1. The question

When philosophers publish articles commending certain views on contemporary problems – for instance, that gene editing should be permitted, that athletes should be allowed to use performance enhancing drugs, that children aged 12 should have the vote – they thereby offer advice. Do they have responsibilities as advisers? If so, what difference should that make to what they publish? Although philosophers have on occasion addressed their responsibilities, they have not done so by attending to what exactly it means to advise.Footnote ¹ I shall argue that philosophers do have responsibilities as advisers, and that this makes a real difference to what and how they publish.

In prospective summary, I shall define advising as a communicative act that has an essential aim, that of persuading others to do what is commended, and examine the ways in which it can go wrong, distinguishing between bad advising and bad advice; and between failings on the part of the advisee and of the adviser. I shall criticise and reject three responses that might be made by the philosopher that make a disclaimer of an advising role: that advice is not meant; that the advice does not make a difference; and that the philosopher is only speaking to other philosophers. Finally, I shall consider the case of the philosopher who offers direct advice to a policy maker and argue that in this kind of case what has been called insincere advice may be justified.

In what follows I shall simplify by regarding the practical normative advice offered as being for the adoption of some policy or law. I shall ignore any distinctions between those who might implement such policy or law, that, for instance, between a regulatory body in civil society and a Government. I also allow that some practical normative philosophy may commend a general approach to a matter, rather than the implementation of a particular law or policy. Nevertheless, even such general comment has implications for what might be considered appropriate practical measures. My concern in this piece is what follows from understanding philosophical commendations as advice of such measures.

I concentrate on the case of normative practical philosophers who advise. There may well be interesting things to say about philosophers of science or epistemologists whose conclusions might have adverse consequences for how, for instance, some kinds of scientific and knowledge claims are understood. That possibility broaches issues that lie beyond the central case of moral advice.

2. Communicative acts

‘Advice’ is defined in dictionaries as guidance or recommendations with regard to future action. Sometimes this is qualified as prudential guidance, but there seems no reason so to limit the scope of advice in this manner, which may thus include moral advice. To advise is to offer guidance. The writing and publication of philosophy articles arguing for practical normative conclusions clearly then counts as advising. Some of these conclusions may be about past states of affairs (for example, whether the Holocaust was a uniquely evil historical event). Yet even such retrospective judgements have implications for what ought to be done in the future.

That practical normative philosophy is advice should not be surprising or strange. Yet the fact that this is barely mentioned in the meta-analysis of practical normative philosophy is both. What follows is an analysis of what we should take from this unsurprising and straightforward feature of some published philosophy.

In the first instance, how exactly should we think of the giving of advice? Advising is best understood as a communicative act in which an adviser addresses a view to the advisee and commends its acceptance. John Searle, in his taxonomy of speech acts, classifies advice (alongside requests and commands) as instances of ‘directives’ (Searle, Reference Searle and Gunderson1975). These are in turn one kind of a ‘performative’ speech act, that, as J L Austin defined it, ‘will often, or even normally, produce certain consequential effects upon the feelings, thoughts, or actions of the audience, or of the speaker, or of other persons: and [it] may be done with the design, intention, or purpose of producing them’ (Austin, Reference Austin, Urmson and Shisà1975, p. 101). To advise is to represent what is advised, the propositional content of the advice, as something to be adopted and where this is appropriate acted upon. Its defining and characteristic aim is to secure acceptance of what is advised.

The initial strangeness of thinking that the philosopher in publishing an article does advise may be diminished by attending to the following three clarificatory comments. First, advice can, to use the legal distinction, be in rem or in personam, that is directed to anyone, the world as a whole, or to a particular individual. ‘The’ advisee need not be ‘an’ advisee and may instead serve as a collective noun for whosoever can hear and attend to the advice. Second, advice can be solicited or unsolicited. It is not necessary that advice should be given only upon the request of an advisee. Indeed, we live and work in worlds where we are surrounded by unsolicited advice: cautionary notices to employees about safe practices, some road signage, health advisory posters in hospital, and so on.

Moreover, ‘unsolicited’ need not mean ‘unwelcome’, although persistent advice to an unreceptive advisee evidently unhappy with being pestered is rude. Indeed, we can easily imagine circumstances in which unsolicited advice ought to be offered: an experienced climber notes that someone nearby is attempting a route up the mountain that only ends in a dangerous impasse; a garage mechanic not at work sees that a car about to be driven off has a serious fault; a doctor in company with a stranger is able to diagnose a condition that needs urgent treatment.

Third, advice can fail to be taken up in a number of ways without thereby ceasing to be advice. Speech acts can misfire if there is no appropriate uptake. Thus, for example, the advice may not be heard by anyone, just as any particular speech act can fail to have an audience. It may not be construed by those to whom it is communicated as advice, just as, for instance, a command may not be heard as such by those to whom the command is issued, or an insult may not be understood by its intended target as denigrating them. Or it may be heard as advice but rejected and not acted upon, just as a command may be refused by its intended target.

In brief summary, philosophical recommendations are, normally, in rem advice that has not been solicited and which may fail in any of the three ways to be taken up as advice.

3. Advice, incitement and threats: the role of reason

Advice can be given without reasons (‘It would be good to do ϕ. Full stop’) but standardly is supported by reasons in favour of doing what is advised. Yet we should acknowledge that a trusted adviser's advice may be followed simply because she is trusted to advise well. Such trust may derive from the experience of past advice that has proved to be reliable, or on account of perceived characteristics of the adviser.

A philosopher might be such a trusted adviser. Nevertheless, we would normally expect philosophical advice to be argued for. In being supported in this manner by reasons we can distinguish advice from threats, in which the reasons for compliance are prudential ones concerning the well-being of the addressee of the threat and not reasons in favour of what the issuer of the threat wants done.

Advice is thus distinct from those forms of counsel in which appeal is made to the addressee's feelings or interests in agreeing with the counsellor. In this vein, H L A Hart and Tony Honoré in their work on causal responsibility in the law consider the case of harms done by a second party in response to an agent's words. They speak of ‘inducement’ which they distinguish from ‘mere advice’ defined as that in which the adviser acts primarily to draw the attention of the other ‘to reasons for and against doing some action’ (Hart and Honoré, Reference Hart and Honoré1985, p. 54).

Thomas Hobbes in one of the few instances of a philosophical analysis of advice speaks in his Chapter in Leviathan, ‘Of Counsel,’ about ‘exhortation’. This is ‘counsell, accompanied with signes in him that giveth it, of vehement desire to have it followed; or to say it more briefly, Counsell vehemently praised.’ He adds that he who exhorts does not ‘tye himselfe therein to the rigour of true reasoning’ (Hobbes, Reference Hobbes and Macpherson1968, p. 134). We may expect the philosopher to at least try to tie the advice offered to the rigours of good ratiocination.

Advice is also to be distinguished from ‘incitement’. John Stuart Mill famously differentiates between those circumstances in which an opinion ‘that corn-dealers are starvers of the poor’ is expressed. It should be ‘unmolested when simply circulated through the press, but may justly incur punishment when delivered orally to an excited mob assembled before the house of a corn-dealer, or when handed about among the same mob in the form of a placard’. Mill thought that ‘opinions lose their immunity, when the circumstances in which they are expressed are such as to constitute their expression a positive instigation to some mischievous act’ (Mill, Reference Mill1859, Chapter III). Note that the speaker in Mill's example does not advise the crowd to attack the corn dealer's house; nor does he even allude to such an action. He simply arouses in the ‘excited’ mob feelings of ill-will towards the house's occupants such that a ‘mischievous act’ is likely to result.

In short, the philosopher advises by giving reasons for what is advised and does not threaten or ‘induce’ or ‘exhort’ or ‘incite’.

4. Responsibility for the consequences of the advice being adopted and acted upon

Advice is a speech act that seeks to induce in the advisee a change of mind (or, we should allow, reinforcement of a view already held but perhaps held uncertainly or tentatively). Such a change of mind can have further consequential effects if the other acts on what is now taken to be the right view as to what should happen or be done. On a simple ‘but for’ causal model of responsibility an agent is to be held responsible for what would not have happened save for the advice. Thus, it would seem that if the advisee is caused to do what the adviser commends as best, then the latter is responsible for what is done.

Yet, inasmuch as advice is supported by reasons this simple model does not get it quite right about the responsibility of the adviser. Consider the ideal limit case – and it is important to stress that this is an ideal limit – in which the advisee does ϕ for those very reasons that the adviser has provided in favour of doing ϕ and no others. Although in such a case the advisee acts as she does because the adviser gave her the relevant advice the adviser is not to be held responsible in the sense of being held liable for what is done by the advisee. If acting on the advice is wrong the adviser cannot be blamed for providing the advice. It is the advisee who is blameworthy for doing what she regards herself – incorrectly – as having decisively good reasons to do. Had she considered the matter alone and unadvised, doing what she did for those reasons then she would be blameworthy. Certainly, she was provided with those reasons by an adviser. However, it was her acceptance of those reasons that made the advised action hers and not that of the adviser.

This ascription of responsibility is recognized in a classic liberal defence of liberty of speech. As a communicative act advice can be freely given and it can also be freely accepted or rejected. Thomas Scanlon defends a liberal principle of free expression that exempts from criminal sanction those instances of acts of expressions in which the connection between them and subsequent harmful acts consist merely in the fact that the former led individuals to believe these acts worth performing (Scanlon, Reference Scanlon1972). He does so because he regards the principle as protecting the autonomy of citizens, their capacity to make their own considered decisions as to what it is best for them to do. The speaker who provides citizens with reasons to do what is harmful should not be held legally liable for the harm subsequently done, if the citizens act as they do only because they believe they have good reasons to.

Like Mill, Scanlon seeks to distinguish those expressive acts that rationally persuade the other from ones that cause the auditor to act in ways that do properly expose the speaker to liability for harms done. These would include ‘incitement’ or ‘threats’.

5. Bad advice

Let us be clearer about liability for advice, not incitement or a threat, that results in bad outcomes. Responsibility can be borne by the advisee or by the advisor; and fault can be found in the advice or in the advising. To take the latter disjunction, bad advice encompasses both bad advising and advice of the bad. What is advised may be ambiguous or unclear; it may be supported by poor reasoning, by claims that are inconsistent; it may be unsupported by the evidence, and by arguments that are fallacious. Or it may straightforwardly recommend what is wrong and the consequences of whose adoption are bad.

Importantly when advice goes wrong fault can be found on the side of the advisee. Thus, there may be a failure fully and properly to understand what is in fact very good philosophical advice. It may be that the advisee embraces the conclusion, what is commended, but without appreciating why it is commended; or does in fact grasp the argument but misconstrues the conclusion. That could lead to actions that are not supported by the advice offered; or, that are not those actions that the adviser intended the advisee to take.

Dan Brock worries about precisely this kind of possibility in thinking about his work on the President's Commission for the Study of Ethical Problems in Medicine. He uses the example of offering philosophical advice that defended the withdrawal of life-sustaining treatment as permissible killing and which appealed to the moral irrelevance of the distinction between killing and letting die. His worry was that non-philosophers on the Commission might be confused about these matters, but that addressing the confusion as a philosopher would mean speaking in terms that might not produce clarity. Instead, it would risk inducing misapprehension of what was at stake with harmful consequences for any adopted policy.

This example is, he notes, only one of many, and his general concern was that philosophers seeking to influence policy might be persuaded not to ‘follow the [philosophical] argument wherever it leads’, but rather be ‘manipulative’ and play ‘a little fast and loose with the truth’ (Brock, Reference Brock1987, pp. 788–89).

Brock would presumably not have tailored his published academic work in this fashion, least of all in the article outlining the worry about doing so. The disquiet he voices there arises from a particular context, that of advising policy makers, and how best to do so as a philosopher but not necessarily in the argumentative voice of one. I will return to the issue of what and how philosophers might directly advise policy makers. In their published work philosophers have an obligation to be clear and unambiguous in what they commend. They also are obligated to offer what are believed to be the best reasons for the favoured commendation. They should be good in their advising.

Yet, turning to the second disjunct, what they commend may be wrong; their advice may be bad in the sense of advice of the bad. Here the philosopher runs two kinds of risk. The first is epistemic, that of mistaking what should be done. The advice being considered is practical, public and policy related. It is to that extent essentially political and the epistemic risks in question arise from what John Rawls calls the ‘burdens of judgments’, namely the ‘many hazards involved in the correct (and conscientious) exercise of our powers of reason and judgment in the ordinary course of political life’ (Rawls, Reference Rawls1996, pp. 55–56). Those risks are run however conscientious one is in the exercise of reasoned argument. Nevertheless, that fact does not exempt the philosophical adviser from the duty to be a conscientious reasoner.

Second, there are the consequential risks that arise from the adoption of the advice. In turn, such risks encompass two possibilities. First, that the commendation is badly executed by those who make policy even where it is fully and correctly understood; second, that circumstances are such as to make the commended action misjudged.

Any adviser must be held responsible for a failure to foresee what any reasonable person would. No adviser can be faulted for a failure to see what could not reasonably have been predicted. To the British Prime Minister, Harold Macmillan, is attributed (some claim falsely), the response to a journalist's question as to what is most likely to blow governments off course, ‘Events, my dear boy, events’ (Knowles, Reference Knowles2006, pp. vi and 33). The unexpected – the unpredictable course of events – can utterly undermine the best intentions of the policy maker.

Nevertheless, if what is advised, even if true or well justified, can be predicted with good reason to have bad effects inasmuch as the advice is misrepresented, misunderstood or misused, the philosopher should consider whether still to offer the advice. Indeed, as Jubba and Kurtlumus (Reference Jubba and Kurtlumus2012) argue the adviser ought to remain silent and not publish the advice.

In sum, the philosopher as adviser must be honest, sincere, and conscientious in the formulation of advice. They should also be diligent in seeking to avoid unclarity and ambiguity. They should also take account of what might reasonably result from the hearing of the advice and possible action by policy makers. To do this they might engage in the imaginative exercise of hypothesising that they have the ear of a philosophically literate Minister, someone who can understand what is advised and who can enact the advice thus understood. What difference would that make to what is advised and how it is advised? Such a consideration is not insignificant and it should properly weigh with the philosopher who recognises that published work is advice.

6. ‘I didn't mean to advise’

The philosopher who publishes practical normative recommendations thereby advises even if such advice is not solicited, is not directed towards any identifiable advisee, and is not heard as advice or acted upon. As such the philosopher bears the responsibility of any adviser to take due care as to what the advice is and how it is offered. Yet the philosopher might seek to abjure the role of adviser, and its attendant responsibilities, by claiming that the advice was not intended.

Inasmuch as such a claim runs counter to the essential end of advising, we can view it as an instance of a speech act that is insincerely made. Other speech acts can also be insincere in this manner and are characterised as abuses as opposed to misfires within the class of infelicitous speech acts. However, we should allow that the philosopher who makes practical normative recommendations but claims not to mean to advise is not lying with the intention of deceiving. Philosophers should of course clearly be held responsible for deliberately and knowingly misrepresenting as what is best that which they believe not to be such. Inasmuch as such mendacious advising resulted in bad consequences, their responsibility would equally clearly extend to those outcomes. The final section briefly considers whether such lying might nevertheless be justified in a policy context,

We should note that insincere directives in Searle's sense can still succeed. A command that is not meant as such can nevertheless secure compliance, and a request not intended can cause the other to provide what is asked for. Insofar as this is the case those who knowingly issue directives without intending to do so face the burden of clearly justifying what they do. What then might be the defence of the claim that advice was not meant?

In fact, there are three kinds of insincere philosophical advice. The first is an explicit disclaimer of any intention to encourage or endorse action that is consonant with the normative recommendation. Rather, it will be said, the point of the philosophical piece is to rehearse an argument without endorsing its conclusion. Here is an example of such a defence. In 2012 the Journal of Medical Ethics published online a paper by Alberto Giubilini and Francesca Minerva entitled ‘After-birth abortion: why should the baby live?’ (Giubilini and Minerva, Reference Giublini and Minerva2013). It developed and defended a position familiar to bioethicists that holds that neonates have the same moral status as foetuses and both lack the status of actual persons. Thus, abortion and ‘after-birth abortion’, namely infanticide, are equally morally permissible.

The Journal of Medical Ethics paper's publication created a storm. In response, the authors offered a defence of their actions in an ‘open letter’ on the Journal's blog (Giubilini and Minerva, Reference Giublini and Minerva2012). To selectively quote:

Is the authors’ claim that they are ‘philosophers not policy makers’ an adequate defence of their insincere advice? It is true that from a claim that ϕ is morally permissible, it does not follow that ϕ should legally be allowed. However, the claim that their work was an ‘exercise in pure logic’ suggests that the authors did not even intend to make normative recommendations. Yet it was written as a piece of practical normative philosophy, and it falls properly within the scope of any normative advice to consider whether, and if not why, what is morally permissible should be legally permitted.

It is true that philosophers are not policy makers. Philosophers should also not write or act as if they were, save as part of an imaginative exercise in evaluating what the implications are of being able to implement one's recommendations. Yet, philosophers who commend a state of affairs that might be policy can at least be asked to consider whether they would legislate such a state of affairs. Would you, we can ask, wish that what you advise to be allowable is allowed?

The only obvious remaining way to read the authors claim – ‘we are not policy makers, we are philosophers’ – is that the task of the normative philosopher is or can only be that of exploring the implications of endorsing some view. This can be done without oneself endorsing (or rejecting) that view. This is an instance of philosophical hypothesizing of the following form: ‘If you believe P then you are committed to believing Q, and that is the case whatever the truth or falsity of P, or whatever might be the warrant for P’.

Such a form of argumentation plays an especially important role in anticipatory criticism of one's opponents whereby those who subscribe to some position can be warned off such subscription by being alerted to the implications of doing so. However, it is normally made clear that this is what is intended. Moreover, there is a difference between hypothesising what follows from subscribing to a position one seeks to criticise and couching one's own view as entirely hypothetical. Engaging in that kind of conditional reasoning should be explained and justified.

The second kind of insincere philosophical advice is that which is not fully intended. Its author may offer reasons in favour of the commendation. However, in this kind of case a significant if not the only reason for making the recommendation is that publication is in the interests of the author. The most obvious instance of this, and one that regrettably seems to be of increasing contemporary influence, is a concern to secure extensive citations of one's work. This can be achieved if a conclusion is outrageous or is significantly at odds with the views of others.

A philosopher may publish work of this kind whilst still believing what is argued. Yet what is argued for does all the same give the author reasons to subscribe to the argument and its conclusion that lack epistemic integrity. Philosophers need not be at fault for working in a context that favours such publications. Yet they are at fault for succumbing to its pressures and for engaging in philosophical advising whose warrant is insufficient. Insincere and thus bad philosophizing is the result.

The third form of insincerity is that of the philosopher who is a policy maker, or who may reasonably believe that their words will make a difference to policy. As such, I shall return to it in the context of considering the philosophical advisor to the policy maker.

7. ‘It wouldn't make any difference anyway’

The philosophical adviser may minimise the responsibilities of offering advice by claiming that it is unlikely to be acted upon. This claim should be taken seriously. Philosophy is a small academic discipline, and the readership of philosophy journals – and indeed of any individual article – is tiny. There are other relevant considerations that would appear to support the disclaimer of any real influence.

In the first place, it is proper to acknowledge the difference between personal advice and practical policy advice. Practical normative recommendations in respect of law and policy are the focus of this piece. Philosophers do not normally offer advice in respect of personal matters.Footnote ² Yet they do advise in a general sense as to what it would be good for individuals to do or refrain from doing. In respect of such matters what is commended is an action that can be performed by the addressee – tell the truth, do not break the promise, make the donation to charity, and so on. By contrast, in respect of policy it is far less clear how exactly the commended state of affairs can be brought about, and by whom.

This uncertainty or lack of clarity occurs because what is commended is a state of affairs whose achievement can only be down to the efforts of many. We can know that this is the case without knowing exactly what it is that everyone must do. Not only is it not clear what exact combination of individual acts will bring about any state of affairs. It is also, and in evident consequence, not at all clear what each of us needs to do to help to bring it about. What we can term the problem of collective action is as follows. Bringing about a desirable state of affairs requires the actions of many and may be brought about in possibly many different ways. Thus, there are very real, perhaps insuperable difficulties in spelling out who should do what to the overall end of bringing about that state of affairs. To that degree it is simply not possible, or realistic, to expect any normative judgment as to the desirability of a state of affairs to be action guiding in the way that a simple judgment in respect of a personal moral matter that ‘You ought to do Φ’ is. Adam Swift and Zofia Stemplowska concisely summarise the problem of collective action in this manner: ‘We can know what we collectively should be doing without knowing what any of us individually ought to be doing’ (Stemplowska and Swift, Reference Stemplowska, Swift and Estlund2012, pp. 387–88).

In sum, publishing the normative commendation of some policy is not immediately and directly action-guiding in the way that a normative judgment, ‘You ought to do Φ’ is. Yet the former judgment does guide action in a weaker sense. It points in the direction of, and brings into the foreground of deliberation and discussion, the question of what kinds of action would be required if that policy was to be adopted. It does so without indicating to anyone who attends to it what exactly must be done here and now in the light of that belief.

Moreover, the philosopher who commends a policy or state of affairs can at least know whether these could be brought about, even if in saying this we still need to be careful to distinguish what is possible in principle from what is plausible in the real world. To the extent that what is advised might be realized the adviser must consider what role they might play in bringing it about. In the unlikely event that everyone who plays a role in bringing about the state of affairs does so for just and only those reasons provided by the philosopher, then their role as adviser exempts them from responsibility. Nevertheless, they should be sure that what they advise is supported by good reasons, and not just reasons everyone accepts.

If their advice only succeeds in persuading others to do what is not in fact advised or what in being attempted falls short in bad ways, then the adviser is responsible for what results if such outcomes could reasonably have been foreseen. What at the outset was defined as the ideal limit case – when the advisee does what is advised for just those reasons advanced by the adviser – is just that, both ideal and at the bounds of possibility.

Nevertheless, the philosophical adviser might still protest that the chances of their advice being acted on are slim to nil. Two responses are in order, from respectively the side of the advice and the side of the adviser. First, the adviser is surely responsible for what might be the case if the advice – whose characteristic aim, remember, is to secure acceptance and action as a consequence – is acted upon. Consider a billboard put up on a deserted desert road with the message: ‘Life is hopeless. End it now’. What it advises will be read by few and acted upon by even fewer. Yet that it might be acted on by even one person is a good reason not to put up the billboard.

Second, it is a failing of an adviser not to advise with that characteristic aim in mind. That the advice will not, as it were, hit its target does not absolve the adviser from the responsibility of trying to aim at the target. The good adviser advises in order to persuade and advises well only so long as what is advised would be good whether or not it is accepted. A philosopher who sincerely and reasonably believes their advice will have zero chance of being adopted might be thought not to advise but rather to bear witness to what is right. Yet those who testify are obligated to assert what they honestly believe to be the case and it matters that what is thus affirmed might be acted on whether or not it is. The aim of advising, if not always its outcome, is to persuade whoever hears the advice to do what is advised and that remains true even if no-one hears it or acts on it.

There is a further important point to be made about the kind of philosophical advice under consideration. In the case of personal advice a simple statement of what ought to be done by the individual implies that it could be done and indeed clearly displays what it is that must be done. In the case of policy advice there is normally no such simple entailment of who can and should do what is needed. Rarely can a policy adviser be assured that what is advised will be straightforwardly acted on. Yet there is an important further difference between the personal and policy advice that has to do with the significance of the latter. This tells in favour of exercising more caution in such a case than when offering personal advice. This has to do with what it means to counsel action in the political sphere. Here, we are talking about the exercise of the power of law over individuals.

Caution is urged by those who recognise the significance of this fact. For instance, Thomas Nagel has insisted upon a certain restraint in the urging of such a coercive exercise when individuals intractably disagree morally (Nagel, Reference Nagel1987). It should not, he argues, be just that I, who have power, force you to do what you may disagree with, that it is a matter of one side of the moral argument winning. Moral disagreement requires that any exercise of power be reasonably justified to the other. Thus, Nagel urges that those who have moral certitude as to what should be done ought further to assure themselves that they are certain enough to be justified in forcing others to comply. There is, in other words, a further burden of justification upon those who advise political change that goes beyond what is needed to be sure that what one advises is warranted.

Of course, those who advise do not exercise coercive power, but they counsel those that do. Hence, the particular concern they should have as to whether their advice is followed. For when it is a matter of personal advice the advisee is only whosoever faces the moral dilemma that might prompt the advice. When it is a case of advising the adoption of law or policy, all within the purview of the relevant jurisdiction are subject to the coercive consequences of that advice being adopted. The recipient of personal advice may choose whether to follow it; the citizen subject to law and policy has no reasonable choice whether to comply.

8. ‘I am not writing for them’

There is a subtly different response that the philosopher might make from that which professes either not to be speaking as a policy maker or not to make any difference anyway. This is that the philosopher who publishes is speaking to other philosophers. As such responsibility extends only to the influence that one might have on one's intellectual peers. Moreover, the way in which a philosopher speaks to other philosophers is very different from that in which they might seek to persuade those who make policy.

This response is unpersuasive as a renunciation of a broader responsibility. The philosopher certainly does not speak directly to the maker of policy. Nevertheless, the philosopher does speak about what the maker of policy is concerned with. The philosopher aims at changing policy even if the target is a long way removed from the published piece. The distance in question is also somewhat minimized once one acknowledges that the influence of the philosopher can be indirect. The published work may be read by others, students of philosophy, academics, and intelligent lay persons, who in turn can influence policy makers or themselves make policy.

It is true that philosophers must publish work that can persuade other philosophers. Their mode of rational persuasion may be obscure to non-philosophers. However, making that concession should not be taken to suggest that speaking to one's peers and speaking to a lay audience must always be in different voices. There are countless examples of philosophers whose work, without a significant change of language, has served to influence both categories of reader. Peter Singer's work on animals and on the duty of assistance we owe to the global poor is a good example.

Yet philosophers need to be careful and judicious in how they present their cases. We noted earlier Dan Brock's worry about how the non-philosophers who make policy might be confused in their understanding of philosophically important distinctions and arguments. The worry should motivate, at a minimum, a commitment to clarity and precision. It should also motivate a willingness in non-philosophical company to present key claims in a form that avoids the risks of damaging misinterpretation. Once again, the obligation is to be clear and unambiguous.

9. Making policy

What then of the case where the philosopher is directly offering advice to the policy maker? Historically there are good examples of philosophers who have acted as counsellors of the ruler: Machiavelli using his own political experiences to proffer advice to his fictional Prince, Aristotle tutoring Alexander, John Locke serving as amanuensis to the Earl of Shaftesbury. However, the most interesting contemporary examples are of those philosophers who have served on commissions or committees with an advisory role. As such they are not directly employed by the Government yet are commissioned to provide them with advice.

The political philosopher, Dennis Thompson, has written briefly and insightfully about the responsibilities of the political adviser (Thompson, Reference Thompson1990, pp. 52–57). It helps then to distinguish in brief terms between the political adviser and the philosophical adviser. First, the advice offered differs in kind. A political adviser may only advise as to a range of options indicating for each the possible pluses and minuses, but not advise the Government to adopt any particular option. A philosophical adviser is expected to make clear recommendations of what would be the best real-world outcome. Or at least that is what the editors of the relevant academic journals expect.

Second, a political adviser may advise on what it is best for their employer to propose – not inasmuch as this is the best law or policy as such, but insofar as the law or policy secures for the relevant member of Government certain benefits, such as popular support, continuance in the preferred position, or promotion to a better job. The philosopher publishes only what they regard as the most defensible practical recommendation.

Third, the political adviser is normally partisan, advising a Government which has clear ideological and political commitments. If the adviser is to retain the Government's ear and job, their advice should at least not be dissonant with those commitments. The philosopher is – at least in principle – non-partisan, their defence of any law or policy not motivated by support for any political party.

However, the political and philosophical adviser do share one important thing in common. They are both offering advice in the present moment and with an eye to what might be done. An excellent and revealing example of what this means is provided by Mary Warnock who chaired the Committee that produced a Report commissioned by the then United Kingdom Government to consider what – in the wake of the first successful test-tube baby – ought to be the law and policy in respect of fertility treatment. Her concern was to defend what was philosophically respectable, but which could also be feasibly adopted as policy.

One key element of the Report was the recommended rule that would limit the life of in vitro created embryos to only 14 days. Warnock thought it important to ensure (and to reassure the public) that certain things could not and would not be done to the human embryo. One way it did this was to characterise the human embryo as having a ‘special’ status. This characterisation has subsequently frequently been described as a ‘fudge’ or a ‘muddle’ (Fox, Reference Fox2000; Brazier, Reference Brazier1999). Philosophers were also unhappy with the Committee's recommendations seeing them as merely endorsements of what was, at the time, ‘roughly acceptable’ rather than morally defensible positions (Hare, Reference Hare1993; Lockwood, Reference Lockwood and Lockwood1985). The selection of the 14-day threshold was seen as part of this putatively indefensible approach.

Yet Warnock deliberately favoured an approach which combined both purely philosophical arguments (of a kind that would find favour in an academic journal) and arguments that had an eye to what makes for good policy that can muster public support (and thus of a kind that would find favour with politicians and policy makers). She was clear, in respect of the latter, that the recommendations made ‘had also, as far as possible, to be acceptable to society as a whole’ (Warnock, Reference Warnock2003, p. 505).

Warnock's approach was to support the 14-day threshold by an ‘amalgam’ of arguments. Some appealed to metaphysical and moral reasons supporting the threshold chosen. Some appealed to the feasibility of adopting the threshold in question as policy: that it would be seen as regulatable by using a clear, precise and determinate limit; that it would be seen as an acceptable compromise between a prohibition on any embryo research and the absence of any constraints; that it would serve to prevent a slippery slope from acceptable to unacceptable forms of research; and that it would allow research that yielded significant, beneficial results.

Warnock did not abandon philosophical argumentation in order to be a successful adviser. Rather she recognised the limits of relying solely on what philosophers might count as good reasons. Her advice was based on both good philosophical reasoning and good political or pragmatic reasons of public acceptability.

The Report's recommendations were substantially accepted and led to the creation of the Human Fertilization and Embryology Authority that regulates fertility treatment and embryo research in the United Kingdom. It is widely recognized as an exemplary body of its kind, and under its regulatory guidance there have been thousands of babies created using fertility treatment and world-leading research done on embryological development. Times may have changed to a point where some of the considerations in favour of the 14-day threshold proposed by Warnock no longer have the force they once did. That does not show that they did not have force at the time; nor that the kind of consideration that formed a part of the amalgam she favoured should not still be taken seriously. In short, her example displays the way in which advice authored by a philosopher can – and perhaps should – be tailored to fit the constraints of policy adoption, namely what could be supported publicly and would be adopted officially.

It is instructive then to compare Warnock's approach with that of Dan Brock whose reflections on being a philosopher on advisory and policy-making bodies were cited earlier (Brock, Reference Brock1987). The disjunctive title of his article, ‘Truth or Consequences,’ suggests a fundamental choice has to be made between two very different normative imperatives: that of the academic philosopher is to pursue the truth, ‘to follow arguments and evidence where they lead without regard for the social consequences of doing so’ (Brock, Reference Brock1987, p. 786). By contrast ‘the first concern of those responsible for public policy is, and ought to be, the consequences of their actions for public policy and the persons that those policies affect’ (Brock, Reference Brock1987, p. 787).

Of course, that contrast is too simple. A significant part of the independent, critical, and unconstrained evaluation of one policy option rather than another will appeal to the consequences of adopting each. What else might a respect for ‘evidence’ mean if not a concern correctly to identify and to appraise the significance of relevant consequences? Striving to reach the truth on what is the best policy to adopt, as much as what is the best policy as such, must mean following the evidence and arguments.

Yet, a philosopher's arguing for what she thinks true or right might have bad consequences. A philosopher who advises that something, P, is the best policy as such might successfully persuade their non-philosophical audience of this but in a manner that means it does not fully or properly appreciate P, or the reasoning that warrants P. Thus, endorsement of P is not an unqualified good since those taking it up may by mistaken inference or by simple misunderstanding endorse views (and thereby feel justified in pursuing courses of action) the philosopher cannot support.

Brock's example – of the relationship between the permissibility of ending life-sustaining medical support and the moral importance of the distinction between killing and letting die – illustrated his worry that using philosophical arguments to persuade non-philosophers of his views might have bad outcomes. These are of good philosophical reasoning being insufficiently well understood.

Of course, it is not only those who make policy who may misunderstand a philosophical argument. Philosophers are guilty of misconstruing what has been said by another philosopher, as many an aggrieved author has alleged. Yet the effects of being misunderstood differ according to the context and policy making is a very different context to the academic forum. Moreover, the defining disjunction of Brock's article obscures the important distinction that his example relies upon, namely that between the pursuit of the truth and how that truth (and its pursuit) is communicated. It is not enough to demand of academics that they should only pursue the truth, that pursuit being viewed as an unqualified good. It is wrong to insist that an academic should have no regard for the consequences of what is said publicly. This is especially true when philosophers are speaking to those who may misunderstand what is advised and act on the advice that has been misconstrued.

There is an obligation on those who offer practical advice to those making policy. It is, first, as Warnock recognized to accept the distinction between what is best as such and what is the best that might feasibly be accomplished, and to advise accordingly. It is, second, as Brock's piece and example demonstrates, to advise in a manner that acknowledges the dangers of one's advice being misunderstood.

Warnock's sensitivity to the constraints of feasible policy making should alert us to an important distinction. Her concern was to defend the best policy that could be adopted. There is also the question of what is the best policy that can and should be adopted because it can be. Understanding what this means helps to provide a warrant for the third, and justified, type of insincerely proffered advice that was briefly alluded to earlier. The key distinction is between reasons in favour of a policy and reasons in favour of the policy being adopted. Reasons in favour of a policy – that it is the morally best approach to a particular matter – are not, and do not provide conclusive reasons in favour of adopting the policy. It might be that the best policy as such should not be adopted. Imagine then that adopting the best policy would lead the Government to lose the crucial support necessary for it to remain in power, and grant that it would be, on balance, better for that Government to stay in power. Conversely, the best policy to adopt might not be the best policy as such. This might be precisely because it would ensure that the Government retained power, even though what the Government ought, in an ideal world, to do is adopt the best policy as such.

Moreover, it could be that recommending what is in fact the best policy to adopt as the best policy simpliciter is needed if one's advice is to succeed in persuading the policy maker. Thus, someone with an eye to the broader consequence of adopting one policy rather than another might insincerely advise that the best adoptable policy is the best policy as such and do so both because of the benefits of its being adopted and because an admission of the real reasons for advising would undermine the chances of the advice being effective. A philosopher might publicly morally compromise when she endorses what it is privately believed does, perhaps significantly, less good than agreeing with others but in making the public compromise thereby help greatly to improve the world (Archard, Reference Archard2012).

10. To conclude

The responsibility of the philosophical adviser who publishes practical normative recommendations of policy is not one that can be denied or significantly minimized. Every philosopher who advises in this manner should imagine that their recommendations are being made to a Minister, someone who can hear, understand, and act on that advice. This simplifies matters for the purposes of the argument. Of course, the philosopher can be taken to advise their peers, the public, an interest group or a social movement. Still, the philosopher is responsible for what is done in the name of such advice. Their obligation is to be clear and unambiguous in their statement of what is recommended, and conscientious in their formulation of the best arguments for that recommendation. They should be honest and sincere in making their recommendations, the only exception being a defensible endorsement of what is best in the circumstances, that is feasible and likely to be adopted; or one whose endorsement and adoption promotes the good more broadly considered.

This may seem unduly modest and perhaps even rather too self-evident to warrant being explicitly laid out as has been done here. In brief response, it should be enough to ask all practical normative philosophers if they do see themselves as advisers, do in consequence of that recognition of their role discharge the stated obligations, and if they can see no failings in this regard, on their own part or on the part of some of their peers.