Dreyfus & Dewey on Mature Moral Agency
Abstract: Mature moral agency is conceived by most dominant normative ethical theories as involving mature moral agents that reason with moral principles in order to form moral judgments. This conception of moral maturity is normatively anchored by the tenets of rationality: it consists in and thus, is achieved through a process of principled moral deliberation in accordance with norms of good reasoning. In this paper, I argue that this conception of moral maturity fundamentally misconstrues that in which mature moral agency consists and thus, the means by which it is achieved. Accordingly, in the first section, I argue that we must abandon the idea that moral principles can provide moral guidance for and impose normative authority on mature moral agents. Then, in the second section, I develop the insights of Hubert and Stuart Dreyfus and John Dewey into the outline of an alternative conception of mature moral agency.
The conception of mature moral agency upheld by most dominant normative ethical theories is one in which mature moral agents engage in a process of moral deliberation that is guided by moral principles in order to arrive at moral judgments that involve both evaluations of the situation at hand and decisions about the appropriate action to take. This conception of moral maturity is normatively anchored by the tenets of rationality: it ideally consists in and, thus, is achieved through a process of principled moral deliberation (i.e. moral deliberation guided by principles). In this paper, I argue that this conception of moral maturity fundamentally misconstrues that in which mature moral agency consists and thus, the means by which it is achieved. First, I argue that we must abandon the idea that moral principles can provide moral guidance for or impose normative authority on the lives of mature moral agents. I then develop the insights of Hubert and Stuart Dreyfus (hereafter 'Dreyfus') and John Dewey into the outline of an alternative conception of mature moral agency.
If moral maturity consists of the ability to deliberate with moral principles, then mature moral agents are those whose judgments and actions are explicitly guided by such principles. In this section I argue that there are several problems with this conception of mature moral agency: 1) moral principles fail to provide anything more than "entry level" moral guidance and are, therefore, incapable of producing mature moral judgments, and 2) the role of principled moral deliberation in the actual generation of mature moral judgments is misrepresented. Moral principles, I argue, are the products of mature moral judgments. They are not employed in the process by which these judgments are achieved.
Problem 1: Moral Principles ~ Normative ethical theories typically specify moral principles that can (and, in fact, must) be employed by moral agents in order to yield appropriate moral judgments and determine appropriate action. Now, such principles come in at least two distinct forms: formal principles (i.e. moral principles without specific action-guiding content) and substantive principles (i.e. moral principles with specific action-guiding content). In what follows, I consider substantive principles first, then formal. Both forms of principles, I argue, are incapable of providing the sort of moral guidance necessary to achieve mature moral judgments (i.e. judgments that are appropriate for the situations they are judgments of).
Substantive principles provide moral agents with direct, unambiguous, action-specific moral guidance. Examples of such principles are: 'Never lie' or 'Thou shalt not kill'. These principles take the form of rigid, context-less rules. Such rules function to decompose situations into basic feature/response patterns (i.e. certain features require certain responses) that are recognizable even for those who lack experience.
The problem is that although substantive principles provide specific instructions on what one ought to do, such principles are oblivious to the complex details of particular moral situations – details that, in the end, play a role in determining the appropriateness and justification of moral agents' judgments and actions. Because of their inability to adapt to context, substantive principles greatly restrict the moral agents' capacity to recognize and respond to the situation-specific characteristics of particular environmental circumstances: they reduce the range of genuine response options. For instance, while it seems generally correct to say that lying is morally reprehensible, it is possible (if not plausible) that situations might arise in which it would be morally acceptable, if not praiseworthy. Substantive moral principles do not give moral agents the means to deal with these cases, which are numerable, and so moral agents are unable to use them to make appropriate moral judgments and determine appropriate action.
Of course, one might suggest that this problem is solved by the introduction of situational ceteris paribus clauses, thereby creating conditional substantive principles. I think that up to a (very limited) point, this is exactly what happens. But, the number of qualifications necessary to produce moral judgments that are appropriate to the actual situations they are judgments of would quickly render the principles in question unmanageable. Thus, the usefulness of such principles has a narrow window: as Dreyfus writes, "Like the training wheels on a child's first bicycle, these first rules allow the accumulation of experience, but soon they must be put aside to proceed."
One solution to this problem is to introduce content-less moral principles: i.e. formal principles. Examples of purely formal principles are: A moral agent must always strive to "uphold justice" or "maximize utility". The difficulty with formal principles is that they are only useful for identifying and evaluating moral issues at a very abstract level. At best, they orient moral agents by pointing in the appropriate direction. That is, they provide a perspective from which the situations moral agents face should be considered. As an example, the "maximize utility" perspective makes maximizing utility the primary orienting objective, organizing evaluation so that only the issue of maximizing utility is salient, thereby eliminating other issues from consideration.
The problem is that such principles are unable to provide information about what actually would, in a given situation, uphold justice or maximize utility. That is, formal principles fail to provide moral agents with the tools necessary to evaluate the complex details particular situations confront them with, or to develop appropriate moral solutions. In addition, reliance on a single perspective, by itself, often leads to moral dilemmas: e.g. maximizing utility by saving ten lives through sacrificing (or enslaving) one.
Moral principles are like the tips one finds in books on chess when first taking up the game. Such tips are attempts to distill the knowledge possessed by chess masters into a set of basic rules. As helpful as these rules are to a novice, it would be a mistake to think that the games played by masters are guided by them – explicitly or otherwise. In fact, it would be impossible for masters to play a master-level game of chess if they were. Likewise, it is impossible to produce mature moral judgments and actions using moral principles, because they do not possess the sensitivity or flexibility necessary to respond appropriately to the complexities of particular situations. As a result, moral principles, whether formal or substantive, prove incapable of providing the moral guidance necessary for mature moral agency.
Problem 2: Moral Deliberation ~Moral psychologists largely accept the conception of mature moral agency I'm arguing against. That is, they assume that mature moral agents both can and should deliberate with moral principles to form judgments and determine appropriate action. Currently, principled moral deliberation is considered to be an indicator of moral maturity. In this vein, Lawrence Kohlberg's work has been taken as empirical evidence that moral development can be tracked according to the adequacy of principled moral deliberation. Yet, by requiring moral agents to evaluate hypothetical moral dilemmas and give explicit reasons for how and why they formed their judgments after the fact, Kohlberg's methodology misconstrues the role principled moral deliberation normally plays in moral evaluation. It does so by: 1) giving moral agents abstract written scenarios to analyze (scenarios that leave out most of the particular details that are relevant in real-life moral situations) and 2) asking those agents to articulate reasons for their moral judgments, something which naturally leads to a retrospective appeal to moral principle(s) (e.g. "Why shouldn't the man take the drug? Because that is stealing and people shouldn't steal"). But, this does not mean that they actually employed these principles in a deliberative process during the evaluation itself. In fact, far from it. There is much evidence to suggest that people commonly engage in post hoc reasoning when articulating reasons for judgments and actions that had already occurred.  As Dreyfus suggests, the more mature moral agents' judgments are, the more likely what they are actually doing is what experts in a host of different domains do when trying to explain, describe, and/or justify their actions after the fact: they are appealing to the recognized rules/principles of the domain in an effort "to find words within the jargon to talk about something that [they] don't think is particularly describable."
In social psychology, a growing body of research has begun to offer an alternative picture of moral judgments, one in which moral agents engage in moral judgments from the 'bottom up' instead of the 'top down'. Such research suggests that moral judgments are primarily intuitive, "gut-level", emotionally-guided evaluations., These evaluations appear to be able to both identify morally relevant issues and provide insight into appropriate action without requiring explicit deliberation or reference to moral principles. In fact, appeal to such principles, if it occurs at all, happens (as in Kohlberg's studies) only after the fact, when moral agents are called upon to explain and/or justify judgments already made. Thus, such research suggests that moral judgments and actions are not typically guided by principles, but are in fact prior to them.
Interestingly, the independence of principled moral deliberation from the process by which moral judgments are formed appears to have some empirical support. When shown that principled deliberation leads to a different moral judgment than the one they originally gave, moral agents often fail to give up their original judgment. In fact, they report still feeling compelled by their original judgment, which they find cannot be dispelled by their recognition of the more highly deliberated alternative. Yet, if, on the other hand, they are encouraged to perceive and/or imagine the judged scenario differently, their original judgment can be over-ridden by a new one.
Similarly, research on the development of wisdom suggests that moral maturity (one component of wisdom) does not involve an increased ability to deliberate with moral principles, nor to assume some sort of stance of ideal rationality. Instead, moral maturity appears to involve highly developed moral sensibilities (which involve emotional and other forms of deeply engrained, intuitive knowlege) that lead to spontaneous judgments and actions that are appropriately responsive to the situations encountered. In other words, mature moral agents are those that are able to imaginatively draw upon a broad network of experiences and intuitive understanding, allowing them to "feel" their way through complex situations much in the same way that we unconsciously and spontaneously "feel" our way through the physical environment through which we walk. We do not deliberate about the moves we ought to make – we just make them.
The point is that we have to be careful not to confuse the product with the process. To see this, imagine the Mona Lisa. And then imagine a paint-by-numbers replication of it. Regardless of their superficial similarity as a product, the process that went into each painting was completely different: the former involved creative painting, while the latter involved following rules. Importantly, this also means that even if the creative master-level product of the Mona Lisa can be decomposed (after the fact) into a novice-level pattern of rule-following procedures, we cannot equate the process through which the Mona Lisa was originally made with the process through which its paint-by-numbers replication was made. Similarly, while it may be possible (at least in part) to explain and/or justify the product of mature moral judgments by making reference to moral principles, it would be a mistake to hold that the process through which such judgments were formed was, therefore, principled moral deliberation. This is the case even if the decomposition of mature moral agents' judgments into principles makes them potentially reproducible by another moral agent through principled moral deliberation. I say "potentially" because, considering again the Mona Lisa example, it is important to remember that the paint-by-numbers decomposition of the original painting allows only for the reproduction of that (and no other) painting. When it comes to moral judgments, however, the possibility of being faced with the exact same situation (which would call for the exact same judgment) appears to be incredibly small. And moral judgments are only mature moral judgments when they are appropriate to the situations they are judgments of. This means that the reproduction of the same judgment in different situations would not necessarily make that judgment mature. Not to mention the fact that the process by which that judgment is generated would not be the same process used by mature moral agents. In the latter case, the moral judgment "leads" the principles; in the former case, it follows them.
This is not to say that mature moral agents do not make reference to moral principles; they do. But such reference represents an effort to articulate their experience, not to govern their behavior. That is, for the mature moral agent moral principles are descriptive, not prescriptive. As Philip Kapleau writes (in the context of Zen Buddhism): "Remember, the [principles] are not moral commandments...Rather they reveal how a deeply enlightened, fully perfected person...behaves. Such an individual doesn't imitate the [principles]; they imitate him". As such, moral principles are not necessary to provide justification for the actions of – and thus, cannot impose normative authority upon – mature moral agents.
Dreyfus on Expertise ~ Dreyfus has presented an account of "ethical expertise" that largely captures the essential features of mature moral agency. Comparing the development of moral maturity to the development of other forms of expertise (e.g. chess, driving, martial arts, etc.), Dreyfus shows that while moral principles play a role in moral agents' early moral judgments, they necessarily drop out of the picture as maturity is achieved. The moral novice relies on principles to introduce her to the rough "parameters", or basic rules, of moral practice. Over time and through experience, though, such reliance is gradually replaced with the capacity for "immediate, intuitive" – that is, non-deliberative – responsiveness to the details of particular situations. This responsiveness springs from a cultivated body of embodied knowledge (i.e. know-how) that affords a heightened awareness and affective attunement to one's environments (both internal and external). Once an agent has advanced past the early stages of development, principled moral deliberation is no longer necessary. The conception of moral maturity as consisting in principled moral deliberation is thus turned on its head.
Dreyfus writes, "...an expert [in this case, a mature moral agent] does not solve problems. He does not reason. He does not even act deliberately." Rather, he acts spontaneously, intuitively, and decisively: that is, immediately, without deliberative effort, yet with the appropriate judgment or action. Just as the professional ski racer is able, without deliberation, to apply the right amount of pressure to her skis in order to bring herself rapidly around a steep turn; just as the master chess player is able to respond immediately and decisively to the moves of her opponent; just as the Native American Indian immediately perceives what manner of man and/or animal has passed through his territory; the mature moral agent is able to intuitively perceive and understand the morally relevant features of complex situations and to act spontaneously in a morally mature manner in response to such situations. As one Japanese martial artist writes:
There is no choosing. It happens unconsciously, automatically, naturally. There can be no thought, because if there is thought, there is a time of thought and that means a flaw...If you take the time to think 'I must use this or that technique' you will be struck while you are thinking.
On this alternative conception of mature moral agency, the development of moral maturity comes not through principled moral deliberation, but rather through the conscientious cultivation of the kinds of experiences that, through the attentive development of habitual practice, become ingrained in moral agents' intuitive response repertoires.
Dewey on Moral Character ~ For Dewey, all human action (or, as he calls it, 'conduct') is morally significant. He writes,
Where there is conduct there is not simply a succession of disconnected acts but each thing done carries forward an underlying tendency and intent, conducting, leading up to further acts and to a final fulfillment or consummation...Every act has potential moral significance, because it is, through its consequences, part of a larger whole of behavior.
In other words, all human action is significant because of its bi-directional link to the character of the moral agent whose action it is: the action of a moral agent both "...reveals the existing self and it forms the future self". In this way, the self and its acts – i.e., a moral agent's character and her actions – are unified and cannot practically be separated. This isn't to say that moral agents cannot act, as it were, 'out of character' – rather, it is to say that such acts function to change the structure of the very character that they were originally 'out of character' with.
Now, Dewey defines character as "...the interpenetration of habits". He writes:
...we are given to thinking of a habit as simply a recurrent external mode of action...but habit reaches even more significantly down into the very structure of the self; it signifies...an increased sensitiveness and responsiveness to certain stimuli, a confirmed or an impaired capacity to attend to and think about certain things.
Given, for Dewey, the intimate link between character and action, the cultivation of habits plays a vital role in the cultivation of moral maturity, for it is the process by which moral agents' dispositions to respond to and act in the world (i.e. their character) are formed. Through habit formation, patterns of action and perception are developed that not only structure moral agents' tendencies to respond and act in certain ways – they also structure the way moral agents' perceive and evaluate the world by influencing what kinds of stimuli they attend to and what kinds of features they recognize as morally relevant. The cultivation of moral maturity thus requires not only that moral agents actively attune themselves to the morally relevant details of their environments (a process accomplished by modeling respected others and learning through practice, instruction, and experience how to attend to certain features of their environments), but also that they develop a stable disposition to respond and to act appropriately (through a sustained, mindful, and directed cultivation of engrained habitual responsiveness). As moral agents' actions begin to reflect the (initially untrained) insights of their moral sensibilities, those sensibilities are strengthened and stabilized, thereby expanding their capacities to transform their insights into a stable foundation of what Francisco Varela has called embodied "ethical know-how".  Through repeated activity such know-how is employed and, by way of direct feedback, fine-tuned and increased. Through this process, moral maturity is achieved.
So, for Dewey it is only through the active, conscientious cultivation of moral character that moral agents develop the sensitivity, depth, and experience necessary for mature moral agency. And moral character can only be developed through sustained moral engagement, as the formation of character necessarily involves reliable and stable forms of acting in the world (i.e. engrained, habitual responsiveness). Here we see the similarity to Dreyfus' view: through sustained practice, instruction, and engagement moral agents cultivate a habitual, engrained responsiveness that allows for decisive, immediate, and appropriate evaluation and action. For Dreyfus, this is the achievement of ethical expertise; for Dewey, it is the cultivation of moral character.
Dewey wrote (echoing Aristotle), "...only the good man is a good judge of what is truly good." And this, I think, brings us back to the issue of moral principles, for it seems to me that the mature moral agency is the development of a way of perceiving, evaluating, and acting in the world that just is itself the kind of general orientation of conduct that our moral principles are attempting to capture. The mature moral agent does not need to be reminded that she should "never lie", "uphold justice", or "maximize utility" – and this is not because she already has these principles in mind (i.e. because she has them memorized or has otherwise internalized them), but rather because she simply has become the kind of person (i.e. she has cultivated the kind of character) that interacts with the world in such a way as to reliably bring such states of affairs into being. Thus, she engages the world honestly, she acts justly, and she brings about the best consequences – not because she had it in her mind to do so, but because she has developed to the point that she both recognizes and acts upon the morally relevant features of the situations that she is faced with. To do otherwise would be to go against her character, to act in contradiction to herself. That is, if she were to do otherwise, she would no longer be who she is.
 I use "appropriate" in reference to moral judgments and actions instead of something like "accurate", "correct", or "right" because at least within the confines of this paper I want to remain neutral on the realism/anti-realism debate.
 Hubert Dreyfus and Stuart Dreyfus (1986). Mind Over Machine: The Power of Human Intuition and Expertise in the Era of the Computer. New York: The Free Press, p. 22.
 Of course, most philosophers recognize something like Mill's "out of time" problem, which led him to suggest that people would not really be required to be perfect utility maximizers, for they need only worry about their utility within a circumscribed sphere. However, there are important problems with this qualification. See, for example, Susan Wolf (1982, 1992) and Peter Singer (1972). I would argue that the response that such ideals are only heuristics – i.e., "ideals that we cannot meet but should strive for" – is problematic for several reasons, not the least of which is that it appears that we cannot, in fact, know from a stance of imperfect rationality what the ideal stance of perfect rationality is. For instance, Antonio Damasio's (1994) research on the role of emotions in reasoning appears to show that the view that there could be an ideal stance of perfect rationality is misguided.
 Lawrence Kohlberg (1976). "Moral stages and moralization: The cognitive-developmental approach." In T. Lickona (Ed.), Moral development and "
 For similar criticisms of Kohlberg's methodology, see Carol Gilligan (1982). In a different voice. Cambridge, MA: Harvard University Press and Jonathan Haidt (2001). "The emotional dog and its rational tail: A social intuitionist approach to moral judgment." Psychological Review. 108, 814-834, among others.
 There is a large body of research in social and cognitive psychology indicating that post hoc reasoning and motivated reasoning are two of the most frequently used forms of reasoning: Lerner & Miller, 1978; Kunda, 1990; Kunst-Wilson & Zajonc, 1980; Nisbett & Wilson, 1977; Perkins, Allen, & Hafner, 1983. There is also research showing that most of the evaluations made in social situations are largely automatic and do not involve explicit reasoning: Bargh, 1994; Bargh & Chartrand, 1999; Bargh, Chaiken, Raymond, & Hymes, 1996; Zajonc, 1980.
 Quoted in Dreyfus and Dreyfus (1986), p. 35. This quote came from an interview conducted with an expert-level nurse who was discussing what it was like to try to explain to other less skilled nurses why (and how) she took the actions or made the medical judgments in particular situations the way she did.
 Jonathan Haidt (2001). "The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment", Psychological Review, 108, 814-834; Green & Haidt (2002), "How (and Where) Does Moral Judgment Work?", Trends in Cognitive Science, 6(12), 517-523.
 By "intuitive" I am referring to the form of embodied, non-inferential knowledge often referred to as "know-how" (see Varela, Dewey, Heidegger, and others) that is developed in such a way as to be immediately and non-deliberatively accessed. As John Dewey wrote: "We may...be said to know how by means of our habits...We walk and read aloud, we get off and on street cars, we dress and undress, and do a thousand useful acts without thinking of them. We know something, namely, how to do them." Human Nature and Conduct (1922), as published in John Dewey, The Middle Works, 1899-1924, Vol. 14. Carbondale: Southern Illinois University Press, p. 177-178 (emphases added). See also Varela's Ethical Know-how: Action, Wisdom, and Cognition (1999): "We always operate in some kind of immediacy of a given situation. Our lived world is so ready-at-hand that we have no deliberateness about what it is and how we inhabit it." p. 9.
 Importantly, in my reference to intuitions I am not advocating an intellectualist/rationalist (or "Platonistic") account of intuition. If anything, I lean towards the perceptual accounts, though I think it is clearly false that this would require any sort of special or unique perceptual faculty. The best account, I'd argue, (a sort of "embodied" account) lies somewhere in the insight revealed by the following quote: "...he might have said, if any man could have got him to talk about it, that like the morning dove, the bittern, the Indian, he had a sixth sense. What he thought of as his sixth sense was in fact only what his five senses agreed on and communicated to his mind, acting together, like an intelligence agency, to sort out, accept or reject, and evaluate the impressions that came to them." Mountain Man, Vardis Fisher (1965:2000), University of Idaho Press, Moscow, Idaho, p. 6-7.
 When I say "without deliberation or reference to moral principles" I mean not only that there is no explicit reasoning according to moral principles, but also that there is also probably no explicit reference to (or inference from) propositional states such as beliefs and desires.
 See, again, Haidt, (2001) among others. Clearly, strong theoretical conclusions cannot be drawn from empirical findings such as those discussed here. However, I think such findings are instructive insofar as they point towards theoretical framework(s) that appear to be the most explanatorily adequate.
 See, for example, Baltes, P. Gluck, J. & Kunzmann, U. (2002). "Wisdom: Its structure and function in regulating successful life span development." In C. R. Snyder and S. J. Lopez (Eds.) The Handbook of Positive Psychology, Oxford: Oxford University Press.
 I am indebted to Trevor Curnow (Wisdom, Intuition, and Ethics, 1999) for this example.
 (Roshi) Philip Kapleau (1980). The Three Pillars of Zen. New York: Doubleday, p. 231-232. Emphasis added.
 Hubert Dreyfus and Stuart Dreyfus (1991), "Towards a Phenomenology of Ethical Expertise", Human Studies, 14, 229-250.
 Ibid., p. 235.
 In fact, it is a liability, for it leads to inferior judgments. This is not to say that situations don't arise in which such deliberation is nonetheless necessary – even for mature moral agents (clearly everyone is faced, at one time or another, with a situation that falls outside the range of their experience). The important thing to recognize is that the moral judgments that result from such deliberation will inevitably be inferior to the spontaneous judgments that normally accompany mature moral agency.
 Dreyfus and Dreyfus (1991), p. 235.
 Quoted in Dreyfus and Dreyfus (1986), p. 34.
 Dewey (1932/1960), p. 11.
 Ibid, p. 149 (emphasis mine).
 Human Nature and Conduct (1922), as published in John Dewey, The Middle Works, 1899-1924, Vol. 14. Carbondale: Southern Illinois University Press, p. 29.
 Dewey (1932/1960), p. 13.
 Francisco Varela (1999). Ethical Know-How: Action, Wisdom, and Cognition. Stanford, CA: Stanford University Press.
 Of course, Dewey differed with Dreyfus on a number of points, not the least of which was the role deliberation vs. intuition played in the formation of appropriate moral judgments. As Dreyfus points out, however, (see Dreyfus and Dreyfus ), while Dewey was rightfully suspicious of the "Platonistic" account of timeless, universally true and self-evident intuitions, this resulted in him failing to fully appreciate the deeply intuitive account of moral judgment and character that he was himself developing.
 And by this I do not have in mind a Ghandi or Martin Luther King, but rather just all those every-day people who have, through experience, developed moral character (often referred to in the literature as 'wisdom').