It is widely held that a causal view of mental content has to be externalistic. Jerry Fodor has had a major role in promoting this idea, which unfortunately caused him much trouble, since his Representational and Computational Theory of Mind (RCTM) needs a content which is determined narrowly. Here I defend the thesis that in fact, if you want a causal view of mental content, the kind of externalism both required and possible is not the sort proposed by Putnam (1975). In other words, RCTM can and must do without the notion of wide content, because i) narrow content can be interpreted externalistically (i.e. causally), and ii) wide content is not really consistent with a causal view, at least if we want (as Fodor does) a nomological notion of causation.