Full Idea
If something has beliefs only if something else is disposed to "treat it" (i.e. think of it) as though it does, then we seem at least to have an infinite regress of appeals to believers.
Clarification
An 'infinite regress' is an explanation that never ends, because it always requires a further explanation.
Gist of Idea
If mind is just an explanation, the explainer must have beliefs
Source
comment on Daniel C. Dennett (works [1985]) by Georges Rey - Contemporary Philosophy of Mind 3.2.1
Book Reference
Rey,Georges: 'Contemporary Philosophy of Mind' [Blackwell 1997], p.79
A Reaction
This sounds like a serious difficulty for behaviourists, but is not insurmountable. We need a community of interlocking behaviours, with a particular pattern of behaviour being labelled (for instrumental convenience) as 'beliefs'.