In Prison, Without Knowing Why

This article in the New York Times caught my attention yesterday,
partly because just a few weeks ago I commented on paper at the Central APA meeting on the
punishment of late stage demented offenders:

The number of prisoners in the US with dementia is on the rise. The NYT
article above notes that due to budget constraints, offenders whose cognition
is so compromised that they cannot take care of themselves are being
cared for by other prisoners. This is a risky solution at best.

I don’t think late stage demented prisoners should be in prison at all. Proportionality is a hallmark of retributive punishment, the primary principle of punishment according the US Model Penal Code. To be just, punishment must be proportional to either the crime committed or harm caused, and to the type of offender. A murderer deserves more punishment than a car thief. And a person who is insane or mentally incompetent is not as deserving of punishment as someone with normal mental capacity. This is because one with diminished mental capacities does not know what they were doing is wrong in the same way a competent person does, or they fail to fully understand the consequences of their acts.

This requirement of proportionality is not only relevant at the time the crime is committed. In Panetti v. Quarterman, the US Supreme Court refused to allow Panetti to be executed because he thought he was being executed for his religious beliefs, instead of his earlier crimes. According to the Court, retribution requires that capital punishment “…have the potential to make the offender recognize at last the gravity of his crime and to allow the community as a whole…to affirm its own judgment that the culpability of the prisoner is so serious that the ultimate penalty must be sought and imposed. The potential for a prisoner’s recognition of the severity of the offense and the objective of community vindication are called in question, however, if the prisoner’s mental state is so distorted by a mental illness that his awareness of the crime and punishment has little or no relation to the understanding of those concepts shared by the community as a whole.” That is, if the person doesn’t understand that they are getting their “just deserts,” they feel no connection to society’s values or judgment of their acts.

For this reason, a late-stage demented offender who cannot remember their crime is no longer the person who deserves their just deserts. If he cannot link the suffering imposed by imprisonment to their earlier act, he is not enduring punishment proportional to their harmful act. I thus think where there is no basic comprehension of the connection between current punishment and an earlier crime, late stage demented offenders should be released or civilly committed.

For a fuller (more popularized) treatment of this topic using all the principles of punishment, see:

Any comments would be appreciated.


  1. Katrina, I think the points you make are right as far as they go, and I had a similar response when I read the article. But this also strikes me as a situation in which our moral intuitions conflict with a very ‘real-world’ resource allocation problem. It’s not just the expense of moving these people out of prisons and into nursing homes. It’s also the fact (if memory serves me) that nursing homes don’t want convicted murderers, particularly those still prone to violence and who are happy to chat about their horrific actions, and many of these individuals are alienated from their families. So I think it’s important to look at the problem not only through the eyes of a philosopher of law and an ethicist, but through the eyes of a decision-theorist.

    There’s more to this utilitarian perspective. Not only are we talking about the alarming costs of alternate habitation, but the very real feelings and attitudes of the victims and/or their families. How would these people feel about the fact that a convicted felon who brags about ripping out a woman’s eyeballs can live out the remainder of his or her life in relative comfort? Particularly given that, since dementia has an earlier onset in prison, these individuals may be alive for decades to come? (I’m not claiming they deserve punishment; I’m claiming that the victims’ feelings are a relevant consideration.) Finally, the opportunity to care for a prisoner with dementia has had transformative value for other prisoners, some of which have gotten early parole because of their service. Again, I can’t really argue with anything you say – it’s all technically correct, as far as I’m concerned – but I wonder if this isn’t a case where our retributionist moral intuitions have to be tempered by utilitarian considerations.

    I don’t think it’s enough to take a ‘hard line’ response and argue that utilitarian considerations are irrelevant to the moral issue. It strikes me that a growing (and positive) trend in applied ethics is to take a ‘principles’ approach in which moral problems aren’t resolved by the firm application of a single moral principle, but through the negotiation of several different principles including social costs. So I don’t mean to imply that this is a case where morality gets trumped by social costs. I want to suggest that this is a case where social costs of the kind suggested above represent a genuine ‘moral’ constraint on the right solution.

    BTW there’s another great article in the NYT that came out only about a week before about the growing population of mentally ill prisoners in Arizona. I can’t find it right now but a similar sort of case could be made for it.

  2. gualtiero piccinini

    Justin, your points are well taken but they mostly apply to early stage dementia.  Katrina is mostly talking about late-stage dementia, and I agree with her that it’s inhumane to keep such people in prison, or at least in ordinary prisons.

  3. I would agree with that – I didn’t get the restriction from the initial post. Arguably the case is much stronger for transferring them as the dementia progresses. Although part of my assent is governed not by the retributionist ethic but the fact that we have a much smaller proportion of prisoners, so the costs are more manageable. Perhaps a metric such as ‘proportion of lucid days to non-lucid days’ would be helpful – as it decreases, the case for transferring them is greater. Though there’s an interesting consideration here, which is somewhat tangential: many people with dementia have ‘lucid’ and ‘non-lucid’ days (or moments); in the former, they remember their actions and identify with the person who carried them out, but in the latter they no longer do so. From a strict retributionist perspective, should we conclude that punishment is acceptable on the lucid days and unacceptable on the other? That could make for some interesting arrangements.

  4. Katrina Sifferd

    Thanks for your comment, Justin, and for giving me the head’s up about the other article. I’m certainly supportive of considering things such as social cost when attempting to solve problems within the criminal justice system. It may be very upsetting for victims and family members of victims to learn that the person who harmed them is being released from prison. However, where the offender is truly late stage demented I think it is possible that victims may feel the pull of the moral principle I discuss: that is, they may feel their retributive sentiment is no longer justified.

  5. Kingson Man

    Another ethical quandary: as treatments for e.g. Alzheimers improve, prison doctors may be in the position of curing their patients so that they may be competent for execution – lucid enough to be put to death!

  6. Katrina Sifferd

    That is an interesting and worrying thought. In this case, the conditions for application of punishment come and go – a real problem, as you say, for the retributive account.

  7. On second thought, I don’t know if it’s a serious problem. If one accepts a Lockean-style account of personal identity and responsibility (which is what I had in mind initially), in which the ability to remember one’s crime is necessary for just punishment, then the oscillation between lucidity and non-lucidity would imply an oscillation in the conditions for the justness of imprisonment. But you don’t seem to appeal to that, as much as you appeal to a deterioration of mental capacity, including, but not limited to, whether and to what extent the individual can remember the crime. I suspect that even for a Lockean account, many of these prisoners, even on their lucid days, do not actually ‘remember’ the crime, but they confabulate the events surrounding the crime based on the testimony of others. In that case, even if a demented prisoner boasts about the crime, and his or her description of the crime is consistent across time, that doesn’t mean that he or she is deserving of punishment. So I suppose that ‘just punishment’ would depend on a general evaluation of mental competence not unlike the evaluation of competence for medical patients who want to forgo treatment. Two relevant links – one is a recent article about mentally ill prisoners in Chicago (not Arizona) and the other is a supplemental series of quite disturbing photos from the one you mention.

  8. Bill

    That quandary has already been encountered with the use of anti-psychotics:

    a href=””> here

  9. Biil

    …punishment is acceptable on the lucid days and unacceptable on the other?
    (end quote)

    I don’t think that follows, else we might think it right to forbid prisoners to sleep beyond some minimum to stay alive.

    Yet another mitigating factor in their imprisonment is that as dementia progresses, most such demented persons who are impaired enough to no longer remember that they committed a crime may also forget that they are even in prison.

    That type of spatial disorientation would tend to decrease their discomfort roughly in proportion to the severity of the dementia which makes them unaware of their crime.

Comments are closed.

Back to Top
%d bloggers like this: