non-optional when there are norms. As we shall see, the question is precisely whether, given certain conditions, this way of understanding the notion of “duty” is intelligible.

5 It is true that norms, on the standard conception, are not restricted to constituting or imposing duties or obligations. That is, “to oblige” is not the only normative function. They may also confer upon their addressees the faculty or permission to choose between alternative actions p and not-p. It is obvious, however, that it is not the permissive or facultative norms that are in question when examining the idea of non-eligible behaviours, nor are such norms able to resolve the practical question involved in choosing between p and not-p. In this sense, it cannot be held that they determine the action that ought to be done because here, this depends on the decision of the agent involved. That is why I shall restrict the discussion to those norms that impose or constitute duties. In any case, whether norms would make any sense if all existing norms (or all those belonging to a certain normative system, e.g. “moral” or “the law”) had the only function of allowing all alternatives is an open question.

  • 7 A radical distinction in the ontology proposed by Kelsen, for example. But whichever the conceptio (...)
  • 8 It must be noticed that a belief in the existence of a norm or a duty is also an event. It is in t (...)
  • 9 At least, that is what I have attempted to prove in Caracciolo 1996.

6 Relatedly, we must acknowledge that saying that when norms exist “some behaviours cannot be chosen,” means the same as “they are not allowed” and as “they ought not to be performed”. But in no way does this mean that they are not empirically possible. Strictly speaking, what action agent A will actually carry out is an empirical question that does not depend (at least not directly) on the existence of a norm of which A is the addressee. Therefore, “determination” cannot be understood as the causal or empirical determination of the behaviour. Moreover, the metaphor of the existence of two independent “worlds”, one normative and the other empirical or causal,7 in which actions have occurred, are occurring, and will occur, implies the complete irrelevance of norms with regard to behaviour. Actions only occur due to their relationship with some fact that also belongs to the empirical world.8 In turn, if norms do exist, it should be recognized that they exist regardless of any empirical fact.9 That is, there is in principle no ontological relationship between these two worlds.

2.2 Normativity requires possibility of action

  • 10 Copp 2008.

7This lack of causal relation does not mean that other relations cannot be postulated. Recall the statement according to which it is empirically possible to comply with and also not to comply with any norm. This thesis is usually presented as a conceptual requirement: The idea of “norm” implies that the requirement that makes up some specific normative content must refer to actions that can be done by the agent to whom it is addressed, which also includes the possibility of non-compliance. As a pragmatic conclusion, there is no sense in demanding behaviours that cannot be fulfilled; or as with some moral content, in the sense that no moral reproach is justified if the agent cannot perform the morally required action.10 At the least, because of the relation to impossible actions, the individuals involved cannot consider any practical issues, hence, the generally accepted principle stating that “ought to” implies “can”. If the principle is understood as a conceptual requirement, this means that there is no duty to perform action p, unless p can in fact be performed.

  • 11 Obviously, the same applies to any particular event (see Fisher 2003).
  • 12 I think this distinction captures a perfectly intelligible way of speaking. It could thus be said, (...)

8 There are, of course, numerous open questions regarding the intelligibility and scope of such a principle that cannot be dealt with here. Nonetheless, for the problem I wish to present, it is necessary to indicate a certain ambiguity in the word “action” and, therefore, in the expression “possible action”. If the duty to perform p results from a general norm, “p” must be substituted by the description of a type action, or a generic action, that is, a certain property that defines a set of events. It is admissible to say that a generic action p is empirically possible in the actual world if an occurrence in space and time of some case of p is not ruled out by the empirical laws of the actual world. In this sense – which I shall call possibility1 – any action belonging to the class is possible. But the duty may also refer to a token, or particular action, that will have to be done by a certain individual at a certain moment and in a certain space. That a certain token action – a certain case of the generic action p – may occur, that is, is “possible” at a time t and in space e, does not only depend on empirical laws (i.e., that it belongs to a particular class), but also on the state of the world at time t1. In the absence of the necessary conditions at t1 for p to occur, this particular action cannot be performed,11 that is, it will not be possible in a second sense – which I shall call possibility2.12

  • 13 Trivial cases are those in which a definite agent has a physical or mental impediment which makes (...)
  • 14 See, for example, Zimmerman 1996: chap. 3 and Coop 1997.

9 It can be maintained that the scope of the “ought to” implies “can” principle is restricted to type or generic actions, that is to say, to the idea of possibility1. But this does not seem to be a theoretically interesting restriction. The only actions that exist in the empirical world are token actions. Therefore, general norms are only complied with by means of token actions. So, asking whether a certain agent A can comply with a norm N, that is, satisfy the duty imposed by a norm N, is tantamount to posing the question of knowing whether A can or cannot perform a certain action in the sense of possibility2.13 If one is prepared to admit that in the event of a negative response the duty does not exist, then the principle has extended into the domain of particular actions – it is what happens to the following more or less standard formulation in the environment of metaethics, which I shall adopt in what follows:14

  • 15 There is currently a discussion regarding whether the maxim should include or exclude the possibil (...)

(P) An individual S ought to perform an action p at time t and in space e, only if she can do p [and also omit p] at time t and in space e.15

2.3 Norms imply the existence of reasons for action

  • 16 See the works included in Dancy 2000a. See also Raz 1990.
  • 17 The inverse does not hold, since not all reasons for action result from duties. This is so, even i (...)

10As we know, it is also generally accepted nowadays that the notion of “norm” (or “normativity”) has to be analysed in terms of an idea that is considered more basic, namely, that of the “reason for the action”.16 The reductionist proposal assumes that norms constitute reasons to perform the actions required in their content. Consequently, insofar as this content constitutes duties for action, the statement that there is a duty to do p implies – in keeping with this proposal – that there is a reason to do p.17

11 Three observations are relevant in this respect. First, the idea is not that any requirement of behaviour that can be expressed in normative language leads to a reason for action. For this to be so, there has to be a certain property, called “validity” or “obligating force”, which can be identified with its “normativity”. Indeed, the relationship is the exact reverse precisely because it involves analysing this property in terms of the idea of “reason”. That a norm exists that institutes the duty to do p, can be shown only if a reason to do p exists as a result of the requirement. (Obviously, I cannot take a stand on such an issue here. Nor regarding what the other conditions would be for the existence of norms.)

  • 18 See the important book by Korsgaard 1996.

12 Second, the reductionist proposal can be explained as the result of adopting the idea of rationality as the key to understanding the nature of practical problems. In other words, the correct solution to a question of this kind can only be to perform a rational action. In metaphorical language, the reason is presented as the “source” of the normativity.18 This unified view of possible normative domains is destined to exclude the result – considered to be paradoxical – according to which complying with a duty might just configure an irrational action.

  • 19 Reduction is also a controversial theoretical manoeuvre. As we shall see shortly, this depends on (...)

13 Third, it is clear that if the idea of rationality is to have an impact on the empirical world, where existing actions actually happen, it must be admitted that rational actions are in fact possible. That is to say, we must accept that if an agent A has a reason to do p, then it is possible for A to do p, which means that the principle “ought to” implies “can” can also be indirectly derived by means of the conceptual reduction of the idea of “norm” to the idea of “reason”.19

2.4 Norms objectively exist

  • 20 For example, Hart (1961: 85) says: “Hence it is thought that obligations and duties characteristic (...)

14The existence of norms (if they exist) and their “validity” must be considered objective data in the following sense: they are independent from the mental states of the agents that form the class of their addressees. That is, they dispense with the dispositions of behaviours, beliefs and, especially desires that each of them might have contingently in relation to the actions that make up pairs of reciprocally excluding actions. That is a minimum feature because here the classification of “objective” is opposed to “subjective” and does not lead to any sort of judgment regarding the question of knowing what the other conditions of the existence of norms and duties are. Only the negative condition prevails, according to which one cannot count on certain data from empirical reality for this. It follows from this thesis that duties of action constituted by norms, as well as the corresponding reasons, are objective in the same sense. As mentioned above, it is a feature implied in the notion of “duty” in the strong sense under review in this discussion. The requirement of this independence is, incidentally, a topic that belongs to the tradition of practical philosophy and should not be confounded with an even stronger version of the separation between duty and desire, according to which, far from being independent, one would have to say that “an agent A is duty-bound to do p” implies the absence of A’s desire to do p.20 Such a claim is tantamount to establishing an inverse conceptual connection. The idea of objectivity assumed in this work does not require such a strong condition. According to that idea, the objective character of A’s duty to do p is compatible with the absence of, but also with the presence of, her desire or disposition to do p.

2.5 Action requires motivation

  • 21 Some may argue that mental states are reduced to neuronal states. Once again, however, this compli (...)

15So far, the propositions have reflected a certain conception of norms. Now, within the domain of the philosophy of action, according to the standard conception, one must admit that there is no token action without motivation where by “motivation” we must understand some mental or psychological state attributed to the agent involved in the action. The absence or presence of motivation thus presents itself as a conceptual requisite to distinguish between what an agent “does” – the action she performs – from what “happens” to her without her intervention. On the other hand, as actions are also facts in the empirical world, these psychological states with which motivations are identified are also assumed to be facts without which the occurrence of a certain action cannot be explained.21 Put differently, motivations represent necessary conditions of the empirical possibility of token actions. Therefore, according to this conception, no single action can be performed in the sense of possibility2, in the absence of its corresponding motivation.

  • 22 Smith 1987.

16 Furthermore, as we also know, the general conception admits Hume’s vision according to which motivations are pairs of mental states made up of a desire and a belief: the desire to achieve a certain state of the world and the belief that a certain action is a sufficient means to satisfy this desire. Moreover, beliefs and desires are assumed to be independent states in that no desire – whatever its content – implies the existence of any belief – whatever its content – and vice versa.22 So, according to this assumption, we must discard the alternative of assuming that the belief that p ought to be done implies the desire to do p and vice versa. I am well aware that this is an extremely concise presentation of a complex issue, but I believe it is sufficient for what I wish to show below.

3 The dilemma

17Let us now assume that the following formulation represents a norm:

(N) All individuals belonging to class A ought to perform an action belonging to class P at any time included in interval T.

  • 23 See a similarly structured argument, in another context, in Fischer 2003.

18It seems we must admit, from N, that an individual’s duty to perform a token action p, which belongs to P type actions, and at a certain time t included in interval T, results from her being included in class A. Therefore, according to the theses presented above, the argument can be reconstructed as follows:23

(1) If “a” ought to do p at t, then “a” has a reason to do p at t.

(2) If “a” has a reason do p at t, then “a” can do p at t.

(3) If “a” can do p at t, then she has at t a motivation to do p at t.

(4) If “a” has at t a motivation to do p at t, then she has a desire to do p at t.


From which it follows that:


  • 24 There may be complications in determining the time the principle refers to, that is, the moment at (...)

(5) If “a” ought to do p at t, then she has at t a desire to do p at t.24

19Propositions (1) to (5) imply that if “adoes not have the desire to do p at t, then she is not duty-bound to do p at t. This conclusion is manifestly irreconcilable with the objectivity postulated for norms and, therefore, with the objectivistic conception of duty presented above. Moreover, if we accept this argument, we eliminate the distinction drawn above between “duty” in the weak sense and “duty” in the strong sense, and with it the basic idea of “norm” associated with excluding alternatives. Nevertheless, the argument is presented as a direct consequence of the so-called “ought to” implies “can” principle. Therefore, the set of theses presented earlier appears to lead to a genuine dilemma: either norms (and the duties constituted by these norms) are not objective, or the idea of “norm” does not imply the possibility of compliance. It should be noted that premises (1) to (4) of the argument above only propose to specify a certain conception of norms and a certain conception of actions. So, a revision of the content of these ideas may lead to the dissolution of the dilemma. Below, I shall explore some options to this effect. Nonetheless, I shall conclude that none of these alternatives is satisfactory.

4 No way out?

20Considering that, by hypothesis, all the theses that have to be judged are at first sight perfectly acceptable, it is by no means clear which one would have to be modified, clarified, or dropped.

4.1 Specifying the concept of reason for action

21In the first place, it might be necessary to specify the scope of the proposed reduction of the idea of “normativity” assumed in premise (1),

(1) If “a” ought to do p at t, then “a” has a reason to do p at t.


because a correct elucidation of the notion of “reason for action” should block the way to the problematic conclusion. Indeed, in this premise one must assume a certain idea of what reasons are. Therefore, if the idea in question is erroneous, it would have to be modified. It soon becomes clear, however, that such a step would only transfer the problem and has no resolving effect on the dilemma itself because there are two conceptions that are reciprocally incompatible with regard to the nature of reasons for action.

  • 25 Williams 1981.

22 On one hand, there is the internalist conception of reason for action with a thesis – common to many versions –, according to which the propositions that state the existence of a reason for action, have to be directly or indirectly relativized to what Bernard Williams called the subjective motivational set: for any agent A, A has a reason R to do p only if p promotes or satisfies a desire, or disposition or any other actual mental state of A’s capable of leading her to action. If this is not so, the corresponding proposition is false.25

  • 26 Obviously, a simplified version of a complex discussion is presented here, about which there is a (...)

23 On the other hand, the conception called externalism, enunciated in a general manner, denies internalism in that it supports the independence of the reasons with regard to the contingent subjective components of the agents that have to act. One therefore admits that an agent A may have some reason to act whose existence does not depend on any desire or disposition of A’s.26

24 However, if one assumes the hypothesis according to which norms (and their corresponding duties) are objective in the sense presented above, and an internalist version is adopted regarding the reasons for action, premise (2) must be considered true

(2) If “a” has a reason do p at t, then “a” can do p at t.

  • 27 See a discussion found, for example, in Fischer 2003.
  • 28 Of course, nor does rejecting the reduction of the idea of “duty” to the idea of “reason for actio (...)


because saying that there is a reason (in relation to an agent, or set of agents) implies stating the existence of a motivation (relativized in the same way). But premise (1) is false: the existence of an objective duty does not here imply the existence of a reason. The reductionist proposal must therefore be abandoned. Inversely, if the externalist version is adopted, premise (1) may be considered true, as a result of reducing the idea of “duty” to the idea of “reason”, but premise (2) is false because the existence of a reason does not imply that the corresponding action can be performed. This is what some theorists – those who have to deny, or at least question, the principle “ought to” implies “can” – maintain.27 It is of course obvious that the need to opt for one of the conceptions reproduces the problem: internalism maintains the possibility of action as presented in Section 2.2, and externalism the objectivity of duty from Section 2.4. Consequently, the dilemma is not eradicated by resorting to specifying the concept of “reason for action”.28

4.2 Dropping the motivation requirement for action

25Another alternative is to focus on the very content of the principle “ought to” implies “can” and deny premise (3) in the first place:

(3) If “a” can do p at t, then she has at t a motivation to do p at t.

  • 29 Coop 2008.

26This negation seems to follow if the discussion regarding the possibility of performing a token action is confined to the existence of a contextual opportunity and the physical ability of the agent involved.29 It can thus be said that the action constituting the content of the duty can be performed, even in the absence of a relevant motivation.

  • 30 Zimmerman (1997: 87–90) also says that the uncontrollable desire to perform the right duty, which (...)

27 However, the connection between the principle in question and what Michael Zimmerman calls “psychological impossibility” is by no means new in philosophical literature related to the nature of moral duty, which may of course be extended to any other normative domain. This is what happens to the objection trained by David Ross on the Kantian idea of duty, in that for its fulfilment it requires the agent to act for a “correct motive” because the absence of this mental component in an agent, when doing the action, would make it impossible for the agent to satisfy the moral requirement. The same conclusion is reached in all cases in which the content of a duty demands a specific motivation (this discussion is in Zimmerman 1996).30 The conclusion is that – if the principle is accepted – in these cases there is no duty to do the action in question.

  • 31 It does not matter here whether the motivation is “good” or that using some criterion it might be (...)

28 Since all of this results from the same conception of action assumed here, the only thing needed to accept conclusion (5) is to generalize the argument to all possible contents of a duty and to all of the motivations. Since no action is possible without its corresponding motivation, there is no way to comply with a duty if that motivation, that is to say, a certain desire, does not exist.31 Beyond that, if this conception is admitted, any proposal to eliminate the psychological components in the formulation of the principle “ought to” implies “can” is notoriously irrelevant because in the actual empirical world there are no unmotivated actions. Opportunity and physical ability may be considered necessary conditions for the possibility of token actions, but they are never sufficient.

4.3 Dropping the current desire requirement for having a motivation

29As part of the process of discussing the scope of the principle “ought to” implies “can”, it is worth pointing out that the manner in which premise (4) and conclusion (5) are formulated and understood may also be objected to:

(4) If “a” has at t a motivation to do p at t, then she has a desire to do p at t.

(5) If “a” ought to do p at t, then she has at t a desire to do p at t.

  • 32 This is an objection formulated by José Luis Martí when discussing these ideas at the Oñati meetin (...)

30It may be said that the action is possible – that is, “can” be performed – if the agent can have the necessary motivation, i.e. if she can desire to do the action.32 That is to say, even in the absence of the current desire at time t, so long as the desire is possible, so too is performing the action. This way out can be understood in two ways, but nevertheless, I see both alternatives as insufficient to eliminate the dilemma.

31 In the first sense, what is meant is that if the agent that is the addressee of a norm had had the desire at time t to do the required action p, p could have been performed at t. Certainly, the counterfactual may be true if the occurrence of p – and the corresponding desire – is not incompatible with empirical laws. However, this can only show that p is one case of a type or generic action possible in the sense of possibility1, from which nothing can be inferred regarding the current performance of a token action at time t. This is because – according to the concept of action assumed – it is also necessary for the agent to have a current desire to do p, for p to happen at t.

  • 33 See a discussion on the logic of this type of argument in Smith 1994: 142 and ff.

32 The other way of understanding the objection states that the current action can be performed if the agent can currently desire to perform the action. However, for this version to differ from the former, it has to be accepted that desires can be chosen, i.e. that they are or may be the result of some sort of mental action (which is debatable according to Hume). But then to be able to have the desire to perform an action it is necessary to have another desire – a second order desire – to acquire the desire to carry out the action in question.33 Thus, by this means the problem is just transferred and the dilemma persists, because now those who have the duty to do p are those who have the current desire to desire to do p.

4.4 Dropping any desire requirement for having a motivation

  • 34 For a radical attack on identifying motivation with desires see Dancy 2000b: especially chap. 4, a (...)
  • 35 See Raz 1990: 73 and ff.

33Another challenge available to us results from abandoning the Humean conception and therefore denying that the existence of a motivation must imply, on any occasion, the existence of a desire. Therefore, if this is admitted, premise (4) of the argument would have to be considered false.34 It can also be assumed that beliefs can constitute mental states that lead autonomously to action or, also in this direction, that certain beliefs imply (the inverse of Hume) the existence of a desire. The relevant candidate here is of course the mental state consisting in an agent’s belief regarding what she ought to do. So, one would have to say that if A believes she ought to do p, then she has a motivation to do p. It may be considered, just as Raz indicates,35 that one who accepts norms is one who believes that norms are valid, that is, one who believes she ought to act in keeping with the content of norms, which implies – as a result of this belief – a disposition to behave as required.

34 Nevertheless, it is evident that this change also fails to alter the result that leads to the dilemma. Because although (4) is altered to (4*),

(4) If “a” has at t a motivation to do p at t, then she has a desire to do p at t.

(4*) If “a” has a motivation to do p, either she has the desire to do p, or else believes she ought to do p.


all those who do not have the desire to do p, and do not believe they ought to do p, are excluded from the scope of N, that is, from all those who have the duty to do p. Those obliged – those with the duty to do p – would therefore only be the acceptants, or those who desire to do p regardless of norms, which is certainly incompatible with the objectivity proposed because here too the existence of the duty depends on a certain mental state. It is no wonder then, that the dilemma should still stand because it is quite clear that for any mental state which is identified with the existence of a motivation, there will or might be some individual who lacks this particular mental state.

  • 36 Once again, for the idea of moral duty, see Zimmerman 1996.
  • 37 According to Hart, those who adopt the “external point of view” i.e. those who act motivated by th (...)

35 It is remarkable that in order to save the prevalence of the possibility of complying, a subjectivist version of the notion of “duty” has to be adopted, which is in no way new to metaethics or to the philosophy of law. According to this version, the content of duty has to be determined by an agent’s belief regarding what she ought to do.36 So, if A believes that she ought to do p, then she ought to do p and vice versa. If desire is not taken into account, this is a theoretical response to verifying the fact that any individual can act only in accordance with what she believes she ought to do.37

4.5 Redefining rational agency

  • 38 In relation to morals, see Smith 1994 and Korsgaard 1996.

36The last point I shall consider in this review is the following: in current literature there is an argument according to which, bearing in mind the constitutive relationship postulated between normativity and reason – that is, between norms and reasons –, it cannot be the case that perfectly rational agents lack the motivation necessary to act in keeping with a duty, i.e. according to a reason, even though reasons are objective. This means that it is conceptually necessary for perfectly rational agents to have the desire to do rational actions.38 It is not necessary to take a stand here on this complex argument in order to notice that it cannot affect the structure of the dilemma either. This is because according to its content, agents who are not perfectly rational are precisely those who do not have the desire to act in keeping with reason. Or, synonymously, only perfectly rational agents are those who are in a position to do the actions required by the reason. As for the rest, if norms are objective, it would not be true that “ought to” implies “can”.

Top of page

Bibliography

Ricardo CARACCIOLO, 1997: Existencia de normas, Isonomía, No. 7 (October 1997): 159–178. URL: http://www.cervantesvirtual.com/obra/existencias-de-normas-0/.

David COOP, 1997: Defending the principle of alternate possibilities: blameworthiness and moral responsibility, Noûs, Vol. 31, No. 4: 441–456. DOI: 10.1111/0029-4624.00055.

David COOP, 2008: “Ought” implies “can” and the derivation of the principle of alternate possibilities, Analysis, Vol. 68, No. 1: 67–75. URL: www.jstor.org/stable/10.2307/i25597841.

Jonathan DANCY, 2000a (ed.): Normativity, Oxford, Blackwell.

Jonathan DANCY 2000b: Practical reality, Oxford, Oxford University Press.

John M. FISCHER, 2003: “Ought-implies-can” causal determinism and moral responsibility, Analysis, Vol. 63, No. 3 (July 2003): 244–250. DOI: 10.1093/analys/63.3.244.

Harry G. FRANKFURT, 1969: Alternate possibilities and moral responsibility, The journal of philosophy, Vol. 66, No. 23 (December 1969): 829–839. DOI: 10.2307/2023833.

Herbert L.A. HART, 1961: The concept of law, Oxford, Oxford University Press.

Christine M. KORSGAARD, 1996: The sources of normativity, Cambridge, Cambridge University Press.

Joseph RAZ, 1990: Practical reason and norms, Princeton, Princeton University Press.

Michael SMITH, 1987: The Humean theory of motivation, Mind, Vol. XCVI, No. 381 (January 1987): 36–61. DOI: 10.1093/mind/XCVI.381.36.

Michael SMITH, 1994: The Moral Problem, Oxford, Blackwell.

Bernard WILLIAMS, 1981: Internal and external reasons, in Bernard Williams, Moral luck, Cambridge, Cambridge University Press.

Michael ZIMMERMAN, 1996: The concept of moral obligation, Cambridge, Cambridge University Press.

Top of page

Notes

1 By “general” in this context I understand an idea or thesis that can be assumed by different alternative theories, divergent in other aspects.

2 Hart 1961: 6.

3 Hart presents this idea in a context in which he takes the law as being made up of rules or general guidelines. Whether “duties” can exist without general norms is debatable. In any case, what follows is independent upon this controversy and greatly affects also those who hold that duties are always “particular”.

4 Metaethical realism generally assumes that norms, insofar as they are linguistic entities, describe pre-existing duties, while an anti-realistic conception must assume that duties are constituted by norms. It is not necessary to take sides on this for the discussion that follows.

5 I am aware that some might hold that the correspondence is not one-to-one: while “duty” implies “ought to”, the inverse does not hold. Kelsen for example, held that “ought to” is an indicator of “normativity”, compatible with other normative functions, but I shall also overlook this complication here.

6 These distinctions correspond to the Kantian ideas of “hypothetical duty” on the one hand, and “categorical duty,” on the other. But this terminology may lead to confusion when one considers that some duties are conditional, albeit “categorical” in Kant’s sense.

7 A radical distinction in the ontology proposed by Kelsen, for example. But whichever the conceptions adopted regarding norms, it seems that it is not possible to maintain that norms determine behaviour causally, unless they are identified with empirical entities. At least, I know of no argument in this respect.

8 It must be noticed that a belief in the existence of a norm or a duty is also an event. It is in this sense that norms – unless they are identified as an empirical fact – are said not to be able to directly determine behaviour.

9 At least, that is what I have attempted to prove in Caracciolo 1996.

10 Copp 2008.

11 Obviously, the same applies to any particular event (see Fisher 2003).

12 I think this distinction captures a perfectly intelligible way of speaking. It could thus be said, for example, that rain is a possible event, but that it is impossible for it to rain today in Barcelona. This can be correct if “possible” is understood in two different ways. The counterfactual proposition that sates that if the atmospheric pressure had risen, it would have been possible for it to rain in Barcelona would also be true, if it is the case that a certain level of pressure constitutes a causally necessary condition for rain, which corresponds to the use of counterfactual statements to express causal relationships. According to this suggestion, that an event may be possible2 implies a possible1 class of events, but the inverse relationship does not hold. The need to make this distinction comes from an objection made by Hugo Seleme during a personal discussion, to whom I give thanks.

13 Trivial cases are those in which a definite agent has a physical or mental impediment which makes it impossible to do the required action, or does not possess the necessary skill to act in the manner required.

14 See, for example, Zimmerman 1996: chap. 3 and Coop 1997.

15 There is currently a discussion regarding whether the maxim should include or exclude the possibility of omission (called the “principle of alternative possibilities”) from Frankfurt’s (1969) objection to the thesis that this is necessary to attribute moral responsibility. However, for the argument presented here, there is no need to take a side on this issue.

16 See the works included in Dancy 2000a. See also Raz 1990.

17 The inverse does not hold, since not all reasons for action result from duties. This is so, even if we accept Raz’s statement that affirming the existence of a reason to do an action implies that p ought to be done (Raz 1990: chap. 1, cf. footnote 1).

18 See the important book by Korsgaard 1996.

19 Reduction is also a controversial theoretical manoeuvre. As we shall see shortly, this depends on what is understood by “reason for action”.

20 For example, Hart (1961: 85) says: “Hence it is thought that obligations and duties characteristically imply sacrifice and renunciation” (that is, sacrifice and renunciation of one’s own desire or interest).

21 Some may argue that mental states are reduced to neuronal states. Once again, however, this complication is not relevant for this paper.

22 Smith 1987.

23 See a similarly structured argument, in another context, in Fischer 2003.

24 There may be complications in determining the time the principle refers to, that is, the moment at which it is stated that the agent can do the action and its relationship with the actual time of the action. See a discussion of this in Zimmerman 1997: chap. 2. It is assumed in the text that the desire must exist at the moment of the action. But, I do not think the argument in this paper depends on taking a position on this matter because whatever the time at which the desire is supposed to exist for a certain agent to be able to do an action, the agent in question may lack the desire at that particular time.

25 Williams 1981.

26 Obviously, a simplified version of a complex discussion is presented here, about which there is a vast literature, but I believe this presentation is enough to support what I intend to show in this paper. One way or another, if reasons are identified with facts, as Raz (1990) does, it must be admitted that there are “external” reasons. Dancy (2000b) also represents externalism.

27 See a discussion found, for example, in Fischer 2003.

28 Of course, nor does rejecting the reduction of the idea of “duty” to the idea of “reason for action” dissolve the dilemma, because according to the second proposition relative to the nature of norms, duty implies the possibility of action in all cases.

29 Coop 2008.

30 Zimmerman (1997: 87–90) also says that the uncontrollable desire to perform the right duty, which implies the impossibility of omitting the action, does not count as complying with a duty because the maxim assumes the possibility of not complying.

31 It does not matter here whether the motivation is “good” or that using some criterion it might be considered “bad”.

32 This is an objection formulated by José Luis Martí when discussing these ideas at the Oñati meeting in September 2007. I thank José Luis for the warning regarding the need to consider this alternative.

33 See a discussion on the logic of this type of argument in Smith 1994: 142 and ff.

34 For a radical attack on identifying motivation with desires see Dancy 2000b: especially chap. 4, although the attack is also directed at any form of what Dancy calls “psychologism”. Along these lines, one could also say that actions are events with no connection whatsoever to mental states. I shall not discuss this alternative, although I do believe it is implausible because it fails to explain the sense in which certain events would have to be considered actions.

35 See Raz 1990: 73 and ff.

36 Once again, for the idea of moral duty, see Zimmerman 1996.

37 According to Hart, those who adopt the “external point of view” i.e. those who act motivated by the threat of undesired consequences, do not follow norms or rules, that is, they do not act according to what they believe they ought to do. The consequence of the subjectivist argument is stronger in this direction, because it follows that those who do not accept the norm that constitutes, for example, the duty to do p, that is, those who do not believe they ought to do p, ought not to do p.

38 In relation to morals, see Smith 1994 and Korsgaard 1996.

Top of page

References

Electronic reference

Ricardo Caracciolo, « A dilemma regarding the nature of norms », Revus [Online], in print | 2018, Online since 13 July 2018, connection on 22 July 2018. URL : http://journals.openedition.org/revus/4141 ; DOI : 10.4000/revus.4141

Top of page

About the author

Ricardo Caracciolo

Professor of Philosophy of Law, National University of Cordoba, Cordoba (Argentina)

Address: Calle Quiroga Sarmiento 4590 – Barrio Alejandro Centeno – X5009ENL Córdoba – Provincia de Córdoba, Argentina

E-mail: ricardocaraccio@yahoo.com.ar

Top of page

Copyright

All rights reserved

Top of page
  • OpenEdition Journals