Tuesday, April 09, 2013
Punishment at the Speed of Light
I have written before about why it's a mistake for theorists to measure the severity of incarceration in objective terms, like months in prison. Here's a new argument adapted from Against Proportional Punishment (Vand. L. Rev., forthcoming 2013):
Special relativity teaches us that even if we wanted to resort to objective measurements of time, there is no observer-independent rate at which time passes. To illustrate, assume that in the distant future, we develop spaceships that travel at speeds approaching the speed of light. Due to the effects of special relativity, a person on such a ship will age slowly on average from the perspective of people on Earth. For example, a space traveler might age four years but come back to a planet whose inhabitants are eight years older.
Now suppose that twin brothers, alike in virtually all respects, commit crimes of equal blameworthiness. The only pertinent difference in their circumstances is that one is incarcerated on Earth while the other is incarcerated on a spaceship traveling near the speed of light. Assuming the twins deserve equal treatment, is it more accurate to measure the duration of their sentences based on an Earth clock or a spaceship clock? The answer is neither. As a first approximation, the duration of each twin’s confinement depends on the clock in the twin’s frame of reference.
But if we ought to individualize time measurements based on frames of reference, why stop there? Just as one person’s clock might appear to tick slower than another’s because of special relativity, one person might experience time moving slower than another because of his unique brain chemistry. When assessing the severity of harms associated with confinement, there is no obvious moral reason to consider the ticking of their clocks but not the ticking of their brains, so to speak.
Importantly, experiencing the passage of time is not itself a harm. But it is closely related to harmful experiences, like the distress and boredom associated with confinement. These experiences are clearly relevant to our assessments of how harshly offenders are treated by confinement. (There may be room to consider objective measurements of time when calculating the societal benefits of incapacitating offenders, but objectively measured elapsed time is not directly relevant to our assessments of sentence severity.)
Of course, it is extremely difficult to measure the experiences I'm discussing. In the real world, we have to settle for rough proxies, at least for the near future. But if you're a punishment theorist, you should care about figuring out what really matters in the absence of cost and technological constraints. After you've done that, you can better return to real world problems by trying to approximate, as best we can, the things we really care about.
Posted by Adam Kolber on April 9, 2013 at 07:26 AM | Permalink
TrackBack URL for this entry:
Listed below are links to weblogs that reference Punishment at the Speed of Light:
Adam, in your hypothetical, isn't the obvious answer that we should measure the time of imprisonment based on the space ship clock? The fact that time will seem to vary from a different reference at a relative speed near c seems plainly irrelevant. It makes no more sense to measure the imprisonment time based on that other reference than it does to measure a lawyer's billable hours from that reference point.
Posted by: Orin Kerr | Apr 9, 2013 12:31:43 PM
Thanks Orin! It makes much more sense to use the imprisoned space traveler's clock than an Earth clock. The question is: why stop there? If we care about the duration of incarceration because it affects how much offenders suffer, it would be more accurate still to just consider how much they suffer (which is only partly related to the ticking of a clock). If we care about the duration because it affects how much we deter, then we should focus on how much we deter (which will also depend on the ways people suffer and anticipate suffering in ways not fully captured by clocks).
It is tempting to focus on time because it seems like we're using the same variable for everyone. But just as an Earth clock mismeasures the severity of the space traveler's sentence, his own clock mismeasures it as well. Objective measurements of time, even in one's local frame of reference, are not what we ultimately care about.
I am not arguing that it's easy to measure suffering. In fact, I'm sure sentences of incarceration have a fetishistic focus on duration because time is so easy to measure. But if we don't know what we really care about when measuring sentence severity, we will never know how closely we are approximating it.
Posted by: Adam Kolber | Apr 9, 2013 12:59:53 PM
From the reference frame of a prisoner on a spaceship traveling at a constant near-light velocity with respect to Earth, it's the prisoner on Earth who will appear to age more slowly. Implicit in your hypothetical that we're dealing with a round-trip journey, in which the spaceship accelerates, and so the spaceship prisoner's reference frame is non-inertial. Some might consider this fact significant in deciding which reference point we should adopt to measure the term of imprisonment for reasons that have nothing to do with either prisoner's subjective experience.
Posted by: James Grimmelmann | Apr 9, 2013 1:07:15 PM
Thanks James! Yes, I did, indeed, envision a round-trip journey (for the reasons you stated) and should have made that more explicit.
Posted by: Adam Kolber | Apr 9, 2013 2:07:18 PM
The term "just as" is doing an awful lot of work in this hypothetical, but you really have not established the equivalence of the two situations. Lots of things would be different in space or on other planets, but that does not imply reconsideration of phenomena on earth. The non-objectivity of time passage for space travelers is a consequence of physics. The proposed non-objectivity of several earth-bound prisoners would be a consequence of psychology or neurology. Why is one "just as" the other?
Posted by: Steven Lubet | Apr 9, 2013 3:00:15 PM
Suppose that younger people perceive a year to be a longer period of time than older people, but on the other hand have many more of them left before they die.
Who gets the longer punishment?
Posted by: brad | Apr 9, 2013 3:18:05 PM
Thanks Brad! I don't think that "perception of time" is the relevant variable. But I do think that the amount of distress and boredom one is likely to feel during a given period of time in prison will vary somewhat based on age. Age would just be a rough proxy, though, for what particular individuals experience or are likely to experience.
Does punishment severity depend on how many years of life one has left? I suspect it does. The person who believes he will die in prison may experience his sentence differently than the person who expects to live a long time. And you're right that however you answer that may be counterbalanced by other age-related ways in which people tend to experience prison. So it's an empirical question, and one that likely depends on the particular individuals at issue.
Posted by: Adam Kolber | Apr 9, 2013 3:52:54 PM
Thanks Steven! I agree entirely with the gist of your comment. The argument in the post is just a snippet of a much longer argument. Consider this though: Suppose you hate receiving injections much more than other people do. When you decide whether or not to get some particular injection, you don't ask: "How does the average person feel about injections?" Rather, you introspect because that's what's relevant to your decision. So the longer version of the argument says, let's look at the leading rationales for punishment and claims that they require the more experiential approach.
It is tempting to rely on objective criteria because they are easier to measure. But I think they are just proxies. Suppose it was really hard to measure the passage of time but easy to measure suffering. Would we bemoan the lack of an objective measurement and just use suffering as a rough proxy for the passage of time? I think not.
(Typepad isn't letting me post the full comment, so I'm breaking this up into two comments to see if that helps.)
Posted by: Adam Kolber | Apr 9, 2013 4:02:12 PM
Reply to Steven, Part 2:
If we seek a moral justification of punishment, then we need to consider how much harm we cause prisoners. If we mismeasure that harm, we will not know if our practices are justified. In tort law, we at least purport to use subjective measurements of harm. We don’t give ten thousand dollars to all people with a broken leg. It depends on the individual’s pain and suffering, as best the jury can discern.
Posted by: Adam Kolber | Apr 9, 2013 4:08:06 PM
Reply to Steven, Part 2:
To morally justify punishment, we need to consider how much harm we cause prisoners. If we mismeasure that harm, we won't know if our practices are justified. In tort law, we at least purport to use subjective measurements of harm. We don’t give ten thousand dollars to all people with a broken leg. It depends on the individual’s pain and suffering, as best the jury can discern. So at a minimum, there's a difference between the way we treat harm in tort and in criminal law.
(As an aside, it is true that the relativistic effects of time would only be very noticeable to our untrained eyes in these science fiction scenarios. Nevertheless, the relativistic nature of time applies to us here on Earth, as well; we just need the right tools to see it.)
Posted by: Adam Kolber | Apr 9, 2013 4:19:46 PM
Adam writes: "If we care about the duration of incarceration because it affects how much offenders suffer, it would be more accurate still to just consider how much they suffer (which is only partly related to the ticking of a clock). If we care about the duration because it affects how much we deter, then we should focus on how much we deter (which will also depend on the ways people suffer and anticipate suffering in ways not fully captured by clock."
Adam, what is "suffering," in your view?
Posted by: Orin Kerr | Apr 9, 2013 5:19:16 PM
I think it's hard to say exactly what suffering is, but it's some kind of negative experience. There certainly is a durational component to suffering. Longer bouts of suffering at a certain intensity are worse than shorter bouts at the same intensity. I wouldn't want to argue that duration is entirely irrelevant to suffering, just that purely objective accounts of sentence severity are missing out on a fundamentally important dimension of severity.
Posted by: Adam Kolber | Apr 9, 2013 8:23:51 PM
If your theory is that punishment has to be closely tailored to suffering, isn't it a problem that it's hard to say what suffering is? I can understand if your position is that you know what suffering is but that it's impossible to measure it objectively, but it seems particularly problematic if we don't know what it is *and* we don't know how to measure whatever it might be.
Posted by: Orin Kerr | Apr 9, 2013 11:47:22 PM
Aha, now I see what you have in mind, Orin. If you had asked me, "Adam, what do you mean by time?" I would have given the same answer: "It's hard to say exactly what time is." Maybe suffering is the harder concept of the two. For one thing, suffering is a conscious experience, and nobody really knows exactly what consciousness is. On the other hand, if you're dealing with morality, I think touching on such issues is not a bug but a feature.
Deep philosophical problems aside, however, we have a pretty good sense of what suffering is in the legal system. After all, whatever we do in tort law purports to look at suffering, even though there is no widespread agreement as to exactly what suffering is or how to measure it. I'm not saying that we do it perfectly in tort, but I'm not expecting more of criminal law. Also, we already have this problem in criminal law to the extent that suffering (or similar language) is sometimes an element of a crime (e.g., compare aggravated assault to simple assault) and is also relevant more generally to considerations of victim harm at sentencing.
In any event, even if suffering is a harder concept than time, I'm not sure how much that would count against a theory that refers to suffering. One could say that worker's compensation schemes avoid (at least some) questions about suffering and certainly simplify questions of measurement. But how good/accurate is some particular worker's compensation scheme? Arguably, you can't answer that without knowing something important about suffering. Thanks, Orin!
Posted by: Adam Kolber | Apr 10, 2013 3:22:59 AM