Although an increased attention has been given to the doctrine of the association of ideas as being sufficient to account for most of the operations of mind, yet its nature has been looked upon as too simple and philosophical to require much scrutiny; whereas, that very power of association appears to me the most difficult of comprehension in nature; for how shall any given idea be supposed as associated with some other idea, which idea is not yet supposed to be in existence; one idea only present in the mind, a single simple perception, merely, cannot suggest an after perception, for the suggestion is the perception of the suggested idea itself.
The association of ideas can truly therefore, be nothing more than a compound idea ; than one being of thought, a conception of different qualities in unison. As a state of mind, as a given sensation, it must be immediately united with the action and with the state of the material organs which excite it, and coalesce therefore as one with it: thence merely forming one being, one given state of being.
Lady Mary Shepherd, Essay on the Association of Ideas, and the Interaction of Mind and Body. This is a very interesting sort of argument against Hume-style uses of association. The basic idea is this. Associationism requires us to say that one idea suggests another (according to resemblance, contiguity, or causation, for instance); but if we take association as a primitive (as Hume does) this is utterly mysterious, because it means that an idea that does not exist is called into existence by an idea that does, despite the fact that any number of alternative ideas could be pulled up. Seeing someone who looks like your best friend in third grade may make you think of your best friend in third grade; but only because your mind has already made the required association (that person looks like my best friend in third grade). So the movement from the one to the other can't be association itself; it presupposes it.
Shepherd's own gesture at how to understand associations in the imagination, the conclusion in the second paragraph quoted, is obscurely expressed, but if I understand her properly, she thinks that we should be explaining association of ideas in terms of more fundamental causes -- two more fundamental causes, in fact, the mind, which is capable of directing its ideas this way and that, and the material organs that make up the body. But more than that, her suggestion is that association of ideas is nothing other than the mind directing its ideas this way while the material organs excite the mind that way, and that therefore the sort of cognition involved in association is nothing other than mind and body joined as a single thing at a given moment under certain conditions.
