Saturday, January 16, 2010

Species and Cognitive Enhancement

Here's a fun puzzle case that's been (independently) suggested to me by a couple of different people recently. Consider Moe, a relatively intelligent monkey, and his psychological counterpart Hugh, a severely retarded human. Let's stipulate that, despite their differing species, they are psychological duplicates: they have all the same experiences, thoughts, and feelings. This is at least conceivable. Now suppose we can cognitively enhance just one of them, to turn them into a rational person. Do we have any (non-instrumental) reason to prefer enhancing one of them over the other? Would one of them benefit more, for example, or have a stronger moral 'claim' to such treatment?

It's very intuitive to think that we'd do better to help the human here. (And not just because of rank speciesism: I trust that the intuition would carry over even if Hugh were a retarded instance of an intelligent alien species.) Moe has [teleologically] 'normal' intelligence for the kind of being he is, whereas Hugh is disabled. If we take these teleological, kind-based categorizations seriously, then we're apt to think that what would be 'enhancement' for Moe is treatment for Hugh, and treatment takes priority over enhancement. But should we take those categorizations seriously?

There's something bizarre about the claim that two individuals with qualitatively identical psychologies could differ in their cognitive (dis)ability status, in any sense that really matters. Sure, we could evaluate individuals on a species-relative scale. But isn't there something unacceptable arbitrary about that? Isn't it more principled to evaluate them in their own right, looking only at their individual qualities?

Charles Johnson responds that we "have to distinguish between the cases where a faculty is present but unexercised, damaged, frustrated, undeveloped, etc. and those in which there isn't any faculty to lament the damaging of at all." In other words, perhaps there's an intrinsic difference in the two minds after all. On this proposal, Hugh has a damaged capacity for rationality, whereas Moe has no such capacity at all.

The main problem with this proposal is that it appears baseless. In general, the distinction between present-but-dormant dispositions and absent-but-acquirable dispositions strikes me as philosophically murky. Maybe in this case we could make clear sense of it if there was an appropriate neurobiological difference between the individuals: say if Hugh's brain (but not Moe's) contained all the "wiring" necessary for advanced cognition, and all we needed to do was restore a broken "connection" of some sort. But it seems we could stipulate away any such blatant physical difference: just consider a case where Hugh's brain is actually as small and limited as Moe's, would be just as difficult to improve, etc. Maybe you could try to look further back, say to the level of DNA, and argue that Hugh's DNA shows greater potential for intelligence than Moe's does. I guess that's the most promising option for the teleologist here, but we're now appealing to facts so distant from the actually manifest individuals that it's hard to credit these subtle differences with any great normative significance. (It seems an awful stretch to say that it's a significant difference between the two minds that one of them arose from DNA that could once have been more easily tweaked to produce a rational being in place of the creature that exists today. That doesn't sound much like a present-but-dormant capacity to me.)

The final option for the teleologist here is to turn to speculative metaphysics. Whatever qualitative, physical differences we may or may not find between how these two beings happen to be, it's a deep fact that they differ in what (fundamental kind of being) they are. [My evil twin explains this distinction in more detail here.] Hugh is essentially a human being - thus rational in kind - however severely underdeveloped he might happen to be in his actual manifestation. Moe, by contrast, is simply a non-rational creature, rather than a defective manifestation of a rational creature. This divergence is reflected in our intuitive judgments that cognitive enhancement would merely constitute development (change within his kind) for Hugh, whereas to enhance Moe would transform him into an entirely different kind of creature.

It's worth noting that anyone who accepts this sort of framework should probably also consider embryonic death to be a significant intrinsic bad, for the reasons given by my evil twin. So that's pretty counterintuitive. And I personally find the speculative metaphysics a bit much to swallow. Still, it does seem to explain our cognitive enhancement intuitions pretty well. If on reflection we reject the speculative metaphysics (as I think we should), then we can use it to give an 'error theory' of our initial intuition in favour of Hugh. Really the stark individualist picture is correct, and we should be indifferent between the two options. But we're tempted to favour Hugh because our intuitions are driven by this metaphysical mistake, which leads us to see Hugh as "deprived" in some way that Moe is not.

What do you think?

25 comments:

  1. I share your initial intuition that one ought to prefer enhancing Hugh, but I think my intuition is explained by a much simpler worry than the ones you consider; namely, I expect enhancement to make Moe, but not Hugh, terribly unhappy. Moe, being so different from both his own species and the only species at his same cognitive level, would likely feel like an outcast (or a mere spectacle) for his entire life. Hugh would not likely suffer the same consequences.

    If I bracket this worry about their respective fit in their social worlds (by imagining either that each is in permanent isolation, or that the enhancement is given to a very large group of monkeys, or that Moe is transformed completely into a human), then my initial intuition vanishes, and your arguments against favoring Hugh seem decisive. But worries about social fit typically would be relevant.

    With this said, I think your error theory is more likely to account for why most other people would favor Hugh. We're probably hardwired to lean toward biological essentialism; if I no longer find it intuitive to think that way, it's probably from having read so much bioethics that I have ceased to be human!

    ReplyDelete
  2. Mark - good point. It's helpful to explicitly specify that we need to bracket such extrinsic considerations about differing outcomes, and suppose (unrealistically) that either being would go on to live an equally happy life.

    ReplyDelete
  3. Hi Richard. This is a usefully provocative piece for me, since I am - I suppose - a teleologist. But after some thought, I'm not much worried by the Hugh and Moe case, because I think the way you set things up begs the question. You consider Hugh and Moe at a single moment; but I’m interested in whole lives with developmental arcs, not just in isolated moments within those lives. The difference between Hugh and Moe (where there is a difference) is to do with how they got to where they are, where they could go next, and how their lives are shaped as wholes. Distinguish, for instance, two whole-life versions of Hugh: (1) Hugh was cognitively identical to me until he suffered catastrophic brain injury at the age of 25; (2) Hugh has Down syndrome (stipulate that Hugh(1) and Hugh(2) are momentarily identical). These cases are different, and should be treated differently; but looking only at the moment, instead of the whole life, can’t catch the difference. Hugh(1) has been injured, and his life is going badly as a result; he should be helped if we can. Hugh(2) has not been injured, and there is nothing wrong with him – he just has (e.g.) a lower IQ than I do (although, of course, his society may well make his life variously constrained and unhappy; the social context of Hugh’s life should be changed if we can).

    ReplyDelete
  4. Hi Sam, I absolutely agree that 'developmental arcs' and past differences can make a difference (though not in the case I had in mind). I meant to set things up so that Moe and Hugh are not just momentarily identical, but have relevantly similar pasts, and relevantly similar future potential too. We're meant to be holding fixed all else except their species-kind. The aim is to see whether species alone can make a difference. The 'teleologist' (as I'm using the term here) thinks so: Hugh, but not Moe, is disabled (deprived of some teleologically 'normal' functioning that he ought to possess). Hugh is thus badly off, according to the teleologist, in a way that Moe is not. Do you share this view?

    I'm puzzled by your last sentence -- you seem to be suggesting that cognitive disabilities aren't really disabilities after all ("nothing wrong"). Maybe you mean to suggest an unusually fine-grained teleological view, whereby one's "normal functioning" is given by one's individual (perhaps past) characteristics, rather than one's species-kind. Is that it? You might then agree with me that despite initial intuitions in the Hugh-Moe case, species isn't really of any intrinsic moral significance.

    ReplyDelete
  5. I took it that your strategy against the claim that, just on species-membership, we have no reason to enhance Hugh rather than Moe, was to ask what could possibly differentiate between them. But the options you offer the teleologist in answer to that question are all claims about Hugh and Moe at a time, not claims about their lives as wholes. That’s what I think is problematic (I withdraw ‘question-begging’ – that’s too strong): I don’t see why I’m only allowed to differentiate between Hugh and Moe in that way; and in order to decide which (if either) we have reason to enhance, I need to know more about their whole lives. I don’t think you can coherently stipulate that Hugh and Moe are developmentally identical in all the relevant respects, since if they were, it couldn’t be the case that one of them is human and the other a monkey. Being a human or a monkey is partly constituted by having a particular developmental history and set of developmental dispositions.

    So, I don’t have an answer to your question whether Hugh is badly off in a way Moe isn’t, because the question’s ill-formed as it stands. I do want to say that their developmental histories – including the histories which make them the kinds of creature they are – are relevant to how well their lives are going, in such a way that, of two momentarily-identical creatures, one’s life could be going well and the other’s badly. I don’t want to say that their species is the only thing relevant to how well their lives are going – does anyone really say that?

    (On my last sentence and your – reasonable – puzzlement, I suspect I do have an unusually fine-grained teleological account, but perhaps that’s a point to put to one side for the moment?)

    ReplyDelete
  6. Well, can't you at least co-operate by imagining a case where Hugh and Moe are as similar as possible? So, no "catastrophic brain injury" introduced at the last minute. We may instead imagine some kind of congenital defect -- more serious than Down Syndrome (DS folks are still far more intelligent than monkeys, I trust!), but something along those lines.

    Anyway, if you could spell out a scenario that gets as close as you deem possible to keeping "all else equal", and you think that even then we will necessarily have some reason to favour Hugh, then please spell out this reason for me, and maybe we can make some progress.

    ReplyDelete
  7. Thanks for the clarification, Richard. I thought you might want to exclude the kind of factor I appealed to, but I wasn't sure.

    I've been thinking about whether there would be a way to ground the standard intuition without appealing to latent dispositions or species essentialism, and one thought that came to mind was that one might appeal to the idea of something having gone wrong deep in the history of Hugh's development, whereas nothing has gone wrong in the history of Moe's development. When I say "deep," I mean something as early as a germline mutation in one of Hugh's parents; I do this to try to make the situation as far removed as possible from the kinds of occurrences Sam C invokes, while acknowleding that there must be some cause for why Hugh turns out to be severely retarded (I think this is as "all else equal" as we can get without postulating that Moe and Hugh have come into existence out of nowhere, which I suspect would spoil the question you're interested in).

    The advantage of casting things this way is that it does not require one to hold that Hugh has any kind of latent capacities, or that he is supposed to be more cognitively advanced by virtue of "being human." On this picture, one could even deny that Hugh is (essentially, not biologically) human at all. One simply would argue that, whatever he is (in essence), his cognitive abilities ought to be enhanced, because their current level is the result of a broken process, and (ceteris paribus) broken things ought to be fixed.

    The most serious worry that I have about this is the obvious one: is there is any way to specify which kinds of mutations constitute "breaking" and which ones would not, without appealing to teleology? We wouldn't want to classify all mutations as things that ought to be fixed, because that would mean that virtually the entire history of life needs correcting. But, we also don't want to sneak essentialism in through the back door. Perhaps we can resolve the issue by invoking welfare: maybe, if a mutation causes the resultant being to be worse off than it would have been otherwise, then we ought to count the mutation as defective. This would make our assessment of whether or not Hugh ought to be treated contingent upon his surroundings, but it always would knock Moe out of the running, since Moe's cognitive abilities are not at all the result of mutation. What do you think?

    ReplyDelete
  8. Mark - that's an interesting idea, and a definite improvement over my imagined proposal that "Hugh's DNA shows greater potential for intelligence than Moe's does." But I'm skeptical of the general claim that "broken things ought to be fixed" (with greater priority than a welfare-equivalent 'enhancement').

    For example, imagine that Sue acquires beneficial mutation X. Then a germ-line 'bad' mutation causes Sue's daughter Sally to be conceived without gene X. Suppose we can offer gene therapy to give more people this rare but beneficial gene. Do we have any more reason to give this benefit to Sally in particular, rather than some random Joe? I would think not. (Perhaps some would disagree, but surely not so many as would be inclined to favour Hugh over Moe.)

    ReplyDelete
  9. Richard, you don’t seem to have got hold of either (1) my objection to the OP, or (2) my position. Doubtless my fault for being less than clear, so I'll try again.

    (1) The important passage in my earlier comment was this:

    'I don’t think you can coherently stipulate that Hugh and Moe are developmentally identical in all the relevant respects, since if they were, it couldn’t be the case that one of them is human and the other a monkey. Being a human or a monkey is partly constituted by having a particular developmental history and set of developmental dispositions.'

    Your insistence on keeping ‘all else equal’ is precisely what I’m saying is the problem. There is no conceivable case where Hugh and Moe are similar in all relevant ways, but Hugh is a human and Moe is (say) a golden-bellied capuchin. Being a golden-bellied capuchin is necessarily a matter of having a particular developmental history, including for instance a particular history of protein-building in the womb, fetal development, etc. Note that I’m not relying on any dubious essentialist metaphysics of species: a species is a breeding population with a particular history; an individual’s membership of a species is a historical fact.

    (2) Your demand that I explain why ‘even then we will necessarily have some reason to favour Hugh’ misses my point. I don’t think we necessarily have a reason to favour Hugh (and surely I’ve already said that I don’t think that). As I said in my first comment, I think your example deliberately leaves out the details – details about the whole developmental history of Hugh and Moe – which we’d have to know if we’re to find out which, if either, we have reason to favour. That developmental history includes their species – because having developed in a particular way is what being a member of a species is – but isn’t limited to it. So, my position is that Hugh and Moe’s species may be relevant to our responsibilities to them, but not for any of the reasons you raise and reject. Therefore, the OP’s argument fails against the version of teleology I’m advocating.

    Hopefully that’s a bit clearer.

    ReplyDelete
  10. Sam, please re-read my previous comment. I grant that Hugh and Moe won't be developmentally identical. I simply ask you to imagine a case which comes as close as possible to this. In particular, I asked you to imagine a case where the retardation was due to a congenital defect rather than a freak brain injury in mid-life.

    You complain that I've left out too many details. But I've made it clear how I want you to fill in the details. So please, just go ahead and fill them in. Co-operate a little bit. Do you think we have reason to favour Hugh in the case where his retardation is congential? (If not, then my post is not trying to argue against you.)

    ReplyDelete
  11. I should add: if you think that (even given the restriction to congenital disabilities) the case is still underspecified in a way that leaves it indeterminate whether to favour Hugh, then the most helpful way for you to respond would be to simply spell out the two divergent precisifications you have in mind. You keep saying that "developmental history" is relevant, but until you engage more with the particular cases, it remains entirely obscure just what developmental features you think are relevant (and -- what interests me more -- whether these are developmental features that necessarily differ when species differ).

    ReplyDelete
  12. For what it's worth, Richard, my own intuition is that one ought indeed favor treating Sally over others. I suppose, in the end, errors in deep developmental history feel to me similar enough to the kinds of later errors Sam C mentioned--the further back in developmental history we go, the weaker the analogy becomes, but there always seems to remain at least a slight (and easily swamped) residue of entitlement.

    I grant that many others would not agree; I'm also aware that my intuitions about this whole puzzle have already proved themselves fickle, but, there they are. So, I'm curious, can you identify why Sally's genetic heritage does not seem to you to give her at least a minimal entitlement over plain Joe to the procedure? Why does it not seem to you a morally relevant difference between the two? Is this just a matter of rock-bottom intuition, or are there reasons at play that rely on deeper intuitions the two of us might share?

    ReplyDelete
  13. Hmm, we're probably getting pretty close to bed-rock here, but let me propose the following welfarist argument.

    If Sally is more entitled to aid than Joe, this must be because either (i) she is worse off than Joe, or (ii) she would benefit more from the aid than Joe. But neither of these conditions holds in this case. So there's no more reason to aid Sally than Joe.

    ReplyDelete
  14. I don't understand how you can claim with any certainty that Moe and Hugh are "fundamentally different beings" if they are psychological duplicates with identical thoughts and feelings. It seems to me that you have implicitly stipulated, with your assumptions, that they are not in fact fundamentally different types of beings, and that claim later is an add-on without justification. From this perspective, I'd say we could legitimately toss a coin--in fact should toss a coin, as to favor either is to be unjust.

    ReplyDelete
  15. James - I think a biological essentialist could coherently insist that (i) one's "kind" is just one's species, and (ii) Moe and Hugh are certainly of different species -- they have different DNA and developmental histories: Moe's parents were unambiguously monkeys, whereas Hugh's parents were unambiguously human. And note that there's no contradiction in the idea that beings of different species may happen to end up in exactly similar mental states.

    It sounds like you want to reject claim (i), and maybe hold instead that what kind of being you are is determined by your psychological (rather than biological) characteristics. Is that right? It's worth noting that this is pretty counterintuitive. Most folks would think that a retarded human being (with monkey-level intelligence) is to be classified as the same kind of thing as other humans, rather than as the same kind of thing as monkeys.

    So, while I agree with you that we shouldn't favour Hugh, I might sooner reach this conclusion by downplaying the moral importance of "kinds of being" facts, rather than offering a revisionary classification of what kind of being Hugh is.

    ReplyDelete
  16. um..not to be a smart alek, but monkeys are way down the list, I think you are referring to Apes. From dictionary.com
    any of a group of anthropoid primates characterized by long arms, a broad chest, and the absence of a tail, comprising the family Pongidae (great ape), which includes the chimpanzee, gorilla, and orangutan, and the family Hylobatidae (lesser ape), which includes the gibbon and siamang.

    Monkey: any mammal of the order Primates, including the guenons, macaques, langurs, and capuchins, but excluding humans, the anthropoidapes, and, usually, the tarsier and prosimians.

    There is no way any monkey would have a comparable intelligence to a human.
    And it undercuts the whole argument when you can't even get the basic terminology right. And since the author can't even get this right, it shows just how little he knows of ape cognitive abilities and physiology. Even our nearest relatives in intelligence, Chimps and Bonobos are radically different than us in ways that go far beyond simple measurements of intelligence.
    There could be no way in which Hugh and Moe are psychological twins.

    One element that has astounded researchers is how much different are Bonobos from Chimps. If psychologists had studied Bonobos first, how differently evolutionary psych. might have played out.

    Our course we will favor humans. They look like us, smell like us, come from one of our fellow humans bodies.

    The only way this philosophical examination could be worthwhile is if Neanderthals had survived. But, my oh my, how different would our history have been.

    ReplyDelete
  17. I think you're rather missing the point of the thought experiment.

    ReplyDelete
  18. look, I know it is a thought experiment, but it is way too constrained by its inner inconsistencies, think of it this way, if we could using some kind of bio-technology to vastly increase a Chimp's intelligence, would you do it? Or would you invest that money in a mentally disabled individual? Planet of the Apes notwithstanding, a reasoning, thinking Chimp would seriously freak us out and force us to question what it means to be human, so for a host of reasons I see helping the mentally disabled individual would take overwhelming precedence.

    A talking Dolphin on the other hand, would be way cool.

    ReplyDelete
  19. You seem to have in mind merely instrumental reasons for favouring the human. That wasn't the question. (Re-read the introductory paragraph.) The question is whether it's somehow intrinsically better to bestow rationality on (i) a "defective", non-rational member of a normally rational species, rather than (ii) a "normal", but similarly non-rational member of a normally non-rational species.

    In other words: does one's species have any intrinsic moral significance, over and above one's individual traits? One candidate proposal, explored in this post, is that species matters because it influences whether one's individual traits count as defective (such that "curing" them is imperative) or normal (such that "enhancement" might be nice, but not imperative). But my subsequent discussion showed that this intuitive idea is difficult to ultimately justify.

    ReplyDelete
  20. The question is whether it's somehow intrinsically better to bestow rationality on (i) a "defective", non-rational member of a normally rational species, rather than (ii) a "normal", but similarly non-rational member of a normally non-rational species.

    Yes, it is intrinsically better for the reasons I laid out above, most of humanity simply could not deal with a rational Chimp. There is an intrinsic (moral) significance because mankind would not want to engineer its own self-destruction by manufacturing a competitor.
    Now there is a big difference between an intrinsic significance, because that significance is hardwired into our DNA, as to the morality of it, that is a different matter.

    ReplyDelete
  21. What most folks can or can't "deal with" is extrinsic to the case of Hugh and Moe. We can imagine changing this whilst holding fixed the intrinsic facts about Hugh and Moe. So do this. Imagine that other people don't care one way or another which of the two we enhance. Just focus on Hugh and Moe themselves, not on other people's reactions. Do we have more reason to help Hugh, for Hugh's sake, than to help Moe, for Moe's sake?

    ReplyDelete
  22. Easy,

    Hugo

    The problem is that they are not identical, if they were then both will either be a monkey or a human. In that case we resolve it with a coin-toss.

    Now Moe is probably very successful and happy in monkey society because of his above average intelligence.

    Now Hugo is very disadvantaged in human society.
    So a simple utilitarian calculation.

    We are dealing with a non-linear utility function here.

    You cannot really make everything intrinsic, in that case you have to remove society forever, Moe AND Hugo will never interact with anyone again (that's the only way you are going to get other people/monkeys to not care).

    In that case both will be identical, you reduced them to the psychological only. Again coin-toss is the only objective or utilitarian solution.

    ReplyDelete
  23. Sl - you've missed the point. Of course in practice we couldn't just ignore all the instrumental reasons. But this is a thought experiment -- a theoretical exercise -- whereby we intentionally bracket some reasons to better see what (if any) other kinds of reasons remain.

    It sounds like your response is to insist that if they are psychologically identical, then they must be the same in every respect. But that seems mistaken. See my response to James upthread.

    (Perhaps you're really just expressing your view that the difference in species is not really a morally relevant difference in its own right. That's cool, I agree. But most people wouldn't, hence my attempt to explore whether their view can be made coherent.)

    ReplyDelete
  24. Hi Richard,

    I don't think I missed the point, you presented me with a "Hugo or Moe" puzzle.

    "But this is a thought experiment -- a theoretical exercise -- whereby we intentionally bracket some reasons to better see what (if any) other kinds of reasons remain."

    OK yes the first part of my reply was the real world which is Hugo, rather than Moe.

    In the bracketed world, yes there is a difference in species, yes they are different.
    But what happens in this case is that this is irrelevant now. I cannot incorporate it in the utility functions, so it becomes perfectly linear between the two (because I can only use psychology which IS identical). So when I do my calculation they are the same in both cases.
    So I get the idea you want me to be perfectly rational and objective, yes?
    In that case a flip of the coin.
    If I am biased towards monkeys then obviously Moe and vice-versa. There is incidentally no difference between the biased and objective coin-toss as long as you remain perfectly rational through your calculations up until the decision point, because the utility gained would be the same even if the final decision is biased.

    So yes the species is not morally relevant, but I also don't see how exercising your bias at the end would be immoral in this case, the end result is ultimately the same if by chance or by bias.

    ReplyDelete
  25. My unfiltered response is that it is very difficult to buttress the intuition that one ought to uplift the human's intelligence in preference to the monkey's. If we accepted an Aristotelian conception of species whereby each member of a species has the same essence and each essence is associated with what Elliot Sober calls a 'natural state' towards which the animal would would tend were its development not extraneously affected, then we might want to say that Hugh's natural state involved being more intelligent than Moe.

    However, I think most philosophers of biology deny that species have essences in this sense. There is (and natural selection demands) considerable genetic and morphological variation within any population, so it seems arbitrary to select some statistical cohort of a species population as a the determinant of the natural state rather than an other.

    Most importantly Darwinian species undergo change, sometimes branching off into new species in response to selection pressures and other historical contingencies, not to a telos determined by ahistorical principles of species-membership.

    So it is hard to see what possible role the idea of a natural state for humans as (opposed to monkeys, say) could play in modern biology.

    On an ethical note the natural state model also seems unfair to both Moe and Hugh. Hugh might be less than averagely intelligent but does thus justify us seeing him as a defective human! On the other hand, Moe might benefit just as much as Hugh from an intelligence uplift.


    David Roden

    http://www.enemyindustry.net/

    ReplyDelete

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)