Anthropocentric tradition holds that the (somewhat idealized) human case is the standard for what counts as a real instance of a psychological capacity. But even if we learn about minds from the human case, it does not follow that humans are the standard for objectively correct ascriptions. For the Literalist, the true extensions of the predicates is only now becoming clear; to put this in Sellarsian terms, the transition from the Original Image to the Manifest Image went too far in withdrawing “person-appropriate” predicates from everything except humans. We already accept this de-anthropocentrized perspective for perceptual capacities. For example, we understand what vision is from our own case, but non-humans also possess real, full-blooded vision however different they may be from us. The Literalist holds that this perspective is being extended to cognitive (“higher”) capacities. The fading line between perception and cognition backs her up.
Literalism does not claim that all the psychological properties we possess are possessed by every other biological entity. But the phrase “the psychological properties we possess” bears scrutiny to clarify Literalism by showing exactly what it rules out.
(1) The psychological properties we possess are properties that are exclusively human. This is compatible with Literalism if it means no more than saying that (e.g.) the human visual system is exclusively human in the same way the human hand is exclusively human. Human-typical psychological properties reflect our “species-specific window of viability” (in Andy Clark’s terms) in relation to the world. The mantis shrimp’s visual system is exclusively mantis-shrimpy, too.
(2) The psychological properties we possess are the psychological properties that there are (in the actual world). This is compatible with Literalism. Maybe this is all bad science and that as a matter of empirical fact humans are the only possessors of psychological properties and the new uses are false. Literalism does not insist that statements incorporating the new uses are true, only that they are best interpreted as literal with sameness of reference across human and nonhuman domains. They are truth-apt, not factually defective.
(3) The psychological properties we possess are the psychological properties we phenomenologically take ourselves to have. That we have a specific perspective on our psychological properties is also compatible with Literalism. It does not follow that what we grasp phenomenologically is correct. Once upon a time, around 600 bc, water did not have a hidden essence. Some—Thales, at least—thought it was the hidden essence of everything else. Much later, our concept of water—that stuff—adjusted to scientific theory and discovery in a way that clarified our original referential intentions. Our psychological concepts are open to similar adjustment. Starting from our phenomenological appearances does not entail taking them as the last word about their reference. Mathematical models provide a basis for understanding similarity of states that until recently we could only grasp phenomenologically from the first person.
(4) The psychological properties we possess are the psychological properties we phenomenologically take them to be and this is what they really are. This claim is incompatible with Literalism, since it entails that we already know what psychological properties really are and that this knowledge is based entirely on observation of ourselves. This claim constrains the proper extension of psychological predicates to the features of human embodiment that we are familiar with or just learning about in our own case. It claims that we do not just understand mental states in our own case; we understand mental states from our own case—where “our” case is the human case, not just me. (Note how we ignore the problem of other minds when it suits us.) That is the upshot of the implicit domain restriction of “others” in the Modified Argument in blog post #2. Literalism claims this implicit domain restriction is scientifically unjustified. Anyone who thinks the mind falls within the purview of science – unlike, say, Descartes – should find (4) epistemically unacceptable. Even if in the past we were unable to work around our parochial human perspective on cognition, we can now.
Literalism does not claim that every psychological predicate will end up with a scientifically determined reference, just as orichalcum did not get a square in the periodic table. The psychologically sophisticated Scientific Image can embrace some concepts and ignore others. Nor does it deny that human psychological capacities are special in some way. But what makes human psychological capacities special need not be why they are full-blooded psychological capacities. Finally, Literalism does not impose a priori restrictions on where psychological language may be properly used. The current restriction to biology is empirical, not conceptual or logical, and may well be temporary given significant advances in artificial intelligence.
Hi Carrie, I’ve been loving this series. I’m very sympathetic to Literalism; I feel like it may one of those views that, frustratingly, lots of people want to hand-wavily reject in passing, but few people actually have arguments against.
I was wondering, though, if I could ask you to elaborate on something you’ve alluded to a few times in these posts, namely the relation between 1) extending psychological predicates to non-human organisms, and 2) extending psychological predicates to smaller parts of human beings. Do these two sorts of extension raise basically the same issues? In particular, what do you think of the idea that everyday psychological predicates are ‘maximal’, so that an entity is disqualified from a predicate applying to it by being part of a larger system that the predicate applies to?
Hi Luke,
Glad to hear you like it, and yes, I think you have put your finger on the biggest barrier. Anthony Trewavas, who championed plant intelligence decades ago, remarked at some point that the biggest barrier he faced was just the very idea that plants behaved.
I do disagree with the “maximal” constraint, as evidenced in my previous reply (the remarks where I mentioned Elliot Sober’s work). Many people seem to take for granted that constraints on whole-part relations for objects carry over for activities, and this is obviously not so (the rotating example). Also obviously this means on my view something is deeply mistaken about homuncular functionalist explanation of mind, with the demand for ever “stupider” capacities in functional decompositions until they are “discharged”. That’s also a chapter in the book and I’m elaborating it in future work as well. I don’t mean to be cryptic, but once one sees that compositional relations differ between objects and activities (or events or doings …) the impulse is to search for a psychological exception to this rule (“OK for rotating, but surely not for deciding!”), and that back-and-forth can go on for quite a while!
Thanks Carrie, and sorry for missing your latest comments on the previous post. Maybe I can push you from the opposite side to Jim and Tony – why should we think that there are mereological constraints on *objects*, of the sort that you think there aren’t for *activities*?
I’m not sure there are constraints for objects either. It does seem to be the case that you can’t make a flower out of little flowers, or a cell out of little cells etc. etc. but of course you can (or must) make a physical object out of little physical objects. So whatever the case may be it is relative to some classification scheme more specific than “physical object’. I don’t have much to say on this whole issue (yet) other than to note that there’s no reason to think any constraints on objects are the same as those on activities.
“you can’t make a flower out of little flowers, or a cell out of little cells etc. etc.”
Well, there is the endosymbiotic theory of eukaryotic cells, so it looks like you can make cells out of little cells, and depending on definition flower-clusters like hydrangeas have might count as flowers made out of little flowers (if ‘flower’ is defined more anatomically, you might need to find some weird mutants, obviously).
But I take your point that this isn’t something you’re committed to defending. Just wanted to speak up for those of us who think there are no substantive mereological restrictions even on objects!