“As Aristotle consistently says, the best generalizations about how one should behave hold only for the most part.  If one attempted to reduce one’s conception of what virtue requires to a set of rules, then, however subtle and thoughtful one was in drawing up the code, cases would inevitably turn up in which a mechanical application of the rules would strike one as wrong— and not necessarily because one had changed one’s mind; rather, one’s mind on the matter was not susceptible of capture in any universal formula”  -John McDowell (1)

The above viewpoint articulated by McDowell is also called by him “non-codifiability.”  That is, knowledge of ethical reasoning is non-discursive, it is irreducible to rules, precepts or other, linguistic or not, ways of conceptualization.

However, someone may object that such a view cannot be sustained, that in fact when ethical generalizations are made correctly, they are exemplars of codifiability.  In order to see why this is so, let us distinguish between two kinds of moral generalizations.

1. Simple Generalization:
   In situation X, do Y.

1. Sophisticated Generalization:
   In situation X, do Y most of the time.

It would be granted, I think, that moral reasoning involving type A would be problematic, for the reason that McDowell, merely echoing Aristotle, points out above.  There would be too many exceptions to this kind of rule to be productively reliable.  Furthermore, perhaps, such indeterminate applicability even undermines its status as a rule. 

Thus the Sophisticated Generalization is an improved version of the Simple Generalization, for it accommodates the “what if” scenarios implied in the Simple Generalization.  However, the Sophisticated Generalization, to return to the original objection, seems (problematically for the non-codifiabilist) to both explain moral reasoning and articulate it in a codifiable way.

However, let us see if the Sophisticated Generalization is actual codifiable.  Any statement allowing for variation or accommodation of an exception such as  “In situation X, do Y most of the time,” is really another way of saying that, “In situation X, do Y, except in case X1 do Y1, except in case X2 do Y2, etc.”  If this is the case though, this shows that the Sophisticated Generalization is not a general rule, but a set of particular rules collected into a dictum.  And if this collective of rules cannot allow for the nuance necessary in moral reasoning, for it will be hard to see at which point the exceptions will cease, then the Sophisticated Generalization falls prey to the same fault as the Simple Generalization.  Both are unable to parallel exhaustively, via a set of codes, the complexity or adaptive variation one encounters in day to day moral reasoning.    

(1) John McDowell, pg. 58, Virtue and Reason, in “Mind, Value and Reality”