“That is not only not right, it is not even wrong”
Attributed to Wolfgang Pauli
Most of us have at best a hazy understanding of what politics is. Even those who “do’” politics in the sense of either living “for” or “from” it (to employ Max Weber’s distinction) — journalists, politicians, civil servants, party officials and volunteers, campaigners, lobbyists and so on — are unlikely to have anything more than an indistinct view of what the activity is that they are engaged in. To this list we might add contemporary political philosophers also, who, at least in the analytical Anglo-American tradition, have in recent decades given relatively little consideration to the question of the nature of the phenomenon they study, preferring to focus instead on normative questions about the values and ends that political activity should be oriented toward achieving. Footnote 1 These are important questions: how societies answer them matters and political philosophy certainly has something to contribute to those inquiries. But we might think that political philosophy ought not to concern itself with the nature of actual political practice anyhow, that to do so would be to contaminate the purity of philosophical reflection with the messiness and contingency of real politics.
This essay will argue the opposite: the actual practice of politics places particular constraints on political philosophy that theorists ought to be sensitive to. Put differently, political philosophy should be appropriately guided by the practice of politics about which it seeks to speak. This is not, as we will see, driven by the now familiar concern (of nonideal theory) that contemporary political philosophy has become overly abstract and hence cannot offer action-guiding recommendations for us here and now. Footnote 2 The central claim is rather more in keeping with the realist vein of contemporary political philosophy: that despite the inevitable need for abstraction and idealization, any theory has to retain an appropriate degree of fidelity to the phenomenon that it theorizes in order to be a theory of and for that phenomenon. Footnote 3 Whatever else it might abstract or idealize away, there are certain facts about the actual nature of politics that any political philosophy should take as given.
This essay examines the form such fidelity must take and the limits it places on political philosophy, and it does so primarily through the question of political values. It begins by arguing that a political value must be consistent with the constitutive features of politics as a human activity. The point here is not to draw a sharp distinction between moral values and political values, but to say that the phenomenon of politics itself provides conditions against which any value — including but not limited to moral values — must be consistent if it is to be a value for the political domain. Political values may have their origin outside of politics, but for them to function as values for the political domain, that is, become a value we aspire to realize in practice, a standard of evaluation against which we assess the actions of political agents, or a category through which we seek to understand political life, it must be possible for us to view them as consistent with its constitutive features. The following section then examines the extent to which the preconditions for the realization of values in practice ought to figure in our considerations as to whether they are values suitable for our society. One way of thinking about these sections is as responding to two different but related questions we can ask of values: respectively, “Is this a political value?” which is to say, “Is it a value that is appropriate for the political realm?”; and then, “Is this a political value for us?”, which is to ask the further question as to how far it might be a value that fits or belongs to our social world (which includes but is not exhausted by potential concerns about whether it makes sense for us to aspire to realize that value here and now). The final section then responds to David Estlund’s claim, taken as representative of a particular form of ideal theory, that it is not a defect of a political philosophy if it pays no heed to the realities of politics. Distinguishing between a theory of politics and a theory of a particular political value (justice, freedom, and so on), I shall argue that in the case of the former it is the constitutive features of politics that provide the conditions that give meaning to the enterprise as an exercise in theorizing politics. Moreover, and in response to the claim of ideal theory that the meaning of political values can be understood quite independently of any account of the character of the political domain, I suggest that the meaning of a political value can only be ascertained within an at least minimally plausible description of politics as a distinct sphere of human activity. In these ways it is wrong to maintain that political philosophy need not concede anything to political reality.
I. The General Conditions of Politics
Politics is a particular sort of human activity. It may be related to other spheres of human life, such as morality, economics, and law, for example, but it cannot be reduced to any of them. Footnote 4 One way of thinking about the specificity of political domains is in terms of its constitutive features, those facets of political life that are both essential to politics and help distinguish it from other social practices. These features, by virtue of being essential elements of politics, ought also to be taken as given or fixed points in any philosophy of politics (so I wish to claim). Taking the theorization of political values as our example, in order for values to be meaningfully understood as values for the political domain, they have to be consistent with its constitutive features. These constitutive features we can call the general conditions of politics insofar as they are necessary features of a specifically political domain and, therefore, conditions that any value must be consistent with in order to be a value for that sphere of human activity. By focusing our attention on what must be true of politics as a particular sort of practice we can get a handle on the ways in which that practice itself provides conditions of intelligibility for any appropriate theoretical or philosophical reflection upon it. It is to be expected that people will disagree about what these conditions might include, thought it is important to recognize that the claim about the need to take seriously the general conditions of politics is distinct from claims about what those conditions might be. Nevertheless, as a way of exploring what it would mean for a political philosophy to be consistent (or not) with the general conditions, I shall here focus on what I take to be three strong candidates for inclusion grounded in a particular (broadly realist) way of understanding the nature of the political: that politics is an attempt to provide order via authority and legitimate coercion in conditions of disagreement. Footnote 5
A. Disagreement
Politics is a response to and necessarily takes place in contexts where a group of people require commonly binding decisions on subjects deemed to be of public concern but over which there is disagreement. So, to give some illustrative examples: questions over the distribution of public goods; how the burdens of taxation ought to fall across the population; whether the state should provide public health care; if abortion should be illegal; what the relationship between the state and the church should be; the need for conscription and compulsory education; where the limits of free speech and toleration should be drawn; whether minority groups ought to have special rights, and so on. Our disagreements are not only about ends, however. We regularly disagree as much about the means for pursuing our collective ends as we do about what those ends should be. Even if we agree that the richest ought to pay the highest percentage of tax, for instance, there will still be disputes as to the most efficient and fair way of collecting that revenue from them. Moreover, individuals will often be committed to competing and incompatible normative and conceptual frameworks for coming to collective decisions. Unlike the natural sciences, there is no consensus upon the right methodology or epistemology for determining public policy issues (Christian, utilitarian, socialist, conservative, Kantian, Marxist, Nietzschean, liberal, and so on). Footnote 6 Hence we find ourselves in the ominous position of not only disagreeing what to do but of disagreeing as to the right justifications and considerations that ought to bear on the decision about how we ought to collectively proceed.
Politics is a response to, and necessarily takes place in, these contexts of disagreement and conflict as to what should be done, what Jeremy Waldron has usefully called “the circumstances of politics.” Footnote 7 This makes it important to distinguish, as he remarked elsewhere, the question of what we ought to do from what we ought to do when we disagree about what we ought to do. Footnote 8 While we will all have our views on the first question, politics is specifically geared toward providing an answer to the second. It is also important for the character of politics, and for reasons that we shall come to shortly, that our disagreements about what should be done very often remain even after a collective decision has been made. The circumstances of politics are not resolved once a judgment has been reached, but they provide the permanent backdrop against which politics takes place. This is in large part why the notions of authority and legitimacy must be central to any plausible understanding of politics.
B. Authority
Politics does not have its own substantial content other than to provide order among people who require commonly binding decisions. While there might be many ways in which collective decisions could be reached in conditions of disagreement — a lottery or the imposition of one party’s will, for instance — politics seeks to settle disputes via authority. It may well be possible to settle disputes through brute force alone but we readily recognize that this is to rule coercively not authoritatively, or, in other words, to not rule politically at all. Political rule is authoritative rule. In an authoritative order the ruled obey the rulers because they recognize that their decisions have a certain sort of normative force which creates obligations on them to obey. These obligations are taken to hold (within limits) regardless of whether individuals take it to be in their interest to obey or whether they disagree with the decision that is made. Politics is therefore characterized by a particular sort of relationship of rule whereby the rulers’ commands are transformed into authoritative decisions that those subject to it are obliged to obey.
C. Legitimate coercion
To claim to rule authoritatively is to claim that there is some normative (but not yet necessarily moral) ground that justifies the de jure right to have one’s decisions obeyed. That is to say, political rule must also be legitimate rule. And so any political order must be intelligible to those who live under it as a form of legitimated power. Footnote 9 It is by virtue of the fact that the authority is legitimate that its decisions possess the requisite normative force to transform what might otherwise be conformity due to threat of overwhelming force into a sense in which the decisions of the rulers create obligations on the ruled to obey, regardless of whether they agree with those decisions or not. This need is most pressing when we consider that any authority must be empowered to coercively enforce its decisions upon those over whom it has dominion, for we assume that, given the ineradicability of disagreement, it is always likely that there will be at least some, and sometimes many, who disagree with the decision that has been made. As Mark Philp has put it, “part of what makes rule political is that it is exercised over others who are not inevitably compliant.” Footnote 10 In a successful political order, the majority of people will conform with its decisions on the grounds that it is taken to be a legitimate authority. Yet the need for an enforcement agency that has the ability to impose solutions, where necessary through coercion, remains intrinsic to politics, for without it there is always the risk that disagreements will unravel into the unrestrained conflicts of interest and values that politics is intended to resolve. This capacity may only be a last resort to which an authority must have recourse, and there will always be good reason for associations to try to replace coercion, which can often be inefficient, with more indirect inducements to obedience. Footnote 11 Nevertheless, and for whatever reason, full compliance will not always be forthcoming, and in such scenarios a political order must have not only the de facto ability to secure obedience through coercive force (or the threat of it) but the de jure right to do so also. And so it matters for the question of whether a form of rule is political whether those who live under it see themselves as obeying merely out of fear of coercion or because they understand their situation as one of being subject to a legitimate authority that creates normatively binding collective decisions.
Politics is a practice characterized by disagreement, authority, and legitimate coercion. It may be characterized by much else besides, but these are at least some of the characteristics of the human practice or activity that political theories seek to be about, and the context in which political values are asserted, claimed, debated, critiqued, and so on. Being constitutive of politics, political values must take them as fixed features of the political domain. This means that for a value to be a value for politics it must be fully consistent with their presence. It cannot be incompatible with the general conditions such that a belief about a value is inconsistent with any particular constitutive feature of politics (for example, political freedom is the absence of political authority), nor, as an assumption built into the understanding of the value itself, can it depend upon the general conditions being overcome for their realization in practice. In such cases the value would not be a value for the political domain but for a world in which politics or the need for politics is absent (which is, whatever else we might think about the attractiveness of such a world, not our own). It would not be a value suitable for the activity of politics.
As a way of fixing ideas it will be helpful to try to give some examples of how values might be found inadequate in this way. First of all, political values have to make sense in conditions of disagreement. An elaboration of a value that wishes away such contestation is going to be inappropriate for any political space. For instance, a conception of freedom that assumed that a free society would be one in which all people shared the same ends or values, or even only the same conceptual framework for making decisions on matters of public concern, would not be political insofar as it assumes the absence of the disagreement to which politics responds. The value of freedom makes most sense precisely in conditions where we disagree what to do (a point we shall return to shortly). Likewise (and to use an example defended by Andrew Mason in his contribution to this issue), if a theory of justice requires that all people endorse the same principles in order for a society to be fully just — so that they are motivated by the commitment to acting justly rather than in response to threats of punishment and sanctions — then this is clearly inconsistent with the general condition of disagreement also. Our disagreements about justice are among the deepest and most prevalent that divide us, and it is for this reason that providing commonly binding decisions on matters of justice is one of the most important tasks of any political order. To think that a fully just society would be one in which all people converged on the same principles of justice is, in an important sense, to put the possibility of full justice outside of the political realm and into a world devoid of much of the original impetus for politics in the first place. Footnote 12 That might be an admirable vision of society, but it cannot be a political society.
A failure to take disagreement sufficiently seriously can often also lead to a similar failure to pay sufficient heed to the need for political authority and the possibility of legitimate coercion. One set of problematic values with which we are familiar from both the history of political thought and more contemporary work are those that deny that politics requires coercion at all insofar as freedom, justice, democracy, autonomy, or legitimacy demand that we only obey those laws which we can represent as being ourselves the coauthors of. Under such interpretations, often Rousseauean or Kantian in inspiration, we are only free, the political association only legitimate, and so on, insofar as we obey only our own will. Politics is a sphere antithetical to being ruled by others. Yet the point of political order is precisely to provide an authoritative decision in conditions of disagreement about what should be done (“what we ought to do when we disagree about what we ought to do”), and to enable the legitimate coercion of those who would not otherwise comply where necessary. To assume that people can or will agree, even if only hypothetically, such that they can all obey the law and yet only obey themselves is to theorize for a scenario in which these general conditions, and hence the possibility of legitimate coercion also, are absent. This position denies the very possibility of being forced to act against your will that seems inevitable once you accept that politics takes place in a space of genuine disagreement which necessitates an authority that is not only able to decide but to legitimately enforce its decisions. It is for this reason that the problem that Rousseau sets himself in his Social Contract, to “Find a form of association which will defend and protect, with the whole of its joint strength, the person and property of each associate, and under which each of them, uniting himself to all, will obey himself alone, and remain as free as before,” can only be answered by rejecting the incongruity between force and freedom. Footnote 13 Yet the best route for political theory is not to deny that the tension between legitimate coercion and freedom exists, nor to try and dissolve that experience through philosophical ingenuity, but to help us try to find ways to live with it in an acceptable manner.
Recognizing that politics necessarily entails the possibility of obeying a will other than your own means that we should acknowledge how politics inevitably imposes costs in terms of freedom. The prevalence of disagreement and contestation ensures there will always be cases where we find our government pursuing policies in the name of particular interpretations of values such as justice or equality with which we disagree, and which we rightly experience as a restriction of our activities. And this is true even when we fully recognize the political order as having the right to make such commonly binding decisions. In such scenarios it seems reasonable to resent such restrictions for curtailing our freedom in the way that they do. An appropriate conception of freedom for politics therefore must recognize such costs and make them consistent with what it can mean to live in a free political society. However, any value that gave us reason to resent the mere presence of an authority with the right to coercively enforce its collectively binding decisions does not deserve to be taken seriously as a conception appropriate for the political realm. A freedom that does not allow for the need for authoritative government with the right of legitimate coercion cannot be a political conception of freedom. Such values, if they exist, would not belong to politics. Footnote 14
II. The Preconditions of Political Values
What I have suggested so far is that distinctively political values must be consistent with the general conditions of politics. If they are not, then by virtue of the fact that they are inconsistent with its necessary constitutive features they cannot be in any meaningful sense values for the political domain. There is a related but importantly different question that we might ask of political values also — whether it can be a value for us given facts of the political society of which we are a part. One good way of getting to this issue is via Bernard Williams’s discussion of “Saint Just’s Illusion.”
Saint-Just was a Jacobin leader notorious for the zeal with which he conducted the Terror in his attempt to remake French society according to the ideals of civic virtue associated with the Roman Empire. The human devastation caused by Saint-Just’s attempt to impose republican ideals on French society was clearly an ethical travesty, and we can rightly construct ethical arguments against his actions. Yet there were several dimensions to his mistake, and his failure of ethical understanding was only one of those. One other dimension was of historical interpretation. As Benjamin Constant noted in his essay distinguishing between the liberty of the ancients and the moderns, the ancient conception of liberty employed by the Jacobins, which revolved around and depended upon very strong individual dedication to public life, was calamitously unsuited to the modern world. This trades on a fairly common thought that “the preconditions of political freedom vary with different social formations,” and that, whatever we might take those preconditions to be, where they are absent that conception of freedom is not going to be able to satisfy the need for freedom of that society. Footnote 15 In itself, overlooking the historical preconditions of republican liberty might be fairly harmless, though in Saint-Just’s case this was conjoined with a further — more political — error: a failure to understand the difference between two different sorts of spaces that we rightly and naturally treat differently. The first
is the space of our actual social and political life, within which we encounter various political and ethical demands and ideals, argue with them, adapt ourselves to them, try to form a conception of an acceptable life within them. The other space, of which we may be conscious only in a very shadowy way, is of other conceptions and ideals and world pictures that human beings have had, and may perhaps still have elsewhere, which are not part of our social and political space, are not even starters for a life we might now lead, and are — strictly in that sense — alien to us. Footnote 16
Saint-Just mistook the ideals of Roman republicanism as belonging in that first space, when really it was so incompatible with late eighteenth-century French conditions as to render it an impossibility as a way of life. To be sure, his French contemporaries could recognize republicanism as expressing human ideals, and knew that a historical story could be told that linked those ideals to their own (as we can today). But Ancient Rome was a world so different from theirs in terms of its social structures, economic forms, and people’s needs, beliefs, and motivations, that these values had no place in the space of their actual social and political lives. Footnote 17 The delusion from which he suffered, his “illusion,” was to mistake his utopian vision of the way he wanted the world to be from the way it actually was.
It is important to recognize that the general mistake illustrated by Saint-Just is not to be thought of as a conceptual or philosophical error. It was rather a failure of social and political understanding whereby we take a particular value to belong to our political and social world when it does not. And not because it cannot be made sense of in relation to the general conditions of any political association, but because it relies upon a conception of a social world that is at too great a distance from our actual world, where that distance “must be measured in terms of political considerations of relevance and practical intelligibility.” Footnote 18 Saint-Just was not wrong in thinking that republican freedom represented a politically coherent elaboration of freedom, he rather misjudged the extent to which eighteenth-century France possessed the preconditions for republican liberty, and in doing so missed the point that what was liberty for the Romans could not be liberty for his fellow Frenchmen.
The general point to take from the example of Saint-Just is that if a society lacks the preconditions necessary for a value then it cannot be a value for those people, even if it is consistent with the general conditions of politics. Such values might properly speaking be considered political values, but they cannot be political values for them. They simply do not represent values that can be integrated into a life they can coherently lead. Examples of such values for us today may include: conceptions of justice that rely upon a high level of moral and religious harmony which does not pertain in diverse societies such as ours; aristocratic or war-based interpretations of the virtues of dignity or honor; freedom, if they take basic economic competition to be a violation of what we are at liberty to pursue when modern societies are commerce societies; democracy, if it requires sustained direct participation from its citizens in a culture of widespread political disengagement and disenchantment; basic rights, if certain groups are not to be afforded them by virtue of their ethnicity, race, or religion; or equality that depends upon higher levels of class solidarity than currently exists. It is vital to stress that there are many questions in play here, all of which are rightly to be thought of as matters of political judgment: What are the preconditions of any particular value? Do they hold in our present society? If they do not, just how out of reach for us are they? The latter question may be particularly difficult to answer given that it requires a further judgment about how much the future will look like the present. We must always allow for politics’ ability to surprise us — sometimes dramatically so — but this does not mean that there is no such thing as a considered judgement that can be reached on these matters. Certainly philosophy will not be able to determine any of these questions via its own resources alone and hence cannot simply assume a particular answer in advance, despite how amenable it may be to an author’s favored values. To get anywhere near the truth of the matter it is crucial that our judgments be guided by a sensitive, sincere, and truthful understanding of how our society is structured, the ways in which its main institutions and practices function, and its peoples’ actual beliefs and motivations. To engage in wishful thinking, or indeed the different but related temptation to simply wish away such features of our context as mere inconveniences in the face of our preferred ideal, runs the risk of falling prey to Saint-Just’s illusion and failing to heed the difference between our actual social and political space and those in which our preferred ideals might represent viable options.
III. Ideal Theory and Fidelity to Reality
What has come to be known as ideal theory really represents a spectrum of positions. One such position is identifiable by its being the target of nonideal theory. This is essentially a dispute regarding the extent to which facts about the likelihood that a normative theory might be realized in practice ought to impinge on our evaluation of that theory. Is a theory defective if it is being proposed for a society that lacks the preconditions for its realization, or potentially more worryingly, if the preconditions for its application could not realistically be expected of any human society? Is it to be considered a deficiency of a theory if it is insufficiently sensitive to feasibility constraints that limit what is possible and in doing so effectively renders itself unable to provide “action-guiding” recommendations for us here and now? What if human nature is such that it is highly improbable that people would be able to live by particular principles of justice? Ought that to count against those principles? Footnote 19 Nonideal theorists tend to answer yes to these questions, though there is considerable disagreement as to how facts of human nature or other feasibility constraints ought to alter how we go about doing political philosophy, especially in its normative guise. Footnote 20 Ideal theorists, on the other hand, reject such considerations. As David Estlund has argued in his influential articulation of ideal theory: “moral theories of social justice, political authority, political legitimacy, and many other moral-political concepts are not shown to have any defect in virtue of the fact, if it is one, that the alleged requirements or preconditions of these things are not likely ever to be met.” Footnote 21 More succinctly, “. . . the truth about justice is not constrained by considerations of the likelihood of success in realising it . . . .” Footnote 22 According to ideal theory, the objective of political philosophy is to take up a perspective on politics that, through abstraction and idealization, is unencumbered by facts pertaining to the plausibility that a theory might ever be realized in practice. These are quite irrelevant to the philosophers’ task of getting at the truth of the matter. Philosophy is charged, as G. A. Cohen memorably answered, with telling us what to think rather than what to do, “even when what we should think makes no practical difference.” Footnote 23 If it turns out that justice is not realizable in this world, maybe in any human world, then as deeply regrettable as that may be at least we know what justice is. And that is not nothing, for, as Estlund argues, there is value in understanding something of value. Footnote 24
The concerns I want to pursue here in light of what has already been argued are somewhat orthogonal to those expressed in the ideal/nonideal theory debate, and are more closely associated with political realism. Footnote 25 It is therefore important to be clear that my target is a particular form of ideal theory, and one that might not include the work of certain — nevertheless quite prominent — ideal theorists such as John Rawls (which may give us good reason to be skeptical that the term “ideal theory” remains of much use). Whereas ideal and nonideal theorists are concerned with the question of whether the applicability or realizability of a theory of (say) justice in any way ought to feature in our evaluations of it, the claim I want to focus on and reject is that the nature of the political domain ought not to impinge on our theorization of it, including in determining the meaning of political values.
Estlund’s own version of ideal theory rejects both the nonideal and realist charges. Addressing the latter, he imagines a realist objection to a theory the structure of which goes something like this:
It is characteristic of political societies that they have laws and law enforcement agencies (e.g., police, judges, and juries).
If all people acted moral flawlessly, then there would be no need for law and law enforcement agencies (allowing for the fact that not all crime is immoral).
Therefore, any theory that assumes that people act morally flawlessly cannot be a political theory.
“Would it [such a theory] be assuming away politics itself?” he asks. “If so that sounds like a fatal defect in a political philosophy.” Footnote 26 Estlund’s response to his own question is worth quoting in full:
A lot of work is being done in this objection by a definition. A theory’s subject matter is asserted to lie outside of politics unless it grants a substantial role to laws, police, criminal courts, and so on. Consider a theory that gave compelling arguments for the conclusion that a society could not be characterized by political justice, or authority, or legitimacy in conditions where there was a substantial role for laws, police, and courts. On the definition of politics in question, this would not be a political philosophy. But that is only because politics has been defined out from under it. Fine, let it not count as a political philosophy. This would leave entirely intact its claim to have the correct theory of justice, authority, and legitimacy. Footnote 27
Estlund focuses here on law and law enforcement because he has chosen to address the specific fact (if it is a fact) that people will never be moral angels, but I assume that the point is intended to be quite generalizable: that it is no defect of a political theory if it does not take adequate account of the nature of politics. And certainly such facts about politics do not impinge on the meaning of political values. I want to challenge both of these claims on the grounds that they violate what I take to be a general requirement for the theorization of any phenomenon: that a theory must be consistent with the necessary constitutive features of a phenomenon, those without which it would not be the sort of phenomenon it is, in order to be a theory of that phenomenon. If it is not, then it lacks the requisite fidelity to the phenomenon that it seeks to be a theory of, and hence cannot be a theory for it either.
Moving by analogy is likely to be helpful here. To begin with an admittedly crude scientific analogue: Imagine a scenario in which a precocious young scientist claims to have discovered that all previous theories of how and why hydrogen (in its most common isotopic form) reacts the way that it does with other elements are incorrect, and that she, during her doctoral research, has developed a better theory. When she published her research, however, it turns out that the theory only works if we assume that hydrogen has two protons, two neutrons, and two electrons. What would the right response to her theory be given we know that in reality hydrogen has only one proton, no neutrons, and a single electron? The theory might have the virtue of being internally coherent on its own terms, free from any contradictions, flawed reasoning, or inconsistencies. But even if that were true we would nevertheless insist that it is still a bad theory, though bad in the very specific sense that it is not a theory of hydrogen because of what we know to be true of the composition of hydrogen atoms: it fails to qualify as a theory of hydrogen at all (this is what is taken to be meant by the phrase, attributed to the physicist Wolfgang Pauli, “That is not only not right, it is not even wrong”).
Social practices like politics are quite obviously different from atoms, and it would clearly be a mistake to think that a theory for the former needs to fulfill the same epistemic requirements or human needs as the latter (not least of enabling accurate and repeatable predictions). But nevertheless it remains a common requirement that for something to qualify as a valid theory of a particular phenomenon it has to take its constitutive features as given. So let us move to an analogy of a theory of a human practice that might be deficient in this regard.
Imagine that you and your partner are expecting your first child and, as nervous parents-to-be, you sign up for a course on “What makes a good parent?” led by someone who claims to be an expert in bringing up children. You arrive at the first session and the expert begins by declaring that “Parenting is a voluntary economic relationship with the aim of maximizing profit.” The expert then goes on to set out a series of ideals of good parenting and the values with which it should be concerned in light of that understanding of what parenting is: Parents ought to be assessed according to how much profit they make from their dealings with their children; the maximization of profit is the ideal to which parenting should be orientated; parents have no natural duties of care or compassion to their children and are free to enter into a parenting relationship with anyone they please as long as both parties consent; any obligations that do pertain between parents and their children only arise as the result of a freely made agreement between the two parties, and which both parties are free to leave within the terms of the agreement itself; parenting is a relationship between individuals who are equal in terms of rationality and hence are both to be understood as fully responsible for their decisions; and so on. It is obvious that the right response to this “expert” would be to profoundly disagree with him. But about what should we disagree?
At one level the disagreement will be about what makes a good parent, and one would likely disregard his ideals as wrong and highly improper. But what has really gone wrong is not simply that he is offering a bad theory of parenting, but that he has not offered a theory of parenting at all. He fails to be an expert on the practice about which he claims to have expertise. What he has to say simply misses the mark in terms of being a theory for the practice of being a parent. And the reason why it does so is because it mischaracterizes or outright fails to recognize what we would usually take to be constitutive features of parenting as a particular practice and form of human relationship which is crucially different from other practices such as economic exchange or, more closely, friendship. Such features might include that it is primarily an ethical not economic relationship; that parenting takes place within a family unit (however defined) where certain individuals have particular special responsibilities for their children, and will usually include that period from birth until the child becomes an adult (with the assumption that the two parties are not equal in terms of rationality and responsibility until that point); that it will be a relationship in which the virtues of love and care will be central; that its end cannot be profit; that it might be voluntary but it need not always be (or at least certainly should not be conceived straightforwardly in those terms); and so on. While this is to disagree with the theory being offered, it is crucial that the disagreement hinges on the question of whether it is a theory that resembles any plausible description of the practice of parenting. This is a matter over which there will be legitimate disagreement, and it goes without saying that there are significant disputes as to what it means to be a parent. But at the very least there will be certain descriptions of the practice that we would normally recognize as thoroughly implausible or deeply mistaken (such as the one offered by our expert). In lacking any appreciation of what we might take as constitutive features of the practice of parenting we can say that it cannot function as a theory of parenting, by virtue of which neither can its ideals function as ideals for the practice of being a parent (and that holds even if it might be a plausible ideal for other human relationships). Footnote 28
To bring this back to politics: the first of Estlund’s claims was that it is no defect of a political theory if it does not take adequate account of what politics actually is. Indeed, Estlund’s defense of ideal theory on this point makes it sound very much as if he considers the question of whether something is to be thought of as political philosophy as merely a semantic issue, a matter of competing stipulations of what politics can mean (maybe of the even less important question of what the disciplinary boundaries ought to be between political and moral philosophy). And because these are just competing stipulations it cannot be a defect of a political theory if it fails to “grant a substantial role” to the content of any. These analogies have suggested that the contrary is true. It was clearly a defect of the scientist’s theory that it failed to work with an adequate (in this case we would say true) account of the actual composition of hydrogen atoms. The parenting expert employed what we would consider an erroneous or flawed understanding of what the practice of being a parent consists of. In both cases the failure can be expressed in terms of not being a theory of the phenomenon that they seek to theorize. That failure of fidelity to its phenomenon gives us cause to doubt whether what is being is offered is indeed a theory of that particular phenomenon. By analogy, whether a political theory works with a plausible description or understanding of the practice of politics should also be taken as crucial in helping us evaluate whether it truly is a theory of the political domain, and not something else (or, indeed, a theory of nothing insofar as there is no phenomenon to which it relates). A political theory is deficient if it is unintelligible as a theory of politics. Politics is, of course, more like parenting than hydrogen, not least in that it too is a human practice that stands in need of interpretation, which means that it is perfectly reasonable to expect there to be competing accounts of the character of the political domain (there are less likely to be competing accounts of what hydrogen is). Yet we rightly think there are better and worse, more or less plausible, descriptions of politics. Especially where the description is wildly implausible we have good reason to critique those theories as not being theories of politics, in the same way as we can criticize the expert on the grounds that he failed to offer a theory of parenting. It might be a theory of something else — economics, morality, law, and so on — but it cannot be taken seriously as a theory of politics. So it would hardly seem appropriate to think of that qualification, which necessitates a concern with the actual nature of politics, as irrelevant to the task of political philosophy.
Of course, the claim against Estlund’s ideal theory cannot be that it makes the same elementary mistake as the scientist or the parenting expert in working with an implausible description of politics. That cannot be the claim because his response to the realist charge was that ideal theory need not employ any description of politics at all. The worry is not that ideal theory makes factual errors, but that it does not think facts about politics matter either way. Yet the point of the analogies was to show that it is a failing of a theory if it cannot reasonably be understood as a theory of its purported phenomenon, and that in order to make that assessment we need to ask if it works with a plausible description of what it seeks to be a theory of. And so if we do not know what politics is, then we cannot know that what we are talking about or trying to understand is politics. Footnote 29 The concern for ideal theory is that this assessment cannot be made without reference to an account of the political domain that it denies is required. And so we are left with no criteria for assessing whether what ideal theory offers is really a political theory, a theory for the political realm, or a theory of something else altogether. Here theory must concede to reality.
Part of what seemingly encourages Estlund to dismiss these concerns about the fidelity of a political theory to political reality is an assumption that questions about the actual character of politics can only be matters of arbitrary stipulation or argument by definition. There is little reason to think this is right. Few would deny that there is no distinction between true and false beliefs about politics, and certainly our confidence in this need not be undermined by any concern that this distinction cannot be drawn with the greatest precision, nor that there is no supra-human vantage point available from which we can identify such beliefs once and for all. Footnote 30 Indeed, we have many theoretical and empirical resources that can help us understand the nature of the political domain and distinguish more from less plausible accounts. As we have seen, no particularly sophisticated theory is required in order to assess certain accounts as wildly implausible by any reasonable standard, and such accounts can usually be recognized as such by most people. This is unlikely to exclude much that any sensible person would propose (the analogies offered above were purposefully outlandish to demonstrate a point), but we should not think that it excludes everything. Philosophy’s tools of abstract reasoning and conceptual analysis can also get us some not inconsiderable distance in helping us think through what a coherent theory of politics is going to look like. Yet what counts as a credible description ought to rely as much on empirical considerations as it does on more conceptual inquiries regarding the nature of politics. The resources of abstract reason such as conceptual analysis will play an important role but can only get us so far, at which point our understanding will need to be supplemented by empirical considerations and the sort of analysis that is provided by other fields such as anthropology, sociology, economics, psychology, and history. Political philosophy must therefore be highly impure, not in the sense of turning to the real world once we have in view the ideal which we now wish to put into practice, but in terms of using our best understanding of what politics is like, how it functions, its limits, ends and means, to set the very conditions of what is going to count as philosophizing about politics at all. Footnote 31 This is not mere stipulation, but rather being guided by the best understanding of politics we can muster. Footnote 32 And certainly it is not philosophy alone that is going to get to adjudicate on the question of whether a theory appropriately comprehends politics or not.
Estlund’s second line of defense is that even if we grant the claim that a theory that does not take into account relevant facts of politics should not be called political philosophy, that does not impinge in any way on its ability to get at the truth of political values. Here Estlund’s argument seems to trade on an implicit demarcation between theories of politics and theories of political values that is worth making explicit. Theories of politics provide an account of what politics is. In doing so, they also help us explain why politics is the sort of activity it is and not something else, and potentially explain its relationship to other human activities also. It is these theories that I have been arguing so far must take the constitutive features of politics as given in order to be meaningfully understood as theories of politics (and that it is a failure of a theory if it cannot be understood in this way). Attempts to provide theories of politics have fallen somewhat out of fashion in recent decades, though twentieth-century texts such as Carl Schmitt’s The Concept of the Political, Michael Oakeshott’s On Human Conduct, Sheldon Wolin’s Vision on Politics, Bernard Crick’s In Defence of Politics, and Hannah Arendt’s The Human Condition demonstrate that such lack of interest in the question “What is politics?” is a relatively recent development. Footnote 33 Theories of political values seek to elucidate the meaning or content of particular political values, and the exemplar of this way of doing political theory in recent decades remains John Rawls’s A Theory of Justice. Footnote 34 Estlund (like Cohen) rightly takes ideal theory to be a theory of political values, rather than of politics, insofar as its aim is to uncover the meaning of justice, freedom, and so on. And his response to the realist charge above indicates that he thinks these theories are sufficiently independent, at least to the extent that one can provide a theory of political value without also needing or presuming any theory of politics, and, furthermore, that even if there were some connection between the two, a defect of a political philosophy that it were not political would not have any bearing on a theory of political values. The meaning and content of political values, Estlund claims, does not to depend upon any particular stipulation or description of politics.
Yet, as the example of the parenting expert demonstrates, the descriptive and the normative cannot be so easily separated. A theory of the values appropriate for any practice is going to depend upon a viable account of the practice in which those values are to function as ideals to be realized, normative standards of assessment, or categories of understanding. Put differently, a plausible description of the practice must precede the provision of a normative theory for it. If it does not, then just as the ideals and judgments of our parenting expert were found to be deficient because they took parenting to be something other than it really is, the same will likely be true of those offered by political theorists. Flawed political recommendations will inexorably follow from false beliefs about the political sphere to which they are supposed to apply. We should quite expect that the values and ideals appropriate for a domain characterized by one set of constitutive features would not be appropriate for a very different domain where those features do not hold. It would be surprising, for instance, if it turned out to be of little consequence for the question of what a free society would look like whether or not disagreement and the use of legitimate coercion were constitutive features of political orders. Likewise, we should surely think it pertinent to the question of equality in political societies whether political power and authority must be asymmetrically distributed among its populace or whether that is only a transient feature of pre-egalitarian politics. In this sense, the meaning and content of values cannot be understood separate from the domain in which they are to function and the activity which they are intended to help guide. Again, those are not judgments one can make without some appreciation of what makes politics the specific sort of practice that it is, which of its features are fixed and which can be changed. And this is a condition that pertains to normative judgments or prescriptions derived from any nonpolitical sphere applied to politics, such as economics, religion, or technology, as much as morality (as Estlund explicitly conceives of the findings of ideal theory, hence his talk of “moral theories of social justice, political authority, political legitimacy, and many other moral-political concepts . . . ”). To reiterate an important suggestion made in the introduction, the point here is not to draw a strict demarcation between moral and political values (or political and other values, for that matter), but rather to insist that we should hesitate before taking any value seriously as appropriate to the political domain by virtue of its being in whatever other sense desirable, attractive, or preferable.
Any suggestion that reality ought to play some role in constraining or determining philosophical reflection upon politics, especially when it is engaging in normative considerations of what should be done, what justice demands, and the like, often comes up against the criticism that this makes such reflection objectionably conservative or disagreeably status-quo affirming insofar as it allows reality to constrain the possibility of radical critique or undermine the possibility of positing utopian visions of how life might be otherwise. This worry is misplaced. Remember that it is only those values and ideals that are inconsistent with the general conditions of politics that are inappropriate for the political realm. This will not be an empty category of possible values, and certainly it will include interpretations that some political philosophers have offered throughout the ages, including today. But there is no reason to expect that it rules out anywhere near as much as it potentially allows. What can count as political values for us or any other society will always be a subset of all possible political values simply by virtue of the fact that no political order can possess the preconditions for every political value simultaneously. The preconditions for modern liberty make impossible those for liberty of the ancients, for example. Yet this allows for the possibility of there being political values that nevertheless cannot be political values for us, and utopian or radical speculation which takes the form of exploring those values that fall into the former category but not the latter is a reasonable activity. We might ask whether spending a lifetime engaged in exploring values that could only pertain in conditions so radically different from our own is a sensible pursuit. And there are questions about the intellectual motivations for wanting to study politics in that way rather than through a focus on our actual political world which we still struggle to properly understand. Yet there is a significant difference between an enterprise being mistaken and its being unwise. And the general point remains that only the most outlandish utopianism — that which is unintelligible as political at all — is ruled out by the account that has been offered here.
