Blog Coordinator

« NOWAR 3 Program is Here! | Main | CFP: 3rd Annual Philosophers' Cocoon Philosophy Conference »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Why can't we expand the 'width' of our tracing to include the sorts of decisions Bert made that shaped his bad habits and character? I'm presuming he has the capacities to understand facts like: when you don't keep a calendar or keep a schedule, you are likely to forget some important things (something he's presumably experienced), more important appointments (like taking care of kids) are more important to remember, etc. And he has the capacities to alter his behaviors in light of those facts (e.g., to get a calendar). And he had the opportunities to exercise these capacities. Plausibly, he even made some conscious decisions *not* to exercise these capacities--e.g., thinking things like, "I'm not gonna tie myself down to a calendar that tells me where to be when" or "Screw my ex--she can call me to remind me about the kids." In any case, this is the sort of tracing picture I have in mind when I consider him blameworthy and deserving of punishment (though details about the case might suggest diminished degrees of blame).

And I'd think the law's negligence standard uses something like this view of tracing. We expect people to take reasonable precautions to avoid harm, including shaping their habits (e.g., driving) to meet reasonable expectations.

Eddy, it would make more sense if the law's negligence standards traced back to general capacities and not specific decisions, because requiring proof of specific thoughts like "screw my ex -- she can call and remind me" beyond a reasonable doubt seems unreasonable. That is, looking for such specific thoughts may be beyond the fact-finding skills of the court.

But if we are going to find someone negligent based upon the possession of these general capacities, do we really need tracing? Can't we just say Bert violated some objective standard of care and that he had the capacity to meet that standard of care? Bert had the capacity to understand his duties to his kids, and to meet them, at the moment he decided to go to Vegas -- no need to trace backward in time to prove that. It is true that it makes sense to understand these capacities as diachronic, but I'm not sure the idea that the law tracks diachronic capacities requires that we trace back to any particular decisions or points in time.

I'm not sure what the law should (or can) do about these cases. But from the perspective of attributing (degrees of) blame (MR) and potentially deserving to be punished, I'm inclined to think that 'wide tracing' matters. It seems to make a difference to whether we attribute ill will to the person or think they have the relevant capacities, so maybe it's more of an evidential condition for both quality of will and control theories. But I think I'd defend a stronger claim and say that it matters (beyond the evidential role) whether the person exercised the relevant capacities in making a decision--or more often, a series of decisions--that influenced their bad action even though that action did not involve a (conscious) decision or an exercise of the relevant capacities.

Does this example help: Hert, unlike Bert, is a good father but from a culture where leaving kids (e.g., age 11) alone for a weekend is no big deal (and is certainly legal). He has the relevant capacities to know the U.S. law, having lived here a few months, but he never exercised those capacities even though nothing prevented him from doing so. When he leaves his kids for the weekend, I think he's way way less blameworthy than Bert, and I think it's because of relevant differences in their histories (though those differences are also evidence for very different qualities of will).

Dear Katrina,

I agree with you that tracing will not help here. But I also agree with Eddy that merely having the capacity to do as one should is not enough. My view, sketched in Amaya & Doris (2014), is that moral responsibility for lapses requires a further non-glitch condition. That is, the lapse ought not to be the result of a mere glitch or hiccup in "normal" functioning but the result of the agent functioning as he or she normally does. This is why habits might ground attributions of MR in cases of negligent behavior.

Would you agree that something like a non-glitch condition is necessary? Do you think it is met in Bert's case.

Hrmf, Eddy beat me to the punch--it was sounding like the presence of ``screw my ex!''-type thoughts was only pulling, at best, some evidential weight. What really seemed to matter, even in Eddy's comment, was whether Bert possessed and/or failed to exercise the relevant diachronically construed capacities. Plus, perhaps, the presence of some kind of non-glitch condition? That sounds potentially promising to me--looking forward to reading that paper, Santiago!

But then I'm not sure I know what to say about Hert, other than it seems like a case where moral responsibility and legal culpability might, for a variety of reasons, diverge somewhat. Well, OK, I'll say one thing: compared to Bert, Hert will have a very different sense of what King (2009) calls the expected negative value of the potential harm involved in leaving his kids alone for a weekend, won't he? I suppose we could ask whether Hert has a grossly inaccurate sense of this expected negative value, and if so whether he is blameworthy for *that*, but I'm not sure I see how any of that helps the cause of tracing, wide or otherwise--nor the general notion that conscious states are necessary for responsibility.

The thing I found most telling when this case was introduced was that Bert often tried to make changes to the schedule in the past and had failed to take the kids when he was scheduled to on more than one occasion. This strikes me as a very significant feature of the case. It suggests that Bert was at least aware, at some point, that he had a negligence problem with regard to his children but did not improve or correct his habits. The details matter a good deal here. For instance, did he knowingly make no effort to keep a calendar or schedule of his dates, even while knowing that he had failed to make good on those dates in the past? Or was it that he just didn't see that there was any risk in neglecting to do so? Or did it not even occur to him to keep a calendar and/or reminders at all?

It is hard for me to see how this case isn't analogous to the classic drunk driving case. It is highly unusual (and highly unlikely) for a person to go through life without knowing the effects of alcohol. This is why it is close to impossible to be (morally) excused for drunk driving, and why it is such a canonical case of negligence. It would take a whole lot of convincing to show that Bert's case is unlike the drunk driver's (at least for MR), and that he never knew the risks of his negligent habits.

Ok I'm going to go a bit rogue here to posit that tracing concerns only relate legally to challenges to mens rea, which (in US law generally) in fact is presumed in most cases of criminal action as determined by 1st phase trials. That's why murderers are guilty if they committed the crime--they are presumed to be capable of mens rea (and not at all analyzed as such in such proceedings).

People acquitted by insanity trials (2nd phase) successfully challenge that assumption of mens rea. They were not rational and/or compelled to have acted as they did.

Negligence legally and philosophically falls troublingly in between. The question here is: what constitutes responsibility as resulting from an act arising out of circumstances ranging from a potentially culpable absence of proper intent to an incapacity to conform to proper intent? It's thus a spectrum: as one more approaches an assessment of incapacity of someone to have acted rightly, one is less responsible; as one more approaches an assessment of a proved capacity to form proper intent that was simply lacking, then one is more responsible. Of course episodic or event-indexed excuses are available to prove incapacity as a function of tracing for anyone: a mini-stoke can make anyone negligent in trying to drive. But day-dreaming, gawking, texting, and engaging in sex while driving (it does happen apparently!) all raise the stakes of the question of impropriety of intent to drive safely. And these assessments seem to ally with judgments of persistent and chronic perverse attitudes of people rather than episodic interruptions of something of otherwise more laudable intent. Did one day-dream because she just discovered her partner cheated on her? Did one gawk because he just saw his second pileated woodpecker of his birding career? But texting and sex-driving seem prima facie more a matter of chronic attitude than anything else.

So tracing might be relevant to either episodes or attitudes with respect to responsibility, but contextually.

Alan – I think I agree that, in negligence cases, a certain level of mental capacity is presumed (a rebuttable presumption). However, in murder cases the court does look specifically for intent – if using the MPC standards, the court looks to see if the defendant killed purposely or knowingly. In legal insanity cases the defendant may be proven to have killed purposely by the prosecution, but when the defense is given their turn, they may try to prove the defendant actually thought his victim was an evil alien from the plant Kripkean intent on extinguishing the human race. This would show the defendant didn’t understand the nature and/or quality of his actions and thus ought to be found not guilty by reason of insanity (NGRI). (At least in theory this is how the insanity defense works. More likely he’ll be found guilty but mentally ill and go to prison anyway.)

Legal insanity cases are unique because such defendants usually seem to have the requisite mens rea, but they are fundamentally mistaken about the circumstances of their action. (Sort of like in mistake of fact cases.) Negligence cases, however, are cases where the mens rea seems to be missing but the person is still culpable. I agree that if the mens rea is missing due to no fault of the defendant, like in your stroke case, she is not guilty due to some dysfunction of their cognitive system. But the dysfunction in the stroke case means the defendant did not commit a voluntary act, and in that case it seems we have no need to ever get to a worry about mens rea.

Eddy - Hmmm. I think Hert is less culpable because we have less of an expectation that he know the laws of child neglect and abandonment in this country (even though ignorance of the law isn't supposed to be an excuse, I think it sometimes is). We might understand this difference between Bert and Hert in the way you suggest: Bert's exercise of his capacities to care for his kids in the past indicate that he understood his obligations under the law, and thus is responsible when he fails to meet them; Hert has no such history, and thus is less culpable. But if Hert had been in the country for a full year and remained oblivious to the laws, wouldn't we fault him even if he had never exercised his capacities to know and meet his parenting obligations under the law? In this case we seem to have another objective standard in play (after one month, ignorance of the law is a partial excuse, after a year or so, it isn't), and again no need for tracing. But I think I might be carving up the relevant capacities in a different way than you are in your example.

Santiago – I like your “glitch” condition, because it isn’t as demanding as a tracing condition, and sounds like what a court tends to do: it seemed in Bert’s case the court was trying to determine whether the lapse was in character or just a one-off. Levy also argues that if we have a pattern of lapses we might assume morally responsibility for a lapse. But I wonder how easy it will be to identify a lapse as a “glitch” or instead as part of a pattern or normal operations. Bert did exhibit a pattern of being sort of irresponsible about his appointments, including his parenting schedule. But he didn’t exhibit a pattern of overall non-caring toward his kids (he paid his child support, came to some school events, etc) and he certainly had no pattern of abandoning them (or even not showing up for his scheduled parenting time). He manipulated the custody schedule, but then tended to show up when he said he would. So I'm not sure if his lapse was a "glitch" on your theory, or a part of a pattern of behavior on Levy's.
I look forward to reading your paper!

Izzy – Bert did fail to take kids on his scheduled weekend, but he had always called his ex ahead of time and told her he couldn’t take them. I think Bert certainly should have known he had a “not making his kids a priority” problem, but not necessarily a “forgetting about your kids” problem. Although I see your point that Bert probably should have known not keeping a calendar of his parenting schedule was going to mean he forgot the schedule at some point.


I agree tracing probably will not allow us to blame Bert, but I tend to think that even if it did, tracing would be a way of ascertaining that Bert is blameworthy, rather than a necessary condition for blameworthiness.
On the other hand, after considering some hypothetical scenarios, I also tend to [very tentatively] think that whether Hert actually spent a year or a month in the US or how he exercised his capacities in the past is not crucial with regard to blameworthiness, even if it may be crucial to the question of how we ascertain blameworthiness.

The scenarios are very unusual to say the least, so this is all very tentative, but I thought I'd post one of them in case it might be of interest:

Let's assume the universe has infinitely many stars, planets, etc. (which might or might not be true for all I know, but it's a matter that shouldn't have any impact on blameworthiness conditions), and Boltzmann Bert is physically identical to Bert before making the choice to go to Vegas (except for the fact that his location in the universe is different), and lives on a planet also physically identical to Earth (including the rest of the inhabitants), in a planetary system just like ours.
However, the whole planetary system is only one day old, and resulted from some weird quantum event (infinitely many of which would probably happen in an infinite universe like that, as far as I know).

So, Boltzmann Bert - who is cognitively identical to Bert - also chooses to go to Vegas, and the kids are left alone for a day. Is Boltzmann Bert any less blameworthy than Bert? Given the fact that they're cognitively identical when they make the choice, I think not. But Boltzmann Bert didn't shape his character by means of any previous decisions. He's just one day old.
A similar scenario involving Boltzmann Hert suggests that previous exercise of a person's capacities (at least, in this context) aren't crucial to the matter of blameworthiness (assuming the assessment that as long as they're cognitively identical when they make the choice, they're equally blameworthy, is correct), even if in practice, they may guide our blaming practices.

Katrina--thanks so much. I did oversimplify. My point was based on the 1st/2nd phase trial distinction of criminal action versus responsibility as the respective descriptions usually employed. Typically criminal action phases have the degree of mens rea assigned by the charge-as-tried and as assigned by prosecutors--first degree murder, second degree homicide, involuntary homicide, etc. While the charges are often challenged by the defense in terms of "my client didn't do it" (OJ murder trial, e.g.)--and that was what I was referring to--they may also be challenged as in the Boston bombing trial of Tsarnaev at the level of proper charge in terms of mens rea--the defense is trying to erode the juries' confidence that the defendant had the level of intent warranting a first-degree charge. While this challenge to degree of mens rea is relatively rare in 1st-phase trials (as far as I know), you are right that it occurs as a defense strategy.

2nd-phase trials--Hinckley, Dahmer, now Holmes--which presently are no longer possible in all jurisdictions--specifically have the issue of mens rea as the issue and difference between some kind of guilt and acquittal. (I know you know all this--I'm mostly doing this for the benefit of readers.) I agree that it is very difficult to secure acquittal in these proceedings nowadays. As I said in my Holmes posting, though I think by the letter of law he should be acquitted, he probably won't. But, and as I think happened in the case of the Dahmer jury, these are now cases where jurors simply go by their collective gut-feelings rather than deliberations that adhere to procedure, standards of evidence, and judicial instruction.

One comment to Angra: the traction of Boltzmann Hert's responsibility is significantly attached to a comparable case of someone with a history like Bert, so conditionally if Bert is responsible, then so is BH. But Bert's responsibility *may* at least in part depend crucially on a history that BH lacks, and since everything in BH's world is the result of a very recent quantum event in any case, no considerations of history can play any role in determinations of responsibility of anything in BH's world because the luck of the quantum event sweeps all such matters under its rug. But at the same time Angra, please let me say that you are one of the most interesting commentators I've run across on the several blogs you add considerable value to with your reflections.


Thank you very much for your kind words; I appreciate them, as well as your thoughtful reply to this and other posts of mine.

On the issue of Boltzmann Bert's responsibility, I was relying on an intuition (at least, I seem to have that intuition) that if two agents are cognitively identical, they're also identical with respect to moral obligations, regardless of whether they have the same history or not. Upon further thought, I realize that "cognitively identical" was not a good choice of words, because it might be that one is already including moral obligations under that concept. But that terminological issue can be resolved as follows: Bert and Boltzmann Bert are identical with respect to any description of their minds/brains at the specified time (i.e., not including descriptions of what happened earlier) not involving moral language, as well as identical in terms of physical capacity, looks, DNA, etc.

So, my impression is that they also have the same moral obligations, but you raise a very interesting point about the traction of the argument.

My intuitive assessment still is that the fact that their brains/minds are presently the same with respect to any non-moral terms description trumps considerations about their previous behavior, but if you think otherwise, it may be that my intuitions on the subject are unusual - this is just tentative. On that note, the following variant might be relevant: Let's say that on a second such planet (say, B-Earth 2), his ex-girlfriend calls Boltzmann Bert 2 and tells him that he has to pick up his children. Do you think after she called him, he has the obligation to pick them up (just as Bert would), or that he doesn't (perhaps, because he didn't actually have those children, she wasn't actually his girlfriend, he didn't make any previous agreements, etc.), even if he mistakenly believes that he has an obligation to pick them up?

Two clarifications.

I went with Boltzmann Hert rather than H Bert in my reply to Angra because that was the last reference I saw. Error mounted on error. My point still should be clear enough.

One thing Katrina I should have said in my reply about insanity. While Hinckley (and M'Naughten even) clearly showed signs of premeditation for their acts, which is one necessary condition for 1st-degree men rea, the Model Code under which Hinckley was tried distinguished rational intent from non-rational intent. Only rational intent (substantial capacity to appreciate wrongfulness) is held under the MC as a necessary condition for guilt. I gather he was held NGBI for not satisfying that condition under the burden of proof by the prosecution to the standard of beyond reasonable doubt (namely that he was sane). The same issues and prosecutorial proof standards seem to apply in the Holmes case, and that's why I find it so interesting as a test of procedures and jurors' judgment.


I agree. Often it is hard to determine whether episodes of forgetting, misremembering, not realising, etc., are glitches or side-effects of normal functioning. If the lapse is recurrent, that's probably evidence of something wrong about the agent's moral outlook. "She forgets important stuff because she doesn't care enough," "he didn't realise he was messing things up because he tends to be insensitive to these sort of things," etc.

One off lapses, however, can fullfil the non-glitch condition too. In fact, they can do it without impugning the person's moral sense. That, at any rate, is the point of the paper with Doris. In such cases, attributions of blame are justified. But what justifies them is not the agent's will being morally defective.

Here's one possibility in the case you describe. (I hope I have the details right.) Bert was always there for his kids. He cared deeply for them, was a good dad, etc. Yet, his lapse was not a mere accident. It was, let's suppose, the consequence of his sloppy routines of time management.

Are these sloppy routines evidence of a moral defect? Perhaps they are: maybe Bert doesn't care much about honouring his commitments. But this need not be the case. Bert's sloppy way of managing his time might be evidence of a defect but not a moral one.

Would this work?

Goes without saying but I am enjoying very much your posts


What you suggest is a kind of version of the manipulation argument that puts pressure on compatibilist views insisting that a deterministic causal history is not MR undermining. Here is where many compatibilists about MR part ways, depending on whether you are an internalist or externalist about MR. Many philosophers require not only the right sort of mental conditions for MR, but also the right sort of causal history (viz. one that is presumably not undermined by the truth of determinism).

I would say (perhaps controversially) that Bolt Bert is not responsible because I think one is responsible for one's conduct only if (at minimum) it is the consequence of one's own causal will (necessary, but not sufficient). Thus, if we can't causally trace back a given lapse to certain blameworthy intentional states (e.g. reckless planning), then you can't be held responsible for that lapse. Which is to say, Bolt Bert didn't actually intend those (reckless) conditions responsible for his lapse, whereas Bert (ostensibly, assuming he is culpable) in fact did. So, if Bolt Bert forgets to pick up his kids, he's off the hook, but since he's now carrying out causally efficacious actions, he's henceforth responsible for future events that follow from his own willful effects. I hope to develop this view more fully soon.


What attributions of blame are you suggesting is justified in this case? Causal blame? Legal blame? The thing I would find most peculiar about the case is if it turns out that the tendency to cast moral blame is justified even if (hypothetically) we are unable to locate culpability in terms of conscious intent (whether in the moment or through tracing back in time). It seems to me that if we remove conscious intent from his prior poor planning behavior (so that it would not be due to something like the lack of care over one's commitments), then we remove moral blame. If, for instance, he has a disability wherein he simply isn't capable of managing his time in the way normal adults do, despite his genuine intentions to do better, then I would say he isn't responsible for his lapse.

Hi everyone, what a great discussion!

Alan, thanks for all of the clarification, very helpful. I've got to write a chapter on legal insanity this summer and no doubt I'll be emailing to tap your expertise (emails you are free to ignore, of course). :)

I love that Angra brought in Bolt Bert and Bolt Hert. On a purely synchronic capacitarian account, Bert and Bolt Bert have the exact same capacities and thus would both seem to be responsible for the lapse based upon some minimal capacity to understand the law and conform their behavior to it. However, I don't really like this sort of synchronic account of capacity, because, as Izzy notes, the lack of conscious intent seems problematic if we are looking only at the timeslice preceding the criminal harm. Instead, I think it makes more sense to think of the capacity as diachronic - Bert is responsible if he had this minimal capacity over time, because he knew his parenting schedule and could have performed actions over time to ensure he did not forget it (set reminders, etc). I think my quibble with Eddy is actually fairly minor - I don't think we need to find a conscious decision not to set a reminder, or any particular past decision, to hold Bert responsible. We just need to make a rebuttable presumption that Bert had these capacities over time. On this reading of the case, Boltz Bert did not have these capacities over time, and thus he would seem NOT to be responsible. Except...I agree with Alan that we may want to attribute Bert's causal history to Boltz Bert if BB is a swamp-man like identical of Bert.

Santiago, I still haven't had a chance to read your paper - hopefully this weekend! - but I'm really hoping your non-glitch condition maps onto my ideas of the conditions that would rebut the presumption of capacity.

Katrina and Izzy,

Given your replies and Alan's, it seems to me that my intuitions on the matter probably are unusual after all!
With respect to Izzy's point about Bolt Bert's resposibility for future events, it seems to me that on this account, Bolt Bert 2 - who gets a call from [the person he believes to be] his ex-girlfriend telling him to pick up the children - has a moral obligation to pick them up.
That seems to combine a a diachronic account of blameworthiness (so, Bolt Bert is not guilty for not picking them up), with a synchronic account of moral obligation (since Bolt Bert 2 has an obligation to pick up the kids even though he does not have the causal history of Bert, never actually had those kids, never made a previous agreement, etc.). Is that a correct intepretation?
Or does Bolt Bert 2 only acquires a moral obligation to pick them up when he gets the call from her?


My initial reaction is to say that Bolt Bert 2 only acquires a moral obligation once he gets the call. I think he only acquires moral duties once he becomes an agent or capable actor in the world and begins to create his own experiences. (This gives him the opportunity to take responsibility, as it were). So Bolt Bert isn't culpable for the lapse (since it was not caused, at some level, by his own conscious will), but both Bolt Bert and Bolt Bert 2 is (more or less) culpable for any future "poor planning" related lapses.

Izzy, thanks for the explanation.

I'm traveling right now (having attended a very nice workshop on the epistemic condition on responsibility at VU University Amsterdam, where teaching was getting special attention), and so I can't participate as I normally would. (It also partially explains my absence till now.)

First, I just wanted to thank Katrina for the careful presentation of my paper. It's always a nice surprise to see your work discussed (especially in a positive light!).

Second, I still find negligence, whether in there law or ordinary morality, to be in need of much more defense, especially when it's tired to sanctions as in Bert's case.

I have many thoughts on the topic, but since I can only use my phone to pay currently, I'll have to leave it at that. Katrina's done an excellent job presenting my main view, so most of what I might say would be repetitive anyhow.

I will just as a quick complication for those find of tracing. If you think there were conscious decisions in the past that help explain his failure to remember to pick up his kids, then Bert's conduct is arguably reckless, rather that outright negligent. If I realize that something I do now may increase the risk of something bad happening in the future and do it anyway, then if that risk is later realized, I can't quite claim I was unaware of the risk (even if at the time of the bad action I wasn't aware I was causing it).

Thanks again, Katrina, for engaging with this interesting topic (and for discussing my work)!

My phone thought it knew better. "Tracing" got special attention at the workshop.

It seems to me Bert can still claim he wasn't aware of the risk at the time. The thought that he's putting his kids at risk may never have occurred to him in the moment, since as far as he knew, his ex had them. This lapse in judgment, however, may have been caused by risky poor planning in the past. Which is to say, he's morally responsible for the lapse based on his actions in the past.

Now, of course, it may be the case that not only did he have a lapse, but he was aware that there was a chance he was mistaken about whether he had his kids, and that given his past, it would be wise to check with his wife before he left. This would certainly increase his level of culpability, but I still consider each of these examples to be instances of negligence. Reckless behavior just occurs to me as a kind of negligence. I am not sure a whole lot rides on the distinction, though.

You are right though that on this view culpability would be determined in either case by locating some conscious decision in Bert's past that lead to a negative outcome. If Bert managed to go through life without ever once being informed about the consequences of his actions by any parent, friend, acquaintance, educator, or relation (which I find highly unlikely given what we know about the case), then I don't think he's morally responsible. He couldn't reasonably have been expected to have done otherwise. Legal responsibility may be another matter, but insofar as it attempts to map MR, I don't think it can qualify without finding reckless conscious choices somewhere in the agent's past.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan