A post by Neil Van Leeuwen

Philosophers who wish to understand what imagination is often feel tempted to pose the following sorts of question:

1.     Is X really imagination? [for various candidate mental states and/or processes X]

2.     Does imagination require Y? [for various mental phenomena Y]

For the X place in 1, we get candidates like the following: episodic memory, beliefs that incorporate mental imagery, beliefs concerning counterfactual scenarios, suppositional thought that lacks mental imagery, florid delusions that don’t generate behavior like typical beliefs do (and hence might be imaginings), dreams, etc. The idea seems to be that if we can ‘correctly’ categorize such phenomena and ‘borderline cases’ in terms of whether they ‘really are’ imagination, then we’ll be closer to having a handle on what imagination fundamentally is: that is, if we work out the extension of “imagination” all the way to the corners—so the thinking goes—we’ll be much closer to being able to figure out what it is that unites that extension . . . which will give us what imagination is!

For the Y place in 2, we get candidates like the following: mental imagery, a distinct cognitive attitude (distinct from belief), counterfactual contents, fictional contents, conscious awareness, simulation, ongoing active processing, etc., all of which have been floated as candidate phenomena for being requirements or necessary conditions on imagination. The approach is to find all the Y for which you can say the following: If a mental state or process doesn’t involve Y, it can’t be a state or process of imagining. The hope is that, once we figure out all the Y. . . that will give us what imagination is!

And since we all want to know what imagination is, we fall into posing questions like 1 and 2.

This is speculative, but I think the tacit model people have in mind in the background is something like the natural kind concept GOLD and the corresponding word “gold” (and its semantics). As mid-20th century philosophers like Putnam and Kripke taught us, words like “gold” rigidly designate and have always rigidly designated a certain chemical substance, which we now identify on the periodic table as element 79 (Au), and they did so even before it was known what exactly that substance was. The semantics of “gold” presupposed that there was, as it were, an essence of all things gold, and it just remained to be discovered (eureka!) what that essence was (where “essence” can be minimally understood here as causally efficacious structure that is shared by instances of a natural kind and largely explains their surface features). As we now know, that essence of gold eventually was discovered (hooray!).

If “imagination” were like “gold” in terms of its semantics, questions of the forms 1 and 2 would make perfect sense. Question 1 would do the work of separating off the spurious apparent instances of the natural kind so that we can be working with only true instances of the kind (imagination) we are trying to understand (just like fool’s gold must be set aside if we are trying to find the essence of gold). Question 2 would do the work of getting at just what the underlying essence of the imagination natural kind is, so that we could give it a crisp explication in more exact terms (just as happened once chemists worked out the chemical structure of the element gold).

But what if “imagination” in its semantics is more like “air” than “gold”? That is, what if it’s not a crisp clean natural kind term that rigidly designates a single discrete essence/causal structure, but rather has as its extension a cluster of phenomena that, though certain features are often present, doesn’t have a discreet, cohesive essence that is shared by all instances in the extension of the term?

It doesn’t take much work to see that if that’s what “imagination” is like, questions of the form 1 and 2 are misguided. Any air, of course, does have to be gas. But aside from that, the semantic requirements—if that word even make sense—are extremely loose. The air we usually breathe is about 78% nitrogen. But it would be reasonable enough to refer to what comes out of an oxygen tank as “air,” even though it has no nitrogen at all. What about a room filled with 99.8% nitrogen, 0.199% carbon dioxide, and 0.001% oxygen? Would that really be air? That question lacks a point in the way its analogue had a point for “gold.” So since the “air” version of question 1 is misguided, if “imagination” is like “air” in its semantics, 1 itself is misguided. Related considerations would easily show that questions of form 2 would also be misguided.

As you’ve probably worked out by now, I do in fact think “imagination” is more like “air” than “gold.” What this blog adds to what I’ve said before is a clearer statement of why that implies that questions of the forms 1 and 2 are misguided and—to make matters worse—counterproductive insofar as they tacitly reinforce the wrong (“gold”) model of inquiry into imagination.

At this point I suspect there may be some anxiety welling up in some people reading this, which may fuel some unneeded resistance to what I’m saying. The driving thought behind the anxiety might be something like this: if “imagination” isn’t like “gold,” then there’s no such natural kind as imagination—it’s just the junkyard of the mind—in which case we have no legitimate object of scientific study. We’d be little better than eliminativists! We’d be stuck researching the manifest image and not the scientific image! The whole world will look down on the metaphysical mediocrity of what we’ve dedicated our lives to studying! Heavens! And so on . . .

Yet we imagineers needn’t start clutching our pearls just yet.

And that’s because the “air” analogy contains a positive lesson as well, one that should give us comfort, even as it demands we revise the kinds of questions we ask.

Even if air itself is not a natural kind in the way that gold is, there are still natural kinds that compose air that are just as legitimate as any natural kind ever was: oxygen, nitrogen, carbon dioxide, etc. It just that those natural kinds needed to have terms introduced as they were discovered. It follows that the thinking behind the metaphysical anxiety described above is confused: from the fact that “imagination” is not a natural kind term, it does not follow that there are no natural kinds of which instances of imagination are composed and which may be legitimate objects of both philosophical and scientific study. Analogously, from the fact that “air” is not like “gold,” it does not follow that those who study air do not have a legitimate object of scientific study—obviously!

The questions just need to be recast: instead of looking for a natural kind that imagination is, look for the natural kinds of which its instances (in its somewhat nebulous extension) are composed. And then we can perhaps ask questions of the forms 1 and 2 with respect to our newly minted kind terms. Imagination on this picture would not be a junkyard, but a collection of elements and compositions thereof, each worth studying in terms of its own dynamics and the dynamics of how it composes.

Modified rapture—as Fodor would say—imagination is not like gold, but like air.