From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

Actions, situations and reasons

In an ordinary language a reason for action is any consideration of an agent that counts for his or her undertaking a given action. In philosophical analyses, however, some additional conditions are introduced making the meaning of the term more precise. The description of human actions in terms of their reasons has some important advantages, and particularly for moral philosophy, in which not explanatory but normative reasons are the primary focus of attention. Firstly, reasons for actions can be derived from (or generalized into) some rules of increasingly general character, and finally they can be placed within a theory of practical rationality. Secondly, if morality provides us with special kinds of reasons for actions, their importance can be understood and evaluated against the background of practical rationality in general. Thirdly, if there is no grounding of morality and rationality in the external word, the foundation for them can be found within the internal procedures of the acting subject or can be achieved by an agreement of different persons.

Another important advantage is closely related to XX century metaethics. As its history has clearly shown, moral judgments have proved to be too difficult to allow for one dominant and convincing analysis. According to the compelling argument of M. Smith, this is so because moral judgments have two important characteristics that are very difficult to square: ‘objectivity’ and ‘practicality’ (Smith 1997). ‘Objectivity’ means that in our everyday moral arguments moral terms are used as referring to some objective ‘moral facts’. The ‘practical’ character means that if an agent sincerely formulates a moral judgment, he is also motivated to act in accordance with it. Subjectivism, emotivism and other forms of non-cognitivism have correctly described the ‘practical’ character but they have overlooked the ‘objective’ character of moral judgments. Intuitionism and other forms of cognitivism, in turn, have correctly recognized the ‘objective’ character but they have lacked the explanation of their ‘practical’ character.

There has been an interesting debate on the concept of a reason for action over the last few decades. The stage for the debate is set by three main problems (Cullity et al. 1997). One of them concerns the relation between a reason and the agent’s prior motivation. Are reasons for action dependent on a prior set of desires and emotions of the agent or not? The second issue concerns the nature of a reason for action. Does a reason recognize some external normative entities or is it constructed by the agent in accordance with its internal criteria and procedures? The third issue refers to the level of generality that can be ascribed to the reasons: are reasons of a particular or of universal character?

The contribution I would like to make to the debate in this paper is mainly related to the opposition between particular and universal reasons. I hope that it will throw some more light and bring some more order into this part of the debate. The starting point for my consideration will be the following questions: Is there an order within the sphere of the reasons for action? And if there is, what kind of order is that?

Let us outline an analysis of human action at the beginning. Any such an analysis should contain at least the following elements: an agent, a situation in which the action takes place, the intention of the action, the act itself, and the results that it brings about. Of course this description can be extended by introducing more detailed elements but this outline will be sufficient for our present purposes. One more point is only necessary: we assume that the intention of an agent is captured by his or her reason for action.

Now can there be a unique, purely particular reason for action, as some authors hold? The hidden assumptions underlying any possible answer to this question refers to the concept of the situation in which the action takes place. These assumptions almost never have been formulated explicitly. I shall attempt to show how these hidden assumptions affect the concept of the reason of action, and, on a more general level, the whole debate between different moral theories.

A situation can have some unique morally relevant properties but it can also have some other properties that, though rare, are not exceptional. And of course a situation can have many properties that are quite common and shared with lots of other situations of the same type. There are probably all these types of situations in our lives but the exponents of different ethical theories usually focus their attention on one of them. Exponents of moral particularism implicitly assume that there is no regular order among the situations of our action and, consequently, that no generalization of the particular reasons of action is possible. Of course this does not mean that an agent does not use its reason in its deliberation as to what to do. What it means is that an agent does it in such a way that it does not allow for any generalization. Moral particularism is an extreme position. In some more moderate positions it is assumed that there is much more order in the situations of our actions and that some typical situations can be distinguished. As a result of this strategy a normative content can be formulated with regard to each type of situation. A paradigm example of this strategy can be found in the works of Aristotle, and among the authors of XX century it can be traced in W. D. Ross and Thomas Nagel. Virtues, duties and reasons for action are of a general character but their generality is limited to a given type of situation. Of course within the framework of contemporary deontology insoluble conflicts between duties and reasons for actions are possible.

A different aspect of the order within the realm of the reasons for actions can be related to the question as to whether there are general rules or principles under which our reasons fall, which resembles the regularity described by the laws of nature. This is apparently a higher form of order than that introduced by the existence of general reasons for action or duties. Deriving his reasoning from Wittgenstein writings on following the rule, John McDowell argues that our reasons for action do not fall under a system of laws that resembles the laws of nature (McDowell 1994). Some other authors from the communitarian camp argue that there is an order in the domain of our reasons for action but that our reasons should be placed within ‘the narrative stories’ of our lives and not in a system resembling the system of the laws of nature.

If we assume that our universal reasons for actions fall under general laws and that these laws constitute a consistent system then we arrive at universalistic positions in ethics, in which the highest possible order of reasons is present. In Kant’s classical position, all our reasons for actions falls under ‘maxims’ and, in the case of moral reasons, ‘maxims’ have a common logical form, which is the famous ‘moral law’. Moral reasons in Kant’s system have always overriding force over other, non-moral reasons and no insoluble conflicts among moral reasons are allowed. The assumed order and regularity within the situations of all our action are so high that a logical form of duty – the categorical imperative – is valid for all of them. Mutatis mutandis, a different attempt to establish such a high order can be found in the position of rule utilitarianism. All those underlying assumptions can be traced in the ways of the conceptualization of the situations of actions.

Moral particularism versus a hybrid theory

In the last part of this paper I shall illustrate my ideas very briefly by applying it to the position of J. Dancy and T. Nagel. Dancy is a distinguished exponent of recent moral particularism and the detailed and careful discussion of the concept of the situation of action appears explicitly in his writing (Dancy 1993). He vigorously argues that every situation of our actions is of particular character but at the same time it has some objective moral properties. The ontological relation between natural and moral properties is a ‘holistic’ relation of ‘resultance’. There are no simple correlations between natural and moral properties as any moral value has a ‘resultance base’ of individual ‘shape’. A strong pain – to take an example – usually counts as the reason against an action that is its cause, but there are some situations, in which it does not. Dancy defends a clear form of relation between moral values and reasons for actions: he is a moral realist (a strong cognitivist) both with regard to values and reasons for action. There are non-natural moral values that supervene over natural properties of situations and they generate moral reasons for actions. What we do in our judgments is we recognize these values and our reasons and morality are based on them. As they always are of a particular character no generalization is possible.

Two questions at this point seem natural. Does not the rationality of our action demand some more regularity? Is it enough to state that an act is done for a reason? But Dancy has an answer to these questions. He argues that his position fulfills the condition of rationality in judgment, that is the condition of logical consistency. Yet this condition does not consist in a subsumption of a given reason under a general rule. Deriving from Wittgenstein’s writings he argues that the condition at issue is of a different character: it is based on the judgment “that the new case is relevantly similar to the previous one” (Dancy 1993, 83).

Much more ordering in the sphere of reasons (and situations as a consequence) has been assumed in the position of Thomas Nagel (Nagel 1970 and 1986). In his thinking any reasons for action must by of a general character, which by implication means that there must be some general types of situation. (Nagel himself does not discuss explicitly the situation in which an action takes place). Here, the condition of minimal rationality clearly consists in an identification of given reason as an example of a general type.

The idea of a reason for action that Nagel consequently defends is general in two senses. In one sense, it is general with regard to a person: if something is a reason for one person then it also must be a reason for any other person. In another sense it is general in respect to what counts as a reason for an action: if a property counts as a reason then any other property of the same type must also count. Let us take an example: someone has some bodily sensations and is afraid of serious illness. They can, of course, motivate the person to visit her doctor, but do they count as the reason to do it? In Nagel’s thinking they do, but only if they are general: they count in the same way for every agent and any other sensation of the same type counts in the same way. It is clear that the order in the realm of our actions (and the situations) that is implied by Nagel’s idea of practical rationality is higher that it appeared on the surface. It is worth observing that it is higher than that stipulated by Hare’s idea of universalizability, as the latter makes allowances for some differences of agents’ characters.

Although all reasons are general in a double sense, Nagel does not require that they must fall under general laws, neither is the order in the realm of reasons sufficient to ground morality. In Nagel’s thinking, for a grounding of morality one more condition is required, that is one referring to the idea of objectivity with regard to reasons for action. Finally, morality is grounded in the sphere of ‘objective’ reasons, and these are reasons that can be recognized and endorsed from the ‘impartial’, objective point of view. The tendency to reach such an objective point of view is one of the important features of the human mind. There can be no doubts that the recognizing of ‘objective’ reasons depends on the rational nature of our agency. But it is not clear whether an agent discovers the ‘objective’ reasons or construes them, in other words, is Nagel’ s stance a kind of moral realism (strong cognitivism) or just a kind of metaethical constructivism (weak cognitivism)? On the one hand, Nagel apparently does not consider the ‘objective’ reasons and values as a part of an external world, on the other he does not explicitly, offer any constructivist procedure either.

Ethical theory in Nagel’s thinking is a theory on how to generalize all valid reasons for action. What Nagel tries to achieve is a complex, ‘hybrid’ theory which accommodates both the reasons for action coming from consequentialism (valid for all agents) and the reasons coming from deontic constrains and personal commitments (valid for particular agents) (Thomas 2009). What is more he also attempts to find a space for ‘subjective’ reasons coming from our personal projects and choices. One can obviously have some serious doubts as to whether such a synthesis can ever be achieved but one can be sure that Nagel is trying to do justice to the complexities of our lives as far as possible.

Although much work still remains to be done, the idea of grounding morality in the domain of our practical rationality seems both interesting and promising. J. Mackie argued that all our moral judgments are false. Should there be objective values they would have to be very strange entities, different from everything in the universe and motivating us at the same time (Mackie 1977). A normative theory of reasons for action has resources to answer Mackies’ objections: reasons are built into the nature of our action and they do motivate us if we rationally recognize them.^*