Moving beyond rationalistic responses to the concern about indoctrination in moral education

Abstract

Indoctrination is an ongoing concern in education, especially in debates about moral education. One approach to this issue is to come up with a rational procedure that can robustly justify potential items of moral education content. I call this the ‘rationalistic justification project’. Michael Hand’s recent book, A Theory of Moral Education, is representative of this approach. My essay has three parts. First, I show that Hand’s justificatory procedure – the problem-of-sociality justification – cannot serve the purposes he has in mind; it fails on its own terms and may even cause the teacher to inadvertently slide into indoctrination. Second, I argue that the causes of this failure lie deeper than Hand’s particular approach to the rationalistic justification project; rather, it is the broader project itself that is misguided, largely due to its narrow conceptions of morality and rationality. Third, I offer an alternative way of framing the issue of indoctrination, by drawing on Aristotle’s philosophy of rhetoric. My suggested approach recontextualizes the issue of indoctrination and brings into focus a broader set of relevant features of the teaching–learning situation.

Introduction

Advancing any proposal regarding moral education in the context of a morally pluralistic society such as our own often raises the concern of indoctrination. Are we indoctrinating students, we may worry, if we involve them in a programme of moral education about which there may be reasonable disagreement in society at large? Is there any part of morality that enjoys consensus or can be soundly defended by reason? If not, are moral educators simply persuading students to adopt their own opinions – that is, indoctrinating them? In light of this concern, some may suggest that we should adopt a neutral approach to moral education or drop the project of public moral education altogether. In a recent book, Michael Hand (2018a) critiques both of these options. He worries that only teaching ‘about’ morality – simply displaying a range of options for students to choose from – may nudge young people towards moral agnosticism. In other words, students could decide to remain ‘amoral’, in the sense that they could refuse to commit to any moral standards, given the epistemic uncertainty that appears to surround them (Hand, 2018a: 12). As for dropping moral education entirely, Hand argues that while children may acquire some moral beliefs from their environment and may have a natural tendency towards sympathy, they also absorb various kinds of prejudices, and natural sympathy is, after all, limited. If schools abandon the project of moral education, they are in effect abandoning children to be indoctrinated by other agencies (Hand, 2018a: 7–9). We need, he concludes, a more deliberate form of moral education.

So far, I am in agreement with Hand. However, we quickly part ways. To put the concern about indoctrination to rest, Hand develops a rational procedure that can test whether a given moral standard is robustly justified or not. If a standard passes the test, then it can be taught ‘directively’ – that is, we can go beyond teaching ‘about’ it and can persuade students that it is indeed justified – without fear of indoctrination. As Patricia White (2019) suggests, this approach places Hand in an established tradition of theorists, such as Paul Hirst (1974) and John Wilson (1977), who have made similar attempts to rationally justify the enterprise of moral education in pluralistic societies. In fact, Emile Durkheim formulated an early version of this project (Crittenden, 1993). I will argue below that Hand’s recent attempt does not succeed, even according to its own criteria. It is instructive, however, to examine his approach carefully, as a case study, to see what is misconceived about what I will call ‘the rationalistic justification project’. I have chosen his account in part because it is very clearly argued and recent, but also because it has generated a rather significant volume of debate among scholars in education (e.g. Aldridge, 2019; Clayton and Stevens, 2019; Copson, 2020; De Ruyter, 2019; D’Olimpio, 2019; Drerup, 2018; Ferkany, 2018; Hambrick, 2020; Hobbs, 2020; Kotzee, 2020; Maxwell, 2019; Peterson, 2019; Tillson, 2017, 2019; White, 2016, 2017 – Hand has also written several responses, for example, 2018b, 2019a, 2019b, 2020). Hand’s concern about indoctrination is, of course, genuine and warranted. And we should certainly try to articulate good reasons why certain approaches to moral education are better or worse than others – in avoiding indoctrination and/or for other purposes. However, the rationalistic justification project is misguided, in part because of its narrow conceptions of morality and rationality.

First, I will carry out an immanent critique of Hand’s proposed justification procedure: the problem-of-sociality justification. Briefly, the problem of sociality is generated by the conjunction of three facts of human existence: resource scarcity, rough equality and limited sympathy. Moral standards that ameliorate this problem – that help human social groups sustain cooperation and avoid conflict – pass the test and are said to be robustly justified. Hand (2018b) claims that this justificatory strategy is ‘non-moral’ (p. 371). Armed with this rational test, we can determine which moral standards can be taught directively, that is, those which are not subject to reasonable disagreement. Teachers are encouraged to use the same justificatory strategy to persuade students in class that conflict-averting and cooperation-sustaining moral standards are rationally justified – this, without the slightest concern about indoctrination. This is because of Hand’s peculiar definition of indoctrination, which consists of persuading a student, however indirectly, of coming to a conclusion that is not robustly justified, that is, about which there is reasonable disagreement. I will argue that the problem-of-sociality justification fails on three accounts: (1) it is not, in fact, ‘non-moral’, in that it presupposes certain moral commitments and therefore cannot serve Hand’s purposes; (2) it is radically indeterminate in that it admits of solutions so extremely disparate that it is not a reliable test in the way Hand requires it to be, helping us decide how a given moral standard should be taught; and (3) largely because of (1) and (2), it will be unlikely to lead, in the context of moral inquiry, to the kind of justified belief Hand feels is an important ingredient for full moral commitment.

Second, I will argue that the failure of Hand’s problem-of-sociality justification is in part a symptom of deeper issues – some of which are shared by other efforts that follow what I am calling the rationalistic justification project. In other words, it is not a matter of fine-tuning his argument to resolve the abovementioned problems; the broader trajectory of the rationalistic project itself is misconceived. There are three points I will make here. (1) There is no such thing as a non-moral argument that can, as it were, serve as a gateway into morality. There is no non-moral point outside of morality because, following Charles Taylor (1989), as human beings, we are always and already oriented in moral space. (2) To think of morality as a device with one specific social purpose is to drastically oversimplify it. Adapting a phrase from David Wiggins (2006), I will argue that to say that moral education is a device of some kind is either false or nearly vacuous. If we reconceive morality as broader in nature, we are better able to see the true method of the rationalistic justification project. (3) Points (1) and (2) have a number of implications for moral inquiry, namely, that ethical convictions play an important role in it and that we should pay close attention to the language in which it is conducted; both of these realities are largely invisible from the perspective of the rationalistic justification project.

Finally, if a strictly non-moral justification of moral standards is impossible and morality is understood to be a much vaster and inescapable framework for human beings, is indoctrination then inevitable? I will argue that this broader conception of morality encourages us to recontextualize the issue of indoctrination. If we are to avoid it, we should pay attention not only to the availability of sound reasoning in moral inquiry, but also to the intention and methods of the teacher, as well as to the existing intuitions of students and their broader socio-cultural context. In this connection, I draw on Paul Ricoeur’s (1977) study of Aristotle’s philosophy of rhetoric to offer some insights into how moral educators can steer clear of indoctrination.

An immanent critique of the problem-of-sociality justification

The problem-of-sociality justification occupies a central position in Hand’s theory of moral education. The problem of sociality refers to the propensity for cooperation to break down and for conflict to erupt within human social groups. It arises because resources are scarce, we are roughly equal to one another (in Hobbes’ sense that the weakest could kill the strongest under certain conditions), and our natural sympathies are limited. If conflict does break out, we all lose. Therefore, we all have an interest in sustaining cooperation and averting conflict. Any moral standards that unambiguously ameliorate the problem of sociality, then, are said to be robustly rationally justified – lifting us neatly out of the realm of reasonable disagreement. In Hand’s (2020) words, the problem-of-sociality justification ‘gives all human beings, whatever their religious views and whatever other justificatory theories they may find plausible, a decisively good reason to subscribe to basic moral standards’ (p. 238). The ‘decisively good’ reasons we have for subscribing to the basic moral standards Hand (2019a) has in mind are, in an important sense, ‘non-moral’ (p. 663). A moral reason would not serve Hand’s purposes: students need to be offered non-moral reasons so that they can form a rational belief that the moral standards in question are justified.

Armed with the problem-of-sociality justification, the moral educator can ‘test’ a given moral standard to see whether it is robustly justified, unjustified, or neither. This categorization guides her as to how she is to teach the standard in question to avoid indoctrination. If the moral standard passes the test unambiguously – that is, if it certainly would ameliorate the problem of sociality – it is robustly justified and can be taught directively, meaning that the teacher can lead students to the conclusion, whether directly or indirectly, that it is indeed justified. If the moral standard is clearly not justified, it can be taught directively against, meaning that students can be encouraged to believe that it is not justified. Finally, in some cases, it might be unclear whether or not the standard ameliorates the problem of sociality; should this be the case, it should be taught non-directively, meaning that it can be taught about, but not as justified or unjustified – that is, it is still subject to reasonable disagreement. The problem-of-sociality justification’s capacity to clearly distinguish between these three categories is thus vital.

The work of a moral educator is twofold according to Hand: on the one hand, to develop in students the behavioural, affective and conative dispositions associated with subscription to justified moral standards, and, on the other hand, to rationally persuade them that these moral standards are indeed justified. The former, Hand calls moral formation, and the latter, moral inquiry. Full moral commitment, which is what Hand is seeking, requires both of these. The problem-of-sociality justification comes into play in the context of moral inquiry, which is an essentially cognitive activity. Having presumably figured out ahead of time the category to which a given moral standard belongs (justified, unjustified, or reasonable disagreement), the moral educator can then proceed with moral inquiry in the appropriate manner – whether directively (for or against) or non-directively. In the case of justified moral standards, the teacher is meant to rationally persuade students, using the problem-of-sociality justification, that these standards are indeed justified, therefore resulting in the formation of a justified belief in the student’s mind in a non-indoctrinatory way. Crucially, she is able to do so because (1) the problem-of-sociality justification is non-moral and (2) it is able to correctly identify justified moral standards. In this section, I will take issue with both of these claims, which in turn calls into question the problem-of-sociality justification’s serviceability in moral inquiry.

The problem-of-sociality justification is not non-moral

John Tillson (2019) has argued that, by ‘enlisting non-moral reasons for endorsing moral reasons’, Hand ‘misses the point of moral reasons’ (p. 653). Tillson feels that Hand’s problem-of-sociality justification speaks to the desires of students (e.g. their desire not to live in a state of conflict), trying to persuade them in this way to endorse certain moral standards, but that this misses the point of morality, which often demands that we act against our own desires. In his response to Tillson, Hand (2019a) readily admits that the problem-of-sociality justification is indeed a non-moral (or desire-based) argument, and that this is in fact crucial for his purposes. He writes,

It is generally the case that, in advance of committing oneself to something – a person, project, cause or standard – it is pertinent to ask what reasons there are to make the commitment, and whether they are good ones. The reasons will obviously not be good ones if they presuppose the commitment in question. (Hand, 2019a: 663)

In other words, the argument we use to persuade students that a given moral standard is robustly justified cannot presuppose commitment to the moral standard in question. The problem-of-sociality justification is able to play this role because it (allegedly) presupposes no moral standards. As for Tillson’s accusation that he is missing the point of morality, Hand assures him that he is not trying to reduce moral reasons to non-moral ones. Hand (2019a) explains that, once students commit to a justified moral standard, they may develop additional moral reasons grounded in that commitment – additional in the sense that they are added on top of the non-moral reasons that underpin the original subscription (pp. 662–663).

However, if we examine carefully how Hand wields it, we quickly see that the problem-of-sociality argument is not non-moral in the sense that Hand needs it to be. The first moral standard Hand discusses in his book – ‘do not give offence’ – is instructive in this regard. Does this moral standard – he even considers ‘do not give profound offence’ – ameliorate the problem of sociality or not? After admitting that banning profound offence-giving might indeed ameliorate the problem of sociality, he ultimately argues that ‘while the problem of sociality must be solved, we should favour the solution that places the fewest constraints on liberty’ (Hand, 2018a: 98). He then evokes J. S. Mill:

Mill famously contends that people must be at liberty to break with tradition and custom if they are ‘to commence new practices, and set the example of more enlightened conduct, and better taste and sense in human life’ (Mill, 1989 [1859], p.64). On this view, for as long as questions about the right, best or happiest way to live remain unsettled, we should endorse social arrangements that allow people to experiment with a wide range of lifestyles and practices, including those by which others are profoundly offended . . . (Hand, 2018a: 98)

His conclusion, then, is that the moral standard ‘do not give profound offence’ does not unambiguously ameliorate the problem of sociality and is therefore subject to reasonable disagreement and should be taught non-directively. Hand’s appeal to liberty, however, should give us pause. Is this still a non-moral argument? Liberty to pursue a variety of lifestyles is certainly not directly related to resource scarcity, rough equality and limited sympathy – the ostensibly non-moral ground upon which the problem-of-sociality justification rests. If we are convinced at all by Hand here (and if the student is persuaded by her moral educator), is it not because of our pre-existing commitment to liberty? Or, relatedly, our distaste for tradition and custom? This makes it a moral argument, therefore unfit for Hand’s own purposes. In fact, it might even be indoctrinatory according to Hand – in the sense that, without a robust, non-moral justification for liberty itself, are we not persuading students non-rationally of its justifiability?

The problem-of-sociality justification is radically indeterminate

A number of Hand’s critics have argued that the problem-of-sociality justification is inadequate because it leans towards moral minimalism. Johannes Drerup (2018), for example, suggests that Hand’s argument is ‘prone to status quo bias’ (p. 581). He worries that Hand’s approach, focused as it is on individual standards of conduct and social stability, would have difficulty detecting more complex instances of social or structural injustice. Others (Aldridge, 2019; D’Olimpio, 2019; Hambrick, 2020; Tillson, 2019) have expressed similar worries, suggesting, for instance, that the problem of sociality could be solved in one society at the expense of the oppression of another, or even the oppression of a certain group or class of people within the same society. In response to some of these concerns, Hand (2019a) has simply re-stated the fundamentals of the problem of sociality: first, that we are roughly equal and therefore one group of people could not in fact oppress another group indefinitely because the latter would agitate, threatening social stability; and second, that we have limited sympathy, meaning that we have at least some sympathy and are not unaffected by the suffering of others (p. 661). He emphasizes that the problem-of-sociality argument has real human beings and social conditions in mind.

Hand’s response is inadequate of course. It ignores the fact that contemporary extremes of wealth and poverty have more or less negated rough equality in the Hobbesian sense. Those with inordinate wealth are often immune from harm in the sense that they can, for example, distance themselves from suffering that might visit those who are less economically advantaged. They are thus able to ‘free-ride’ (to benefit from living in a social group without subscribing to the same moral standards as the rest of us) quite comfortably. Moreover, as Aldridge (2019) suspects, perhaps it is the case that ‘the free-riders actually run the show, and that their unsympathetic qualities are even being held up as models for successful living’ (p. 641). While I resonate with these critiques, they are not devastating to Hand’s argument. He could simply bite the ‘status quo’ bullet, as it were, and accept that moral education is by definition minimal. Hand (2018a) says something along these lines in his book, where he claims that ameliorating the problem of sociality requires that people are willing ‘to extend each other a helping hand when one is needed’ but that ‘it is not a requirement of mutual confidence and security that people commit themselves to the global project of alleviating all human need’ (p. 117). The latter may be praiseworthy, but it is not something that falls in the realm of moral standards and moral education because we would not endorse penalties for non-compliance. What would be devastating to Hand’s approach – and this is the immanent critique I will advance here – is if the problem-of-sociality justification was radically indeterminate. If it were, it would mean that moral educators could not employ it to effectively categorize moral standards as either robustly justified or unjustified. In effect, all moral standards would be categorized as controversial – reasonable disagreement would reign, throwing us back onto the initial dilemma.

There are at least two reasons why the problem-of-sociality justification is radically indeterminate. First, while Hand’s critics have pointed out that minimal interpretations of what ameliorates the problem of sociality are possible (i.e. oppression of a group or the status quo), it is also the case that maximal interpretations are plausible. For example, one might conceivably and reasonably argue that the current economic system tends to exacerbate greed, conflict, and environmental degradation – and therefore threatens social stability – so a moral standard such as ‘work systematically to create a new economic system’ or ‘do your best to avoid propping up the current economic system’ might be justified. Hand may suggest that while these standards are praiseworthy, they are not moral per se because we would not endorse penalties for non-compliance. However, if we return to his own definition of ‘penalty-endorsing’, he is explicit that the kinds of penalties he has in mind include ‘the sort of mild penalties by which offenders are made to feel uncomfortable or ashamed, such as rebukes, admonitions, tuts, and frowns’ (Hand, 2018a: 21). Surely, we might frown at someone who is in a position to create some kind of positive change but makes absolutely no effort to do so or demonstrates no interest in this regard, even after having been made aware of some of the ways in which it threatens social stability. If both minimal and maximal interpretations are possible, the problem-of-sociality justification becomes radically indeterminate.

Second, I want to echo Bruce Maxwell’s (2019) concern about Hand’s approach, namely, that very few moral standards will be able to be robustly justified using the problem-of-sociality justification. It is telling that out of the three examples of moral standards that are thoroughly explored in Hand’s book, none of them ends up being robustly justified. In a sense, this should not come as a surprise. Social scientists have yet to discover – not for lack of effort – a fixed set of laws that can tell us which standards will help us avert conflict and sustain cooperation; it is probably the case that some standards may even simultaneously produce conflict and cooperation in different contexts. This might not be a problem for a transcendental argument, but Hand emphasizes time and again that his argument is in an important sense realistic. If we want to be truly realistic, however, we should not ignore the difficulties involved in actually determining, in an unambiguous manner, whether a given standard will help us avert conflict or sustain cooperation. Furthermore, the problem-of-sociality justification vastly simplifies societal dynamics. It creates a dyadic opposition between bad terms (conflict, breakdown, violence, instability) and good ones (cooperation, stability), but societies do not exist on a spectrum between two states. Even without reaching for a high degree of sociological sophistication, we can readily think of many other states in which we might find a society: in decline, developing, progress, stagnation, healthy, sick, decadent, flourishing, in transition – these are just some of the terms that come to mind. Common knowledge and general history offer us a rich, nuanced vocabulary and set of concepts. Hand’s simplification might be innocuous if the problem-of-sociality justification played a less prominent role in his theory, but it is its lynchpin. The complexities of real life render the problem-of-sociality justification unserviceable for Hand because few if any moral standards can be rescued from reasonable disagreement in the way he needs them to be.

The problem-of-sociality justification will not be effective in directive moral inquiry

Building on the points above, it appears that the problem-of-sociality justification cannot be used by educators effectively in the context of directive moral inquiry to help students believe that moral subscription to conflict-averting and cooperation-sustaining standards is justified. Hand needs a non-moral argument that does not presuppose any moral commitments and that can reliably tell us whether a given moral standard is justified, not justified, or controversial. The problem-of-sociality justification does not fit the bill. The first reason this is so is because Hand’s justificatory strategy appeals to moral reasons, presupposing certain moral commitments that have not (yet) been justified. Students may need to already be committed to a kind of Millian conception of liberty, for example. Second, the radical indeterminacy of the problem-of-sociality justification – stemming from the possibility of both minimal and maximal interpretations and its disconnect from real social conditions – renders it unable to classify moral standards reliably; in general, it cannot offer the kind of robust justification of moral standards that Hand wants moral educators to have access to in the classroom.

Let us consider one example, which Hand implies is robustly justified, though he never explores it in detail: ‘do not steal’. Hand would encourage teachers to directively teach students that this is a robustly justified moral standard because it helps us avert conflict and sustain cooperation. She could do this in a didactic or non-didactic way, but she is allowed to bring them to this conclusion, provided they hold the amelioration of the problem of sociality as the reason underpinning their belief that the standard is justified. No one (reasonable) would disagree that ‘do not steal’ is a good moral standard. But that it unambiguously threatens social stability is not so clear, at least on certain interpretations of social stability. Does it really threaten social stability for one person to steal something small, a student may ask? What if I am dying of hunger and I steal food from someone who is very wealthy? Hand (2018a) may argue, as he does at the outset of his book (pp. 2–3), that this is more of an application issue: that there is no reasonable disagreement here about the content or justification of the moral standard in question, but just about how it should be applied in complex situations. But this is precisely the point: the problem-of-sociality justification is an instance of application. Carrying out the justification requires one to think through with a group of students how the moral standard would be applied. A group of students might indeed come to the conclusion that ‘do not steal’ is not robustly justified because it does not unambiguously ameliorate the problem of sociality. According to Hand, they are being unreasonable. Should the teacher then force his or her own understanding on the students? If such a thing could even be done, would it not amount to indoctrination?

Questioning the approach of the rationalistic justification project

So far, I have shown that Hand’s problem-of-sociality justification, which is the lynchpin of his theory of moral education, is not serviceable for the ends he has in mind. It is not a non-moral argument and it is too radically indeterminate – moral educators would not be able to utilize it effectively to overcome reasonable disagreement and therefore avoid indoctrination. While for some this may signal that we should try to amend Hand’s account or perhaps resume the search for a better justificatory strategy, I want to suggest that this search – what I have called the rationalistic justification project – is itself misconceived. In this section, I will keep Hand’s argument in view but take a step back from the immanent critique to describe why this broader line of thinking is misguided. The three points I will raise here build off of the immanent critique developed above: (1) there is no such thing as a non-moral argument that can bring students ‘into’ morality since human beings are already and inevitably oriented in moral space; (2) morality is not a ‘device’ for achieving a single social aim and has a much broader scope, intimately linked to the meaning of life and visions of the good for society; and (3) ethical convictions and the kind of language we employ are both vital to moral inquiry, but this is obscured by the rationalistic justification project.

There is no such thing as a non-moral argument

‘Morality is optional for human beings’, writes Hand (2018b: 370). ‘If we do go in for it’, he continues, ‘we should have good reasons for doing so’ (p. 370). As already explained, Hand wants a non-moral argument that does not presuppose certain moral commitments. And as we have seen, Hand’s argument fails to meet this condition. I want to go a step further here and claim that no argument can meet these conditions. In fact, to look for such an argument is part of a misconceived effort in the history of modern moral philosophy, famously brought into question by Bernard Williams (1985) and Charles Taylor (1989, 1995), among others. According to Taylor (1995), there is a particular form of argumentation favoured by those who are embedded in this peculiar conception of morality, one that involves ‘offering someone a reason for adopting a certain ethical position which would have to be recognized as a reason whatever the interlocutor’s ethical outlook’ or ‘finding some external consideration, not anchored in our moral intuitions, which can somehow show that certain moral practices and allegiances are correct’ (p. 138). This is the search to which the rationalistic justification project in moral education is committed.

Hand thinks he can identify a non-moral argument in part because of his extremely narrow conception of morality, combined with his procedural conception of reason. His conception of morality is in fact far narrower than the typical modern theories that Taylor (1995) criticizes for reducing morality to the realm of obligations. For Hand, morality is simply a certain kind of attitude we hold towards a standard of conduct, namely, an attitude that is universally enlisting and penalty-endorsing. While some critics have tried to peg Hand as a social contract theorist and criticize his theory on this basis (Clayton and Stevens, 2019), he reminds them that the problem-of-sociality justification, strictly speaking, has little to do with morality per se – all it does is determine whether a moral standard is justified (or not) or subject to reasonable disagreement (Hand, 2020). In other words, a standard cannot be called moral if it ameliorates the problem of sociality; this just makes it rationally justified. A standard becomes moral when we take up the moral attitude towards it. Hand is not, then, a typical contractarian, in that he is not trying to come up with a procedure that will delimit the sphere of morality. It seems to me that his attitudinal theory of morality (which he borrows from David Copp, 1995) is more akin to emotivism, in that it focuses on a subjective disposition of the individual – in this case, a particular attitude we could take up towards standards of conduct, namely, one that is universally enlisting and penalty-endorsing.

Hand (2018a) mentions that one of the advantages of this attitudinal theory is that it captures the phenomenon of moral pluralism due to its complete neutrality with regard to content (p. 20). In other words, any standard can become moral, if we take up a universally enlisting and penalty-endorsing attitude towards it – which helps explain the fact that people hold such different moral standards. This is a typical advantage claimed by emotivist theories of morality. By making morality an almost entirely subjective phenomenon, more or less an expression of personal preference, moral pluralism is easy to explain. Another motivation here is that this approach lends a veneer of tolerance to one’s outlook – either by embracing relativism or by bracketing the question of the rightness or wrongness of all moral standards. Hand appears to embrace the latter strategy. Again, the problem-of-sociality justification does not judge a moral standard as wrong per se; it can only speak to whether it is justified. This is one of the features of what I called above Hand’s procedural conception of reason, and it is key to why he thinks he can identify a non-moral argument at all. I find it surprising that Hand’s critics have not focused more attention on his attitudinal theory of morality, given its obvious ties to emotivism and the scathing critiques of the latter that are available. ¹

The existence of a non-moral argument depends, then, on our ability to separate morality from the process of justification itself or at least to minimize its role in the procedure. Hand achieves this by employing an attitudinal theory of morality, which separates morality from reason. But we need not accept Hand’s theory of morality, which seems to me to be one of the most controversial features of his approach to moral education. Are we not indoctrinating students into emotivism by promoting Hand’s approach in classrooms? In any case, I will now introduce an alternative conception of morality, which, even if some might take issue with its details, should call into question the idea that there can be such a thing as a non-moral argument. Following Taylor (1989), I want to suggest, contra Hand, that morality is not optional for human beings.

Taylor (1989) takes a phenomenological approach to understanding morality. He suggests that modern philosophical approaches to morality are unduly restrictive in their focus on action, conduct, and obligation. If we look more carefully at our lives, however, we will see that they are entirely saturated with moral considerations, that we are in fact always and inevitably moving in moral space, or against a moral background, orienting ourselves in relation to what we see as having moral significance and importance. As human beings, we naturally aspire to be rightly placed in relation to these matters of moral significance – towards the goods that we see as worthy of our attention and devotion. Taylor (1989) calls this a kind of ‘craving which is ineradicable from human life’ (p. 44). All human beings, then, have a certain perception or map of the moral landscape, which Taylor calls a moral framework. It is within this framework that we think and make decisions about what it is good for us to be and do. Moreover, having such a framework is constitutive of human agency; someone without a framework would not be recognizably human. The kind of modern moral philosophy that Taylor deplores has supposed that this moral space in which we move is an illusion that we project for ourselves and that it is therefore optional to have a framework. But it is impossible to wipe away or bracket this background, although certain theorists think it can be done. Hand himself is an example of this. He relegates morality to our attitudes, essentially dismissing the idea that there might be objective conditions to which these attitudes are a response. He then sets up the problem-of-sociality justification as an allegedly non-moral argument. And yet, as we saw above, he cannot help appealing in the course of his argument to matters that are of moral significance for him, for example, liberty. In other words, among those matters of significance that help orient Hand in moral space is liberty – in his case, a Millian understanding of liberty.

Morality is not optional for human beings. All of our arguments rely on one or another moral framework, even if we are unaware of this. There are always certain ‘goods’ that we hold to be higher and worthy of our attention – that we rely upon to see what it is good to be and do. It is the conversation about these goods that is obscured and perhaps even supressed by the fruitless search for a non-moral argument. In fact, using Hand’s allegedly non-moral argument to persuade students ends up covertly suggesting to them the importance of certain goods: the survival of human social groups, Millian liberty and so on. In this sense, precisely because it does not draw attention to these goods and pretends to be non-moral, it may be open to the accusation of indoctrination.

Morality is not (only) a device for social stability

I argued above that the problem-of-sociality justification is radically indeterminate because it admits both minimal and maximal interpretations and vastly oversimplifies societal dynamics. Stepping up from this immanent critique, we can note that Hand’s approach – similar to others that operate under the auspices of the rationalistic justification project – conceives of morality as a kind of device, a means to the end of social cohesion. As Hand (2018a) puts it, ‘Morality serves to protect human social groups by motivating us to resist the self-interested reasons, uncharitable suspicions and aggressive impulses that permanently endanger them’ (p. 114). The ‘mechanism’ (p. 114) through which morality achieves this function is by habituating us to experience negative moral emotions (shame and guilt) when we ourselves contemplate flouting justified standards of conduct and by evoking the blame and condemnation of others if we do. By virtue of the particular attitudes that distinguish it, then, morality serves as a kind of device that is useful for maintaining social stability. This is how Hand sees it, in any case. But this vastly oversimplifies the nature of morality.

In elaborating his theory, Hand draws on the tradition of moral constructivism, which includes figures such as Warnock (1971) and Mackie (1977). As I argued above, Hand is not a moral constructivist per se because of his attitudinal theory of morality, but he shares with them the idea that morality is a device – an assumption common to many who undertake the rationalistic justification project. A brief look at David Wiggins’ (2006) critique of Mackie’s approach to moral constructivism will help us better understand some of the issues with Hand’s conception of morality as a device. Mackie (1977), who is concerned about the metaphysically suspect character of such things as moral obligations, initially takes morality to be a kind of device that helps us avoid social breakdown by countering our tendency towards narrow self-interest; his aim is to elaborate the content of this device. Wiggins (2006) finds that Mackie has many insightful things to say about morality. However, he argues that Mackie’s explicit methodology of constructivism is at odds with how he actually proceeds:

Mackie’s real method is this. As he moves from topic to topic and question to question, he draws constantly upon the reservoir of implicit knowledge that we all have, but make explicit only piecemeal and in given contexts, of what matters in this or that sphere of activity . . .

Mackie might find this a disappointing description. But I should reply that he ought not to complain. For this is the right method. The fact that he needs to follow it only mirrors the oft-repeated failure of moral philosophy, well documented by Mackie himself, to settle peacefully for any utilitarian aim or deontological aim or other specific aim as ‘the (overall) aim of morality’. (Wiggins, 2006: 328–329)

I think we can say the same for Hand. As he proceeds from example to example, he draws on this ‘reservoir of implicit knowledge that we all have’ and appeals to our moral intuitions along the way, for example, our love of liberty. Hand may also find this description disappointing. But this is the right method, as there is no other way to go about saying anything meaningful regarding morality.

Furthermore, Wiggins (2006) suggests that, while Mackie’s starting point was fine, he cannot go on to specify ‘an overall end that would be required for purposes of reconceiving morality as a means to that end’ (p. 329). In fact, ‘the idea that morality is a device (or a means to an end) is either, as literally understood, false – or else, as charitably understood, uncomfortably close to vacuous’ (p. 329). Wiggins suggests instead that ‘the real aim of morality, is inseparable from the everyday meaning of everyday life and its everyday extensions and elaborations. It is something practically apparent but apparent only within the business of life itself’ (p. 329). Borrowing Wiggins’ line of thought, we can say that it is beyond question (nearly vacuous) that a good moral education should and will ameliorate the problem of sociality. But that this is the central purpose of moral education is clearly false. An adequate account of moral education, then, would be inseparable from the meaning of life itself.

Finally, it is worth noting that this understanding of morality as a device for social stability, which the Handian moral educator would need to embrace, would surely be communicated to students. In other words, the Handian moral educator is not only helping students appreciate that conflict-averting and cooperation-sustaining moral standards are robustly justified, but she is also conveying to them the idea that morality is essentially a device with this aim. Do we want students to become convinced that this is the purpose of morality? Moreover, is there not an implicit view of the good for society being conveyed here? For instance, that Hand favours solutions to the problem of sociality that impose fewer restrictions on liberty suggests at least one feature of Hand’s normative vision of society. He would also favour, we may suppose, the multiplication of experiments in living (à la Mill), all jostling with one another within the boundaries of a permanent framework of basic moral standards that manages conflicts. What I am getting at is that every theory of moral education implies a vision of the good for society, whether it is implicit or explicit. However, the project of rationalistic justification obscures this link between moral education and a vision of the good by reducing morality to a device. If we want to avoid indoctrination, we should be explicit about the vision of society that accompanies our theory of moral education.

The nature of moral inquiry

Building on what I have drawn out from Taylor and Wiggins regarding the nature of morality, I want to point out two important implications for moral inquiry with students. The first has to do with the role of ethical convictions in moral inquiry, while the second pertains to the role of language. Both of these considerations are practically invisible from the perspective of the rationalistic justification project, concerned as it often is with standards of conduct above all.

Moral inquiry inevitably involves conversations about our ethical convictions – about what is morally significant to us. Moral standards cannot be assessed in a non-moral way; the process of rationalistic justification invariably brings into view one or more ‘goods’ at play. If Hand’s problem-of-sociality justification moves us at all, it is because it speaks to our moral intuitions about what we hold to be good – for example, our sense that conflict should be avoided and cooperation sustained. Even within the limited scope that Hand gives to moral inquiry, then, it is necessary for students to acquire a number of ethical convictions; these would include, at the very least, a conviction that social stability is a particularly significant good that outweighs self-interest (if it comes into conflict with it) and a set of convictions regarding the value of rationality. Hand invokes others of course, such as the conviction that liberty matters a great deal. If one continued to explore various examples of the justification of certain moral standards, many other ethical convictions would no doubt show up. Again, if we follow Taylor (1995) and Wiggins (2006), this is the only way to go about moral inquiry. And if this is the case, we should opt, I would argue, for a more transparent and straightforward discussion of ethical convictions with students in the context of moral inquiry. This, rather than smuggling them in under the cover of an allegedly non-moral form of argumentation, which is an approach that is arguably more open to the charge of indoctrination.

Language also plays an important role in moral inquiry – both the language employed by the moral educator and the language students acquire in this context. The language employed by Hand, while logically rigorous, is not particularly morally inspiring, to say the least. This is not a concern from Hand’s perspective: the point is simply to produce a knock-down argument that is rationally persuasive; in fact, he needs it to be stripped from any ‘non-rational’ forms of influence if we are to avoid (his conception of) indoctrination. But this stripped-down conception of rationality fails to grasp the crucial role of language in moral inquiry. In this connection, Taylor (1995) says that ‘I can only convince you by my description of the good if I speak for you; either by articulating what underlies your existing moral intuitions; or perhaps by moving you to the point of making my description your own’ (p. 140). The way we articulate the goods that are of significance to us can move others and even motivate them to action. There is even moral power in a particularly apt articulation. An essential part of moral inquiry, then, is articulation: finding the right language to articulate what underlies students’ existing moral intuitions and to describe the goods that animate our moral landscape.

Much more could be said about moral inquiry, but I will restrict myself here to these two points, both of which are particularly relevant to the critique of the rationalistic justification project. Briefly, then, moral educators need to be particularly sensitive to the existing ethical convictions and moral intuitions of students, help them articulate what underlies their intuitions and make them aware of the diversity of ethical convictions that should orient our conduct – this in a language that empowers students to approach the good.

Alleviating the concern of indoctrination: The rhetoric of moral inquiry

Having examined a few of the reasons why the rationalistic justification project – and Hand’s approach as my case study – is misconceived and ends up unnecessarily narrowing the scope of morality and rationality, we are thrown back on our original concern about indoctrination. If moral inquiry inevitably brings ethical convictions into play, which are subject to reasonable disagreement, is indoctrination impossible to avoid in moral education? In this final section, I will briefly address this concern directly. While I will not elaborate a full-fledged theory of indoctrination, I will suggest a few ways in which the concern can be recontextualized. To assist me in this regard, I will draw on Paul Ricoeur’s (1977) study of Aristotle’s philosophy of rhetoric.

Before proceeding, we should refer briefly back to what Hand (2018a) says about indoctrination: ‘To indoctrinate someone is to impart beliefs to her in such a way that she comes to hold them non-rationally, on some other basis than the force of relevant evidence and argument’ (p. 6). Hand maintains that if the relevant evidence does not justify a given belief (i.e. if it is subject to reasonable disagreement), the educator who is trying to impart this same belief to students is by definition employing non-rational means of persuasion, such as manipulation, psychological pressure or bullying. There is more to Hand’s particular account of indoctrination, which has also been contested (e.g. Ferkany, 2018), but these initial comments, most of which would be echoed by other proponents of the rationalistic justification project, suffice to set up a backdrop for my explorations below.

Talk of ‘non-rational means of persuasion’ may bring to mind some of the philosophical debates about rhetoric, the classical art of persuasion. Ricoeur (1977) explains that ‘rhetoric is philosophy’s oldest enemy’ – this ‘because it is always possible for the art of “saying it well” to lay aside all concern for “speaking the truth”’ (p. 10). At the same time, philosophy needs rhetoric since it also involves persuasion, given that philosophy itself is a form of discourse. This dilemma generates a number of questions: ‘what does it mean to persuade? What distinguishes persuasion from flattery, from seduction, from threat – that is to say, from the subtlest forms of violence? What does it mean, “to influence through discourse”?’ (p. 11). Some of these potential issues with rhetoric are similar to the concerns about indoctrination. In the following paragraphs, I draw on Ricoeur’s study of Aristotle’s philosophy of rhetoric to shed light on the concern about indoctrination in moral inquiry. I am not of course arguing that teaching can be reduced to a kind of rhetoric; nevertheless, it has rhetorical elements, and this is in fact one of the major sources of concern for those who worry about indoctrination. ²

The key to Aristotle’s response to the worries about rhetoric straying from the truth is the ‘link between the rhetorical concept of persuasion and the logical concept of the probable’ (Ricoeur, 1977: 11). Aristotle (1991) announces, in the opening passage of the Rhetoric, that ‘Rhetoric is the counterpart of dialectic’ (p. 619 (1354a1)). Ricoeur (1977) explains that in this context dialectic ‘refers to the general theory of argumentation as regards that which is probable’ (p. 31). ‘The probable’ here refers to the kind of ‘proof’ that is relevant to public discourse, which Aristotle contrasted with the kind of striving for ‘necessary proofs’ that governs, say, geometry. The kind of proof that belongs to the realm of ‘that which is probable’ is a form of argument that Aristotle called the ‘enthymeme’ (the counterpart to the syllogism in logic). Rhetoric is the method or technique that deals with enthymemes – that is focused on the production of this kind of proof; again, the kind of proof that pertains to public discourse. Rhetoric is thus ‘centred on the persuasive power attached to this kind of proof’ (Ricoeur, 1977: 31); it involves ‘the persuasive techniques from which one might acquire the art of using enthymemes’ (Aristotle, 1991: 620 [1354b21–22]). Rhetoric ‘as an art is concerned with the techniques of persuasion, which are kinds of demonstrations (for we are most persuaded of a thing when we believe that a demonstration of it has been given)’ (Aristotle, 1991: 621 [1355a4–7]). This is how Aristotle chastens rhetoric, preventing it from becoming corrupted.

I want to extract two insights here that should already be of assistance to us. The first is Aristotle’s distinction between different kinds of argumentation and proof that are appropriate to different realms. In this connection, that he associates the realm of public discourse with ‘the probable’ is significant. Moral inquiry may or may not be coextensive with the realm of public discourse, but it is certainly closer to that realm than it is to the realm of, say, geometry. Returning to Hand, we will recall that he demands ‘robust justification’ for moral standards. It is unclear what kind of ‘proof’ the problem-of-sociality justification yields. In any case, I think Aristotle’s insight here – that different areas of inquiry demand different methods of proof and therefore of persuasion – is worth bringing to the table when we are considering the problem of indoctrination in a given area. It may be, for instance, that Hand’s justificatory requirements are inappropriate for moral inquiry. Not because moral inquiry is irrational and necessarily indoctrinatory, but because the kind of proof (and therefore persuasion) that is appropriate to moral inquiry is simply different in nature. That we cannot achieve, say, a ‘mathematical’ kind of justification in moral inquiry should not put into question the overall feasibility of (a broader conception of) rational moral inquiry. It may even be the case that trying to achieve robust justification in Hand’s sense in the context of moral inquiry is a kind of category mistake that may lead moral inquiry astray. I want to venture the suggestion that the kind of ‘proof’ or justification that is proper to moral inquiry is more akin to the logic of ‘insight’: that moral inquiry should help us obtain more insights and deeper insight into the area of morality and ethics. While this suggestion does not eliminate the concern about indoctrination (is it even possible to completely eliminate it?), it helps reconceive the terms in which we approach it.

The second insight that we can retain from our initial exploration of Aristotle’s philosophy of rhetoric is the linkage between persuasion and proof. Again, one of the major concerns about rhetoric is that it might part ways with the truth in its attachment to ‘saying it well’. This is similar in at least one important way to the concern about indoctrination: that in instances of indoctrination, truth and evidence have been set aside, and we are trying to persuade students via some other means. Aristotle rescues rhetoric by tying firmly together the concepts of proof and persuasion: rhetoric is the study of the techniques of persuasion, which involve persuasive and appropriate demonstrations. Applying these ideas to the realm of moral inquiry, then, we can say that moral educators need to keep in view the twin imperatives of persuasion and proof. Hand is, in a sense, already sensitive to this notion: he wants moral educators to be equipped with the means of proof – ‘robust justification’ in his case – because without it he feels they might indoctrinate students. What he lacks is the deeper insight that persuasion and proof are internally related. In other words, we do not only need to join the two together, as if they were separate ingredients; appropriate persuasion is best achieved through appropriate proofs. If the kind of proof that pertains to moral inquiry is the logic of ‘insights’, as I suggested above, we can say that the best insights will be the most persuasive and will help moral inquiry avoid falling into indoctrination. This brings us back to Taylor, who spoke of ‘moving’ others through our descriptions of the good. The most insightful descriptions of what underlies students’ moral intuitions, then, will be the most persuasive.

There are two further aspects of Aristotle’s philosophy of rhetoric that I will bring into play here, since they are directly relevant to the concern about indoctrination. The first is that rhetoric always involves a specific speaker and audience, and must therefore take into account ‘the speaker’s character and the mood of his audience’ (Ricoeur, 1977: 32). In other words, ‘rhetoric is a phenomenon of the intersubjective and dialogical dimension of the public use of speech’ (p. 32). Thus, while rhetoric should not lose sight of ‘the probable’ and the kind of proof associated with this realm, it cannot convey these proofs in the same way to every audience. The same goes, of course, for moral inquiry. We cannot convey ‘proofs’ in moral inquiry without taking into account, for instance, the existing moral intuitions of the students involved. In fact, if we ignore this consideration entirely, we might in some cases stray into indoctrination in the sense that we might be asking students to accept ideas that they cannot take hold of. This is perennial educational wisdom of course: teachers need to keep in view their knowledge of their discipline and their knowledge of their students; or, as one educator has put it, she needs to keep her eye on the horizon and her ear to the ground (Ball, 1993) – an apt image for rhetorical effectiveness in Aristotle’s sense. Because the sort of moral inquiry I have been describing speaks to the existing intuitions of students, it may be in a better position to avoid the kind of indoctrination to which I have alluded here.

The final aspect of Aristotle’s conception of rhetoric that is relevant for our purposes is that we speak not only to a specific audience but also in a specific time and place. As Ricoeur (1977) puts it, ‘Rhetoric does not develop in some empty space of pure thought, but in the give and take of common opinion’ (p. 33). There is an essential link between rhetoric and the cultural context in which it is practised in that the speaker needs to ‘draw from the storehouse of popular wisdom’ (p. 33). Again, the connection with moral inquiry is obvious. Recall what Wiggins (2006) said about the right method in moral theorizing: that it ‘draws constantly upon the reservoir of implicit knowledge that we all have’ (p. 328). Moral inquiry, then, needs to take into account the moral horizons in which one’s students move and find their bearings. By offering insight, of course, moral inquiry refines students’ perception of their moral landscape so that they are better able to orient themselves; it does not leave their frameworks ‘as is’, but it necessarily moves within the common moral landscape. If we ignore the reservoir of implicit knowledge, we are tempted, I think, to offer students a highly artificial construction, such as the problem-of-sociality justification. What is more, while the latter may seem to equip us to cut right through controversy, by failing to consider the local context, we may run right into the arms of indoctrination. In other words, what this feature of rhetoric helps pick out is the need for moral educators to appreciate that what is morally controversial differs from community to community; it does justice to the idea of reasonable disagreement in fact. This does not necessarily mean these topics are out of bounds; they may even be essential to address. But how we address them is immensely relevant to the concern about indoctrination.

To summarize, there are four features of Aristotle’s rhetoric to which I have drawn attention: that the kind of proof should be suited to the area of inquiry; that proof and persuasion are internally related; that the specific condition of the speaker and listener are important to take into account; and that rhetoric is always embedded in a particular cultural context. Again, this does not amount to a full-fledged theory of how moral inquiry can avoid indoctrination. I was simply gathering a few insights that allow us to recontextualize the worry about indoctrination. Whereas for Hand the worry is entirely confined to the idea of persuading students to hold beliefs that are not robustly justified, the above paragraphs suggest that other features of the teaching-learning situation should be considered. To avoid indoctrination, moral educators need to persuade through the use of proofs appropriate to moral inquiry in a way that is sensitive to students’ existing moral intuitions and the community context.

Conclusion

I have argued that rationalistic responses to the problem of indoctrination, of which Hand is a representative, misconceive morality and therefore moral education, reducing the scope of both. Hand’s rationalistic justification strategy does not succeed according to its own criteria: it is not non-moral, it is radically indeterminate, and it cannot guide moral inquiry. There is in fact no such thing as a non-moral argument since, as human beings, we are always already oriented in moral space. Morality is also much more than a device for social stability. Moral inquiry inevitably brings into play our ethical convictions and the language in which it is carried out matters a great deal. Having put into question the rationalistic justification project, I returned to the threat of indoctrination and endeavoured to recontextualize it by drawing insights from Aristotle’s philosophy of rhetoric.