ENS, Jaurès building, room Langevin, 24 rue Lhomond, 75005 Paris
Moral and political arguments based on the consequences of actions would seem to be incommensurate with arguments based on protected values (values about actions, not consequences). Indeed, some empirical evidence suggests that the two have quite different psychological properties. Protected values discourse leads to a greater sense of understanding and a reduced willingness to compromise. Accordingly, encouraging a consequentialist perspective by requesting causal explanations can reduce group polarization. I attribute this to the difficulty people have in engaging in causal reasoning about complex issues. Such reasoning is usually done collectively. Other experimental evidence and analyses of large corpora however suggest that consequentialist and protected values reasoning do not reflect separable reasoning mechanisms. Humans tend to reason at a level of coarseness that is not sensitive to the difference between these forms of argument.