IV. On Soft Normativity

We wish to refer back to the notion of ‘soft’ normativity we introduced above, and suggest that it is an accurate corollary to Searle’s (and Hart’s and Rawls’) views that they characterize the normativity of social institutions assoft - of a piece with the normativity we find in games.[^33] The constitutive-rules-based ‘oughts’ of games are, however, defeasible to a very high degree. Certainly when playing baseball oneought to go to first base after four bad pitches are thrown, but no oneought (in any interesting sense) to play baseball in the first place: any baseball player can walk off the field, can abandon the game, any time he wishes - though of course if a professional baseball player were to do this, he would probably lose his job.

This last remark reminds us that there are other types of oughts in games, in addition to those based in constitutive rules. For example: one ought to remain in the baseball field even after humiliating oneself by missing an easy catch. Players in a game of basketball can ‘foul’ their opponents several times in order to prevent them from scoring, but they ought not to stab or shoot their opponents. One might try to explain the latter sorts of normativity by appealing to the fact that, for example, by embarking on a game of baseball one has in a sense promised not to leave the playing field after making silly mistakes, or that all human beings have in a sense promised not to kill in general and this promise covers also one’s behavior when playing basketball. This strategy, however, robs terms like ‘promise’ and ‘contract’ of their customary meanings. Moreover, at least some of the mentioned obligations seem not to be obligations of the osrt which one could acquire by means of promises or contracts.

Legal and sociopolitical institutions, similarly, give rise to obligations not only of the constitutive-rule-based sort but also of other sorts. According to Hart, for example, Nazi laws are genuine laws in the constitutive-rule-based sense - but they are at the same time laws that oneshould not follow . Famously, Hart charged that Gustav Radbruch’s abandonment of positivism in the post-Nazi era was the result of his “half-digested” understanding of “the spiritual message of liberalism”,[^34] whereby Radbruch had failed to see that even the staunchest positivists share the “conviction that if laws reached a certain degree of iniquity then there would be a plain moral obligation to resist them and withhold obedience”.[^35] Presumably, Hart would agree that this “plain moral obligation” is not a game-related obligation. Significantly however he does not discuss what type of obligation it might in fact be, and this is the sort of discussion that Searle avoids as well.

To see that something is wrong with the identification of all normativity with the normativity of games, we can appeal to Wittgenstein’s remarks on the nature of games in the context of his treatment of the notion of family resemblance in thePhilosophical Investigations .[^36] According to Wittgenstein, no definition formulated in terms of necessary and sufficient conditions can apply to all games. In light of Searle’s views on what we have called soft-normativity, however, it is tempting to suggest thatbeing created by a set of constitutive rules would amount, precisely, to the sought-for definition. Whenever you are in the presence of an entity which exists in virtue of constitutive rules, you areeo ipso in the presence of a game, andvice versa .

This resolution of Wittgenstein’s puzzle comes at a price, however - for it forces an over-large scope upon the notion of game, which now turns out to include socio-political institutions like promising, punishment, marriage, and government. Note that if Searle and our authors are correct, then this would in no way count against it. For whenever Searle, Hart, and Rawls wish to explain the normativity of such institutions they do indeed invariably end up talking about the way in which swinging at the third strike entails that you ought to leave the baseball field. This move, if we are right, is not a matter of happenstance. Rather, it reveals that Searle and our other authors have maneuvered themselves into a position where they do not have the tools to draw the distinction between games and socio-political institutions.

Part of the compelling force of the “why should I play this game anyway?” objection to the thesis that all normativity is soft normativity turns on theconventional character of games. For even if there existed something like a game of life,[^37] the skeptic could still ask a reformulated question: “Why should we not alter its rules?” Constitutive rules, are after all not merely to a high degree defeasible, they are also easy to change: at some point in their history virtually all games had rules different from those they have today.

Hospitals are, by definition, places where physicians and nurses ought to care for patients. If there were a hospital in which nurses and physicians systematically harmed their patients, then we would not be content simply to claim that this institution is, by definition, no longer a hospital and leaving it at that. Obviously, we would claim that the physicians and nursesought to care for their patients, and that this obligation is not merely the result of the constitutive rules governing hospitals and medical professions.

Other sorts of normative claims: that murder is wrong, or that it is appropriate for wrongdoers to apologize, that purely accidental (non-negligent) wrongdoing is not blameworthy, etc., are not only not easily defeated, they are also - and even more conspicuously - noteasily changed . Whereas the number of fouls a basketball player can ‘legally’ make in the course of a game can at any time be changed, the prohibition against stabbing his opponents is not likely to change at all.