The quick and the dirty:
Representational contents are causally efficacious only if representational contents are properties. But not all representational contents are properties. So not all representational contents are causally efficacious.
The slow and the clean:
The bearers of causal efficacy are properties. Something causes something else in virtue of some of its properties and not others. The color of an object, say the redness of a Frisbee, may be inefficacious with respect to the bump on my head. But this does not mean it is inefficacious tout court, since the redness may have the power to make a bull charge.
We may not have to dig very deeply into the notion of content to find some serious challenges to claims of its efficacy. This suggestion is fleshed out further as follows.
The bearers of efficacy are properties. But not all contents are properties.
The key language of the discussion of efficacy involves phrases such as â€œcausal relevanceâ€ and â€œin virtue ofâ€ all of which implicate properties as the bearers of causal efficacy. Consider a red Frisbee. It is in virtue of its mass and velocity that it may cause a bump on my head, and in virtue of its redness that it may cause a bull to charge. If mental events are causes, and causes in virtue of their contents, then contents will need to be properties.
However, more will need to be said about contents to justify assimilating them to property-talk. Consider the content of my belief that not all beer is carbonated. What is the content of this attitude? Plausibly, the proposition that not all beer is carbonated. But it is not clear that this is a property. The content of attitudes is supposed to reduce to the meanings of component representations, ala The Language of Thought Hypothesis. Plausibly, the components here would include a predicational representation x is carbonated, the meaning of which is, I suppose the property of being carbonated. So, in at least one instance, weâ€™ve found a property among the contents, and further it is an efficacious property, but it is not clear that we will be able to do this in every case. Some representations will not have properties as their meanings, as in the cases of the mental analogues of quantifiers and singular terms. And some representations will have inefficacious properties as their meanings, as in the cases of the of the predicate x is outside of my light cone or, better yet, x is an inefficacious property.
Does the inefficacy of content violate Folk Psychology?
Let us grant that Folk Psychology acknowledges things with content and that these things figure into the causes of behavior and or bodily motions. But it is not entirely clear that this alone commits Folk Psychology to the efficacy of content. Consider the following attitudes: I believe that there will be a bake sale tomorrow and that I desire to go to tomorrowâ€™s bake sale. The content of both of these attitudes is that there will be a bake sale tomorrow. For reasons given above, it is unclear that this content has effects on me. Additionally, we may wonder how anything happening tomorrow can affect something happening today. Further, note that there is a lot more to the Folk Psych story than the contents, the other half of the story contains the attitudes themselves. The bake sale may be tomorrow, but I have the desire today, and it is the desire itself that drives my behavior and whatnot. So, it is not clear that the sorts of considerations from Folk Psych, like, â€œMandik opened the fridge â€˜cuz he thought beer was nearâ€ are evidence that Folk Psych contains an intuition that contents are efficacious. It seems merely to indicate that attitudes are causes.