Feeling free

Eddy Nahmias recently reported on a ground-breaking case in Japan where a care-giving robot was held responsible for agreeing to a patient’s request for a lethal dose of drugs. Such a decision surely amounts to a milestone in the recognition of non-human agency; but fittingly for a piece published on 1 April, the case was in fact wholly fictional.

However, the imaginary case serves as an introduction to some interesting results from the experimental philosophy Nahmias has been prominent in developing. The research – and I take it to be genuine – aims not at clarifying the metaphysics or logical arguments around free will and responsibility, but at discovering how people actually think about those concepts.

The results are interesting. Perhaps not surprisingly, people are more inclined to attribute free will to robots when told that the robots are conscious. More unexpectedly, they attach weight primarily to subjective and especially emotional conscious experience. Free will is apparently thought to be more a matter of having feelings than it is of neutral cognitive processing.

Why is that? Nahmias offers the reasonable hypothesis that people think free will involves caring about things. Entities with no emotions, it might be, don’t have the right kind of stake in the decisions they make. Making a free choice, we might say, is deciding what you want to happen; if you don’t have any emotions or feelings you don’t really want anything, and so are radically disqualified from an act of will. Nahmias goes on to suggest, again quite plausibly, that reactive emotions such as pride or guilt might have special relevance to the social circumstances in which most of our decisions are made.

I think there’s probably another factor behind these results; I suspect people see decisions based on imponderable factors as freer than others. The results suggest, let’s say, that the choice of a lover is a clearer example of free will than the choice of an insurance policy; that might be because the latter choice has a lot of clearly calculable factors to do with payments versus benefits. It’s not unreasonable to think that there might be an objectively correct choice of insurance policy for me in my particular circumstances, but you can’t really tell someone their romantic inclinations are based on erroneous calculations.

I think it’s also likely that people focus primarily on interesting cases, which are often instances of moral decisions; those in turn often involve self-control in the face of strong desires or emotions.

Another really interesting result is that while philosophers typically see freedom and responsibility as two sides of the same coin, people’s everyday understanding may separate the two. It looks as though people do not generally distinguish all that sharply between the concepts of being causally responsible (it’s because of you it happened, whatever your intentions) and morally responsibly (you are blameworthy and perhaps deserve punishment). So, although people are unwilling to say that corporations or unconscious robots have free will, they are prepared to hold them responsible for their actions. It might be that people generally are happier with concepts such as strict liability than moral philosophers mainly are; or of course, we shouldn’t rule out the possibility that people just tend to suffer some mild confusion over these issues.

Thought-provoking stuff, anyway, and further evidence that experimental philosophy is a tool we shouldn’t reject.

10 thoughts on “Feeling free

  1. “People are saying…”. Apropos freedom and responsibility, check out this recent blog post by Colin McGinn, “On Evil”.

  2. I’m generally a fan of empirical investigation, but I’ve always wondered what distinguishes experimental philosophy from social psychology, aside from the department administering the study.

  3. Compare ‘cognitive processing’ with ‘responsive processing’
    …cognitive would be one’s thoughts being processed…
    …responsive would be one’s feelings being processed…

    There is history for this in analytic philosophy…
    …from the influence of (Eastern) exercises, meditation and acupuncture…
    …towards sensing-observing areas of the human body…

    That from MRI, images of the use of energy by humans come about…
    …there are processes like the solarplexes and other systems of the body…

    How long have we been ‘experimenting in philosophy’…
    ..Which process was first in evolution…

  4. I love experimental philosophy (X-Phi). I agree with Sellars that philosophy’s greatest service is to map the relationships between the manifest image of the world and the scientific image. But the manifest image isn’t as transparent as most philosophers assume – as shown by the fact that philosophers disagree about what the manifest image says. That’s where X-Phi shines: it can provide copious evidence about what people actually believe and why.

  5. Wilfrid Sellars (Stanford Encyclopedia of Philosophy); Good quick read…

    But, what is between left and right could be philosophized:
    …Between is a concept equal to a left concept and equal to a right concept…

    When in play…a = b = c = a = c = b = a…the subjectivity of objectivity…thank you

  6. With respect to free-will, I seem always to return to Bertrand Russell’s pithy comment: “Freedom is the ability to do as we please, not to please as we please”. An objection is that in some respects we are able to please as we please —e.g. to modify our tastes or dispositions. But the intention to modify our tastes or dispositions is just doing what we please.By the way, completely agree with Paul Torek’s comment. Sellars (and Dennett) have it right there.

  7. While waiting for Peter…

    Could post modern philosophy today be from…

    …to have to give to be to have…

  8. I think you are on to something where you suggest that people don’t see free will in action where there is a rational optimal choice with no downside to taking it, but I was surprised by your suggestion that choosing a lover is a clear example of free will – for one thing, at least in my case, it was certainly never a given that the other party would be of like mind, and I am also well aware of how much chance played a role.

    Free will seems to live in a middle ground between the deterministic and the random, and trolley problems try to go there with situations in which there are clear choices, but no global optimum, and where no choice — even to do nothing — leaves things up to chance.

    When people want to hold corporations responsible for their actions, I think that is usually shorthand for holding the decision-makers responsible, and one reason for their being happier with strict liability than are moral philosophers is that they are being pragmatic, and cannot wait for the latter to all agree on the issue. I think there is a fairly wide understanding, for instance, that while strict liability may sometimes be unfair, without it, greater unfairness overall would result.

  9. There will never be a case that Nahmia suggests of course. Free will will never be attributed to robots because it’s no more meaningful to talk of computer programs having free will than any other machine, such as a car. Or a teacup.

    There are well-established principles for engineering negligence in computer software : these will apply. The “robot” won’t be blamed, but the free-will exercising, decision-making engineers and their employers will, the same as any other action in negligence of the same type. Corporations are bodies of people, computers aren’t. Collectively the individuals within corporations make decisions.

    Free will – or the sense of it – is a purely internal state. It is itself a feeling – the feeling that at any point we can do – at least for a certain array of tasks before us – any one of a multiple number of things. This sits side by side with a certain realisation that the overwheleming majority of what we do, from walking to learned typing, is done almost entirely unconsciously. Free will is an objectively-existing subjective mental state, not to be confused with indeterminism, which is an objective state of affairs based upon (in contemporary ideas) a certain clockwork conception of physics (including magical quantum physics by the way) that must impact the function of the brain.

    I suspect that people are willing to think that computers have agency at the point they are willing to beieve they have human minds. The important thing is that they have to have human features, rather than specific assets such as certain emotions. We think things think like us when they are like us.

    None of which – of course – affects the objective scientific existence of indeterminism or otherwise.

    J

Leave a Reply

Your email address will not be published. Required fields are marked *