JUST AS WE CAN BE MORE OR LESS CERTAIN about empirical matters, e.g., that global warming is caused by humans or that the CIA planned the murder of President John F. Kennedy, we can be more or less certain about normative matters, e.g., that male circumcision is morally wrong or that an action is right if and only if it maximizes happiness. in other words, both normative and nonnormative beliefs can vary in degrees. Recently, it has been argued that this is a challenge for noncognitivism about normativity. Noncognitivists think that normative judgments (primarily) express conative or "desire-like" attitudes rather than beliefs, so the challenge is to account for the fact that normative certitude varies in degrees.
Michael Smith presented the challenge in a 2002 paper and James Lenman (2003) and Michael Ridge (2003, 2007) responded independently. We challenged Lenman's and Ridge's responses in our joint 2009 paper. Andrew Sepielli (forthcoming) has now joined the rescue operation. His basic idea is that noncognitivists should employ the notion of being for (Schroeder 2008) to account for normative certitude. But as we shall see in section 2, the being for account of normative certitude is vulnerable to many problems shared by other noncognitivist theories. Furthermore, Sepielli's favored normalization procedure for degrees of being for has highly problematic implications, as we will show in section 3. We begin in section 1 by explaining the being for account of normative certitude.
1. The Being For Account of Normative Certitude
Sepielli takes noncognitivism to be a psychological theory about normative judgment, and expressivism to be the semantic theory endorsed by most noncognitivists. According to expressivism, the meaning of normative terms is to be understood in terms of the (noncognitive) psychological states they are used to express. A notorious problem for expressivism is to account for the meaning of normative terms when they are embedded in complex sentences. This problem goes by various labels, such as the Frege-Geach problem, the problem of embedding and the negation problem. On Sepielli's view, noncognitivism can account for normative certitude only if expressivism has enough structure to solve the Frege-Geach problem. Sepielli's basic maneuver is to apply Schroeder's recent treatment of the Frege-Geach problem to the problem of normative certitude.
Let us follow Schroeder and focus on negation. Consider the sentence (1) "Jon thinks that murdering is wrong." There are various places where we can insert a negation in this sentence, e.g., the following: (2) "Jon thinks that murdering is not wrong." Now, expressivists face at least two challenges. The first is to explain what kind of noncognitive attitude (2) attributes to Jon. The second is to explain why this attitude is inconsistent with the attitude attributed to Jon in (1). Advocates of traditional expressivism might want to say that (1) attributes to Jon a negative attitude to murdering while (2) attributes to Jon an attitude of toleration of murdering. But then it remains to be explained why a negative attitude to murdering and toleration of murdering are inconsistent attitudes. Since the two attitudes have the same content, the alleged inconsistency between them cannot be explained in terms of their content. According to Schroeder's diagnosis, traditional expressivism has too little structure to meet the second challenge, so his fix is to add more structure. He does so by introducing the attitude of being for. The idea is that to think an action wrong, right, etc., is to take the attitude of being for some other attitude to the action in question. To illustrate, sentence (1) should be understood as (1') "Jon is for blaming for murder" and (2) as (2') "Jon is for not blaming for murder." It is easy to see that the content of the attitude attributed to Jon in (1') is inconsistent with the content of the attitude attributed to Jon in (2'). …