The World Philosophy Made

by Scott Soames

  These models measure the goodness of each society in terms of its inherent features, without having to first locate its relation to an imagined ideal society, which is not required to plot possible improvements. If the range of systems investigated is finite, one may end up ranked more highly than any other. But models needn’t evaluate all conceivable systems. Even if we included infinitely many systems, there need be no maximum score, and so no ideal. But the key point is more basic: if the goodness of a society is wholly determined by its intrinsic features, there may be no need to locate an ideal to guide progress.

  What would normative reality have to be like in order for this not to be so—i.e., for it to be necessary to first recognize an ideal system in order for us to improve our own society as much as possible? For this to be necessary, what we think of alternatives to our society when considering them as possible stepping stones on a path to improving it must sometimes depend not only on the intrinsic values of good-making features of our own and alternative systems, but also on their overall similarity to the ideal. The most obvious way for this to occur involves cases in which reducing the intrinsic goodness of a political system brings it closer to the ideal system.

  How could this happen? Imagine you are Lenin or Stalin. You believe in an ideal Marxist future, but to achieve it you think you must make your present society initially much less good, in order to replace its institutions with new ones that will put it on the path to the ideal. Fortunately, few find this compelling today. Still, it illustrates the logic of sacrificing present goodness for greater goodness later. Put aside the grotesque historical features of this example. Isn’t it conceivable that the route to the political ideal might justify one in reducing the intrinsic goodness of a society now? Gaus is skeptical; he argues that when we carefully trace the logic of such an attempt, we encounter severe constraints.

  To understand why, imagine a scenario with eleven possible systems, the eleventh being ideal, the tenth being most similar to it, and so on ending with the first, which is least similar. Let the inherent goodness scores, S(x), of the different systems be: S(1) = 10, S(2) = 20, S(3) = 15, S(4) = 30, S(5) = 25, S(6) = 40, S(7) = 25, S(8) = 21, S(9) = 30, S(10) = 39, and S(11) = 45. Distributions like this can arise if the goodness-relevant features of societies are institutions the values of which depend in crucial ways on its other institutions. Gaus agrees that changes that make one institution worse and reduce the overall goodness of a society might, when combined with other existing institutions, reduce the distance to a better society, putting us on a path to further changes which, if followed, would eventually raise the overall goodness of our society.35 This is illustrated by the following figure, where the goodness of a society is represented by its height on the y axis (the higher the better) and the similarity of the goodness-making features of two societies is represented by the distance between them on the x axis (which means that the shorter the distance between two societies, the fewer significant changes are needed to transform the one into the other).

  If normative reality looks like this, then the identification of the utopian ideal, S11, might, in principle, play a useful role in improving our own society. But, as Gaus emphasizes, the range of possible configurations in which the putative ideal can play this role is limited. In order for an ideal to help us plot a path to a better society, the overall goodness of a possible system must bear a reasonable relation to the goodness of those closest (i.e., most similar) to it; random or arbitrary correlations between goodness and closeness (similarity) must be ruled out. Otherwise “there is no point in getting close to the ideal point … but not achieving it: [since] its near neighbors may not be at all just.” 36 We need a conception of a perfectly good society to orient changes we wish to make in actual societies, only if the range of possible states satisfies a daunting “fine-tuning condition.” Gaus puts it this way:

  If the problem of achieving justice is not sufficiently complex [if inherent justice is too closely correlated with similarity to the ideal] … all we need is to make the best pairwise choices we can, and we do not need to identify our long-term goal [making the political ideal superfluous]. If the problem is too complex [if justice isn’t correlated closely enough with similarity to the ideal], the ideal will not help, because any move “working toward” it is essentially a leap in the dark.37

  Even if this fine-tuning condition (allowing small changes to have just enough, but not too much, effect on the overall goodness of a society) is satisfied, there is another constraint limiting the potential utility of a philosophical identification of an ideally good society. Suppose our political philosophy tells us that normative reality is as illustrated by the previous graph. Suppose further that we know our society is S2. Because we live in S2, our knowledge of it is much greater than our knowledge of other systems—which are possibilities arising from imagined changes in goodness-making features our theory recognizes (changes in social and economic institutions). With this in mind, we ask, How much faith can we put in the accuracy of our model?

  It would be nice to think that our evaluations of each of the eleven systems was as accurate as that of our own. But, Gauss argues, we shouldn’t believe they are.

  [O]ur current social world is in the domain [it is one of the systems in our model], and the evidential basis for judgments about the justice of the world we actually live in must be greater than the judgments about merely possible worlds [i.e., about the institutions and practices of societies, and their degrees of justice, that would exist if the descriptions of them in our model were realized]. For all nonexistent [merely possible] social worlds, we must rely for the most part … on predictive models to judge their social realizations; for our current world [our own society] we can employ our best model to understand it, but we also have masses of direct evidence as to its realization. Indeed, our models are often developed from our current data.38

  In addition to knowing much more about our own society, it is reasonable to suppose that our knowledge of those closest to it—which would result from making relatively small changes to our present institutions and practices—is greater than our knowledge of other possible societies less similar to it. Since the imagined route from our society to one of those more distant societies is a series of unexperienced changes, the possibility of error in our representation and evaluation of them will increase as we move further away from our own society and its near neighbors.

  This leads Gaus to his conception of a neighborhood, which he uses to challenge the defender of ideal theories in political philosophy.

  A neighborhood delimits a set of nearby social worlds characterized by similar justice-relevant social structures. In this rough continuum of social worlds some are in the neighborhood of our own social world (and many are not); our understanding of the justice of alternative social worlds in the neighborhood of our own social world is far deeper than outside it.… As we leave our neighborhood the precision and accuracy of our estimations of the justice of social worlds drops off sharply … and the reliability of … [our] models rapidly decreases as we move to increasingly unfamiliar worlds. In contrast, within our neighborhood there may be relatively obvious local optima, about which our judgments are reasonably reliable.39

  The Choice. In cases in which there is a clear optimum within our neighborhood that requires movement away from our understanding of the ideal, we must often choose between relative certain (perhaps large) local improvements in justice and pursuit of a considerably less certain ideal, which would yield optimal justice.40

  The kind of choice Gaus has in mind is illustrated in the figure, in which S5 is our actual system, S4 and S6 are in the neighborhood, S4 is a local optimum, S1, S2, S3, S7 and S8 are outside our neighborhood, and S8 is the imagined ideal system.

  Suppose we are at S5, trying to become better. Our model tells us that moving up from S5 to S4 will do that, while moving down to S6 would make us much worse. Since both are in our neighborhood, we are justifiably confident that these pre
dictions are correct. So we are inclined move to S4, despite the fact that doing so would increase our distance along the x axis to S8, and so take us further away from our imagined ideal, whereas moving to S6 would decrease our distance along the x axis to S8, and so would bring us closer to the ideal. It is rational for us to discount this fact because we know the path to S8 will (at least temporarily) make things worse by imposing potentially severe hardships on real people, without knowing that we will ever achieve the benefits we now attribute to S8. Since we realize that our calculation of those benefits is likely to be in error, we are, and ought to be, reluctant to sacrifice real people for what may turn out to be a theorist’s illusion. Indeed, our very conception of the ideal may be changed by whatever we do in pursuing it.

  Gaus’s attitude toward this choice is expressed in the following passage.

  If the ideal is to be … a long-term goal, the ideal theorist must sometimes—one would think often—stress that we should pursue the ideal and so forgo possibilities to create a more just social world by moving away from the ideal to some near social arrangements. It is critical to stress that this must be the case: if the ideal theorist denies that such choices need ever be made, then … we can do very well without knowing anything about the Mount Everest of justice, and should simply climb the hills that confront us. But if we really do … [pursue the ideal] [t]hose who bear the cost of this pursuit will live in a less just world—their pleas must be discounted.… [F]or us to be under the sway of an ideal theory is for us to ignore relatively clear improvements in justice for the sake of a grander vision for the future. And yet this grand vision is ultimately a mirage, for as we move closer to it, we will see that it was not what we thought.… [F]or those who remember their twentieth-century political history, the position that such theorists have talked themselves into is far too reminiscent of less democratic idealists.41

  The full force of Gaus’s point requires thinking that the only way for a conception of the ideal society to be of use to us requires us to knowingly sacrifice ourselves and our fellows to pursue a possibility that we can’t know to be an improvement over our present state. In cases of this sort, Gaus plausibly argues that we are well advised to moderate our search for the ideal. However, this Gausian lesson doesn’t extend to all cases in which we might use a model of an ideal society to guide social change. On the contrary, it allows for the possibility that sometimes a plausible conception of the ideal would provide useful guidance. There are two kinds of case in which this might happen.

  First consider a scenario in which there are two paths, A and B, for improving our society, the initial changes to each which involve equal improvements achievable within our societal neighborhood. After that, however, path A levels out, with no greater improvements, while path B continues upward to still further improvements. In such a situation, it would be reasonable to choose path B, since doing so would increase the magnitude of our possible gains, without sacrificing anyone’s welfare in doing so.42 Next, consider a scenario in which we could move toward the imagined ideal by making a change that would have little or no negative effect on the inherent goodness of a political system (when compared either with other possible changes or with doing nothing). The fact that our justified confidence in the accuracy of our model of the ideal might not be very high doesn’t prevent us from justifiably taking it to be a guide in this sort of case. Thus, although Gaus’s argument limits the utility of an ideal conception of social goodness, it doesn’t render it irrelevant.

  Still, if he is right, the task of constructing more modest models for making limited improvements in which we can be confident is probably the more pressing of the two theoretical tasks. In his eyes, this means committing ourselves to a conception of society that is diverse,43 polycentric,44 non-optimizing,45 and open.46 He sees complex societies as made up of different moral, political, economic, and religious points of view from which no all-encompassing perspective can be abstracted that is strong enough to yield authoritative resolutions of major issues. But, he maintains, a good, just, and open society doesn’t require this. Collective agreements on moral, social, and political matters are, of course, needed, but foundational Rawlsian verdicts on ideal justice or goodness probably aren’t. What is required is a stable fundamental framework of social rules, a kind of moral constitution, governing interactions between people with differing evaluative perspectives, allowing them to agree on solutions to problems that most find acceptable, and better than no solution at all, even if few, or none, find the solutions optimal.47

  Ironically, Gaus argues, it is precisely by not insisting on optimal solutions that we may better approximate our never to be realized ideal.

  The optimizing stance is also self-defeating.… Our analysis … concluded that an individual perspective on justice will almost surely be unable to find its ideal; being confined to a neighborhood, the identification of its own ideal will be elusive. However … other perspectives can uncover parts of the landscape beyond one’s neighborhood; revealing features of the social world that are not salient on one’s own view, they can help bring one’s own ideal into closer view. But this requires … a network of interconnected communities of inquiry.… It is precisely the framework for such interactions that the moral constitution of the Open Society provides. Adopting the optimizing stance toward the moral constitution [i.e., insisting that others interact with you only on terms that your particular perspective finds optimal] precludes one’s own perspective participating in this framework for inquiry [because insisting on optimality blocks continuing inquiry and agreement].… To wish to learn from other views, while insisting that only one’s own view is correct and that all must live by it, is hardly a basis for a community of shared inquiry. By seeking to optimize in this way one forgoes optimizing in the sense of better understanding one’s own commitments regarding justice.48

  This paean to incrementalism, the open society, the incompleteness of our moral, political, and empirical perspectives, and the need for unrestrained inquiry and constant experimentation is very much in the spirit of Hayek. To a certain extent, even Rawls agreed by prioritizing his commitment to the most expansive basic liberty possible, consistent with equal liberty for all—which was not to be sacrificed to pursuit of greater egalitarianism in the distribution of wealth. Beyond this, however, Rawls retreated to a rigid, aprioristic conception of fairness, to which he was willing to subordinate ordinary normative perspectives on our social lives with others. Like Hayek, Gaus takes this to be a mistake. Respecting the moral and political wisdom that has arisen from the trial and error of historically evolving social institutions, they are not willing to sacrifice it to any purely abstract reasoning. But they also don’t take it to be final. Like ordinary empirical knowledge, they take our moral and political knowledge to be capable of continued advance by successful social and institutional innovation. While humble about our present, they are optimistic about our ability to make moral and political progress in the future.

  When one asks about the contributions made by Hayek, Rawls, and Gaus, not just to the world of political and philosophical theorizing, but to the wider world as we know it, the judgment is mixed. Hayek and Rawls were the most influential political philosophers of the twentieth century. Both valued liberty and understood the efficiency of the free market in producing wealth, while differing substantially in their conception of the role of government. Each has influenced political elites in the most economically advanced societies around the world. Although the work of Gaus is too new for that to have occurred, the empirical, microanalytic direction in political theorizing to which it points is a highly promising development that in time may come to be comparably influential.

  KARL MARX: A CAUTIONARY TALE

  Karl Marx (1818–1883) was, it must be admitted, the most influential political philosopher of all time. His thought played a major role in changing the map of the world, destroying social systems, and bringing about new ones that transformed countless lives and
caused the deaths of tens of millions. Though Marx himself was not personally responsible for the grotesque crimes of despots invoking his name, the philosophical system he created could be interpreted, without gross or obvious misrepresentation, in ways that made the unjustifiable appear if not justified, at least excusable.49 The idea that these events were excusable was an illusion. In order not to fall prey to it again, we must recognize the power of philosophical thought. Never a parlor game to be played in an ivory tower, philosophy, which struggles to understand the basic categories of our existence, knows no boundaries. Because it can affect every aspect of our lives, it should always be rigorously scrutinized. Up to now I have concentrated on the great positive contributions it has made. But the ledger contains debits as well as credits.

  Marx began, and always remained, a German philosopher in the tradition of Kant, Fichte, and Hegel, all of whom were preoccupied with unusual conceptions of human freedom. Freedom was a problem for Kant because he believed both that our thoughts and actions must be free if morality is to make sense and that our decisions and actions, like everything else in time and space, can only be understood by us as being caused, and therefore must be unfree. His attempted resolution of this paradox invoked his distinction between our sensory appearances and the realities, of which they are appearances, about which we can, unfortunately, know nothing.

  This distinction penetrated to our very selves. We are all aware of ourselves acting in time and space along with other items of “external appearance”—e.g., plants, animals, and inanimate objects. Perceiving ourselves in this way, as mere “empirical egos,” we take our thoughts and actions to be caused. However, we, the perceivers, are also the real “transcendental egos” of which the empirical egos are merely appearances. Since our categories of perception, understanding, and knowledge are merely ways of organizing appearances, we can’t know what we truly are, including whether or not we are free. Kant’s solution is to preserve the possibility of morality by postulating that we are free.