Neuroethics Symposium on Tobia’s “Personal Identity, Direction of Change, and Neuroethics”

Welcome to our second Brains Blog symposium on papers published in the journal Neuroethics. Our target paper for this symposium is Kevin Tobia’s (Yale University) “Personal Identity, Direction of Change, and Neuroethics.” Below you will find an introduction to the symposium and brief précis of the paper, followed by commentaries written by William Hirstein (Elmhurst College), Jesse Summers (Duke University), Sarah Molouki (University of Chicago), and Jennifer Rowe (independent scholar). Kevin has also provided responses to the comments and a list of recent papers on experimental philosophy of the self and personal identity.


The central finding of Tobia’s paper is that ascriptions of personal identity – whether a person is the same person over time – are not just sensitive to how large the change in a person is, but also the direction in which the change occurs. Tobia’s data seem to show that even very large changes to a person’s personality or character may be identity-preserving if deemed positive (e.g. if they result in a better or more virtuous person), whereas relatively smaller negative changes may erode ascriptions of personal identity (e.g. if they result in a worse or vicious person).

Initially, one may wonder why this would be so. Aristotle indicates that developing both virtuous and vicious traits requires habituation, where choices establish traits that can be seen as fairly stable dispositions to act. From this perspective, virtuous and vicious traits seem symmetrical, requiring effort to establish and maintain. Changes equivalent in magnitude to either sort of trait should impact personal identity in a similar way.

But it seems that both the development of vicious traits and the dissolution of virtuous traits (due to, for example, incontinence) would be labeled negative personality changes on Tobia’s view; and here an asymmetry emerges. While virtuous traits and action only emerge via effortful habituation, poor behavior may be the result of either effortful habituation or by incontinence – a person being blown about by temporary desires for pleasure. A previously considerate person may be deemed “not herself” if she starts doing whatever strikes her as fun in the moment, even if such acts are not an expression of an established trait of selfishness. William Hirstein makes this point in his commentary, noting that the famous Phineas Gage discussed in Tobia’s paper might have become disinhibited. Hirstein claims this new Gage may not be the same person as the old Gage, but claims this is because new Gage is lacking in rational capacities vital to personal identity, usually housed in the prefrontal cortex.

I, like Tobia, see danger lurking in the asymmetry of the folk attributions of personal identity. Although Jesse Summers wonders in his commentary whether personal identity may just be a philosopher’s concept, invented to explain a wide range of folk concepts and practices, I agree with Tobia that there seems to be a fairly stable folk notion of personal identity underlying many moral and legal decisions. If we let our notions of what sort of character traits are “best” taint our sense of the extent to which cognitive interventions change identity, we might, for example, be more willing to allow the state to perform interventions on criminal offenders in the name of rehabilitation. A handful of US states and European nations already allow involuntary chemical castration of sex offenders. Virtual elimination of sexual desire from a sex offender constitutes a fairly large change in his character, one that might be seen as preserving identity due to its possible positive effect. However, most sex offenders who qualify for chemical castration under state statutes do not only desire illegal targets (e.g. they are not pedophiles). Removing the capacity for sexual desire removes an important aspect of a persons’ identity regardless of whether doing so may have some positive effect for society at large, and this impact is good reason to see chemical castration of offenders as unethical. From an Aristotelian perspective, removal of sexual desire also removes the possibility an offender can develop virtue with regard sexual partners – that is, that he can learn to express the right amount of sexual desire for the right partner in the right circumstances.

Another worrying example might be a prisoner who transform from violent, hostile person into a docile child-like one due to early-onset dementia (this is fairly common in imprisoned offenders). In this case becoming incontinent may certainly be perceived as an improvement over an earlier vicious character. But even if the change in the offender’s temperament is in some sense positive, this ought not to distract us from the fact that the demented offender may no longer be the same person who committed the crime for which he is being punished, especially if he no longer remembers committing the crime.

Sarah Molouki’s commentary specifically discusses Tobia’s findings with regard to advance directives. Molouki notes that advance directives are designed to allow a person to make binding decisions for his (fundamentally changed) later self. If advance directives don’t apply in cases where the person is “no longer themselves,” then they never apply. Jennifer Rowe questions in her commentary whether it matters if the subject who has undergone a major change to their character is aware of this change: it might be that the folk would be more likely to judge new Gage the same as old Gage if new Gage consciously embraced the changes in his character.

A big thank you to Neil Levy, editor of Neuroethics, all of our commentators, and Kevin Tobia for a great target paper and responses to commentaries.



Overview: Personal Identity, Direction of Change, and Neuroethics

This paper discusses the personal identity relation and how intuitions about personal identity might underlie a variety of moral and legal judgments. It focuses particularly on the effect of improvements and deteriorations on personal identity attributions. Experimental work suggests that personal identity attributions are affected not only by the magnitude of change, but also by the direction of change; all else equal, changes for the better seem more identity-preserving than changes for the worse. The effect of direction of change on identity attributions may drive a number of other moral, legal, and neuroethical judgments, affecting the way in which we evaluate (e.g.) neuropsychological and other cognitive enhancements, cases of dementia and other apparent cognitive declines, and other practical decisions (e.g. the applicability of an advance directive).


Click here to view the target paper.

Commentary by William Hirstein, and Tobia’s response

Commentary by Jesse Summers, and Tobia’s response

Commentary by Sarah Molouki, and Tobia’s response

Commentary by Jennifer Rowe, and Tobia’s response

Click here for selected recent papers on experimental philosophy of the self and personal identity (curated by Kevin Tobia)


  1. Kevin Tobia

    As a first comment, I would like to express my thanks to all those involved in this symposium. For the invitation, great thanks to John Schwenkler and The Brains Blog, and especially Katrina Sifferd, who also organized the symposium. Thanks also to Neil Levy, Neuroethics, and the paper referees. Finally, I am extraordinarily grateful for the expert commentaries. Thanks to the authors of the four commentaries, William Hirstein, Sarah Molouki, Jennifer Rowe, and Jesse Summers, for insightful, challenging, and thought-provoking comments. I look forward to further discussion of these and related topics.

  2. Sarah Molouki

    I echo thanks and great job to Kevin, organizers, and commentators!

    One thing that came up in several of the commentaries/replies that I found interesting was the idea that certain types of change might be believed to uncover a core or true self and thus preserve identity, (also explored in the work of Newman et al., as Kevin points out). Although the true self is often thought to be positive, William’s commentary highlights the idea that disinhibition is often viewed negatively and thought to lead to loss of the self. Since a disinhibited person is likely to act upon their most fundamental and core desires of all (e.g., hunger, thirst, sexual desire, etc.) it is interesting that the disinhibited self is not in fact commonly viewed as the “true” self. It seems that what defines personal identity does not in fact lie at the level of these “deepest” motivations (one reason perhaps being because they are common to everyone and would thus not serve much of a purpose in differentiating individuals). Rather, identity seems to reside at a “higher” level of personality where individual intentions can lead to differential expression.

  3. Kevin Tobia

    Thanks Sarah!

    These are all intriguing issues you raise. It seems that depending on the case, inhibition and disinhibition might each be viewed positively or negatively. The commentaries here provide some suggestion that disinhibition is in some cases viewed negatively (I think in the technical sense, disinhibition is most often viewed negatively), but in some other cases becoming less inhibited might be viewed positively [“the happy drunk”, maybe also disinhibited disregard of BAD social conventions]. Similarly, although inhibition is sometimes viewed negatively, other times inhibition may be viewed positively, as representing an improvement or enhancement (see, e.g., B. Earp, A. Sandberg, G. Kahane, and J. Savulescu. 2014. When is diminishment a form of enhancement? Rethinking the enhancement debate in biomedical ethics, Frontiers in Systems Neuroscience 8(12): 1-8). I’d be very interested to see some more empirical work directly on the effect of apparent inhibition and disinhibition, each viewed positively or negatively, on identity intuitions. My own (unsurprising) hunch is that any effect of dis/inhibition on identity attributions is going to depend a lot on the positivity/negativity, whether the dis/inhibition appears to lead to an improvement or deterioration. But maybe others hypothesize something different?

    Sarah, this last point you raise about intention is a very interesting one. I agree that some of the recent research suggests the relative unimportance of low-level cognition and importance of personality and morals to identity, but Strohminger and Nichols (2014. The essential moral self. Cognition 131) also report that memory, “especially emotional and autobiographical memory, is fairly important.” So I wonder whether the distinction isn’t exactly about what traits are “intentional.” That said, this certainly seems worthy of study. I’d be interested in seeing is some follow-up work to the essential moral self work looking at intentional vs. non-intentional traits (e.g. is the apparent importance to the self of intentional trait X different from that of non-intentional trait X?).

  4. Katrina Sifferd

    Hi Kevin and Sarah

    With regard to disinhibition, one interesting case is discussed by Burns and Swerdlow here:

    Burns and Swerdlow’s patient started to exhibit pedophiliac behavior quite late in life (in his 40s). After some investigation, a tumor was discovered in his orbitofrontal cortex and removed, and the man’s pedophilia then seemed to disappear. Later, the man once again starting exhibiting symptoms of pedophilia, and doctors discovered the tumor had grown back.

    In the literature many speculated that the man had probably always had deviant sexual desires, but had used executive control to inhibit them. The tumor disrupted this inhibition and thus the deviant desires were expressed. This sort of story seems to privilege top-down control mechanisms as central to our identities: we all have unsavory desires, especially in certain contexts, and we don’t act upon them because we don’t think of the desires as expressing our “true selves.” If the executive processes in the prefrontal cortex are what allow us to be moral agents who can screen our behavior based upon our sense of who we want to be (or, in philosophy-speak, agents who are sensitive to moral considerations and can control our behavior in response to these considerations), it would seem these processes are central to human identity — MORE central than our desires themselves.

  5. Vilius Dranseika

    Very interesting symposium.

    I’m not sure, however, what to think about the assumption that the magnitude of change is held constant across the Gage vignettes. I have a vague hunch that there is something teleological about folk perception of moral character of human beings – expectation of moral improvement seems to be a default option. If this is the case, perceived magnitude of change can be bigger in moral deterioration than in moral improvement (i.e. direction of change influences perception of magnitude of change and this in turn influences perception of identity).

    It would be interesting to hear any reactions by the participants of the symposium to this little argument sketch.

  6. Kevin Tobia

    Vilius, thanks so much! I actually share your hunch that at least part of what is happening here is something teleological. I wonder if that would go even further than just moral character. Consider, e.g., language ability. Say individual A is a two year-old, speaking in one or two word sentences. Years later, we look at twenty-two year-old individual B and are dazzled by his striking language ability. A and B are very different in this respect, but I’d predict that this (in itself) does not shake our confidence that A and B are the same person. In contrast, imagine a case where a person with striking language ability is replaced over many years by a person who now speaks in one or two word sentences. I’d predict that, compared to the former scenario, this one would elicit weaker judgment of persistence. Perhaps part of what explains such intuitions is teleology: twenty-two year-olds are meant to have radically different language abilities from two year-olds (*different in a very specific way*).

    I’m working on some experiments to test this teleological interpretation of the direction of change effect. But it’s also hard to tease apart teleology and improvement. In the case I just gave, it seems equally plausible that persistence intuitions are driven by a general sense of improvement. But I think that in the cases where these do come apart, persistence intuition seems to go with the teleological option, not the improvement one.

    Another way to put it is that it seems (at least to me) that not all improvements intuitively preserve a person’s identity; it seems to be specifically the “teleological” ones that do. Some great improvements would break personal identity. In Aristotle’s words (NE 1166a19), no person would “choose to have everything on condition of becoming other – as, for example, the god right now has the good. He will wish to have the good while continuing to be whatever he is.”

    This other suggestion you raise about direction of change affecting perceived magnitude of change is also very interesting! I think it’s definitely worth testing and thinking about further. I read your comment above as suggesting that if direction of change affects magnitude of change, this lends some support to a teleological interpretation. I’d be very interested to hear more detail about that proposal!

  7. Vilius Dranseika

    Thanks for this interesting comment, Kevin. I very much agree with your exposition.

    Concerning my proposal that direction of change (that is, whether the change is in accordance or in opposition to the direction of teleology in question) affects perceived magnitude of change, it is purely speculative now. VERY roughly, the idea would be, by analogy with mechanics, that more force is needed to reverse the direction of movement of an already moving body and make it move in the opposite direction with a given velocity than to accelerate the body on its former path to the same velocity.

    A couple of months ago I’ve run a little pilot study with your Gage design and added several extra questions. One of the questions was something like this: “How big, in your opinion, is the personality change described in this scenario?” (7 point Likert-type scale anchored at endpoints: “Very small change” and “Extremely big change”). Ratings were very high in both improvement and deterioration conditions and any possible variation seemed to be masked by a ceiling effect. It would be interesting to think about ways to improve the scale and/or to construct a somewhat “softer” version of Gage story in which changes are less drastic.

  8. Kevin Tobia

    Hi Vilius, thanks for the reply!

    This explanation helps a lot. I really like this view! I suppose one small challenge comes from the case of unusually little change over time. Imagine a person (say a two year-old) grows over twenty years and does not change AT ALL in the sorts of psychological properties we think would normally change in that time. If I understand it correctly, there should be an apparent teleology in a certain direction (changing in specific ways from two to twenty-two), but in this case there would seem to be some force that has prohibited or hindered such change. Nevertheless, I’m not sure people would intuit that the “magnitude of change” (between earlier and later individuals) here is especially large, even though it indicates a force very strongly opposed to the apparent teleology. I’m not sure how people would respond to this example, but even if this hunch is true, this might just represent a small limitation to the theory.

    It’s great to hear you ran that study! I also agree with the suggestions you make. Perhaps you could use a much bigger scale (e.g. 100 point scale rather than 7 point scale). That seems a reasonably likely way to find some difference if it exists. A “softer” version of the story might also work, although I worry that going much softer may result in a floor effect in terms of the persistence question (nearly everyone rating strongly that the two individuals as identical).

  9. Vilius Dranseika

    This stalled development objection seems to be a difficult one. If change is to be expected, no change at all can constitute a big abnormality (defined as such in the light of previous expectations) but it still may be weird to call the emergence of this abnormality ‘a change’. Thanks for pointing this out.

  10. Jennifer Rowe

    Hi Kevin and Katrina,
    Katrina, thank you for the link to the discussion of Burns and Swerdlow’s patient, an intriguing case. It seems to be a real-world example of a case that parallels a hypothetical case toward the end of my commentary in which Phineas has become cruel, is aware of this and agrees that it is a bad thing, and laments not being able to act otherwise. Burns and Swerdlow’s patient was aware of the change he underwent, and further was aware of it as being a deterioration: he “could not refrain from acting on his pedophilia despite the awareness that this behavior was inappropriate.” There is also evidence that might be interpreted as his disavowal of his condition, or a strong desire to not be in this condition (confessing a fear that he would rape his landlady seems to indicate a desire to avoid this behavior rather than to commit it). I suggested that in these sorts of cases of deterioration, awareness, and disavowal, we might be inclined to judge the agent’s identity persistent, because his disavowal of the deterioration shows that he adheres to the same cultural norms that he shared with us prior to the change, whether or not he is able to behave in accord with them. Burns and Swerdlow mention that “acquired sociopathy occurs with adult-onset damage [to the orbitofrontal cortex], but previously established moral development is preserved. Nevertheless, poor impulse regulation leads to bad judgment and sociopathic behavior.” I think there is an empirical question to be answered regarding whether our judgments regarding an agent’s identity and his moral responsibility come apart. To me, at least, it seems the patient is the same man with a terrible new behavioral appendage – he has behavioral tendencies which can practically be resected right along with the tumor (although, unfortunately, his and others’ memories of the consequences of the behavior cannot). And yet, when asked whether he should be in a position to make decisions about the care of his stepdaughter or perhaps even decisions about his own treatment, I’m inclined to say that despite his identity being persistent, he nonetheless should be denied the right to make those decisions. His acknowledgement of our shared norms shows that he has the potential and perhaps even the desire to behave in the same ways that those norms would suggest, but his potential inability to control his behavior makes it impossible to count on him to act in accord with those norms. I am willing to say that the patient is the same person that he was prior to the tumor, but that his new condition (which is that the tumor could always reappear and influence his regulation of his behavior) should preclude him from having certain rights and responsibilities that he had prior to developing the tumor. The patient’s moral knowledge is intact, but the executive control needed to act in accordance with that knowledge is damaged. So perhaps retention of moral knowledge and other cultural norms is more bound up with identity attributions, and the executive control processes are more bound up with one’s status as an ethical agent. In other words, there seem to be two separable notions packed into the idea of moral traits: moral knowledge, and the ability to act, or control one’s behavior, in accord with that moral knowledge. Certain cases, like Burns and Swerdlow’s patient, seem to show that moral knowledge and the ability to deploy that knowledge do come apart. I would suggest that in these cases, identity does not track the control processes, but rather the retention of moral knowledge. Something like moral responsibility tracks the proper functioning of the control processes.
    I’m trying to imagine some cases where things are reversed: the executive processes are perfectly functioning, but one’s moral knowledge has been drastically altered (perhaps something like: a person joins a cult whose corpus of moral knowledge is vastly different from that of society, and he replaces all of his former moral knowledge with the moral knowledge of the cult. His control processes are properly functioning, allowing him to control his behavior perfectly in accord with these new cultural norms). Despite the functioning of the control processes, would the change in his moral knowledge be more of a threat to identity than the condition of Burns and Swerdlow’s patient? It’s my impression that it would, but I am curious to know what Katrina and others think.

    Kevin, these thoughts seem consistent with the results in Nichols and Strohminger (2014) paper that you cite, and it is certainly relevant here, but this comment is already running long! I’ll just say briefly that I’d be very interested in any thoughts you may have about whether forensic identity can – or should be allowed to – come apart from moral responsibility. I guess I want to push this intuition that there are cases of severe deterioration in which we still might judge identity to persist (but perhaps we judge that the person later has lost certain rights that he had earlier), and this case that Katrina referenced – one in which the patient retains moral knowledge but sustains damage to his control mechanisms – strikes me as being a candidate for such a judgment.

  11. Kevin Tobia

    Hi Jennifer – very intriguing post! And thanks Katrina for raising the Burns and Swerdlow case!

    I think I share some of your intuitions about these cases, although I’m not yet sure what interpretation I favor…

    One very small point: I think the role of culture / cultural norms seems highly relevant here (this is a feature you also highlight in your commentary). In the post above you write, “I suggested that in these sorts of cases of deterioration, awareness, and disavowal, we might be inclined to judge the agent’s identity persistent, because his disavowal of the deterioration shows that he adheres to the same cultural norms that he shared with us prior to the change, whether or not he is able to behave in accord with them.” My hunch is that this cultural norms piece is key. It seems less intuitive that persistence is preserved when one deteriorates massively, but the present person is aware of such and disavows it for *totally new and distinct* reasons. Maybe the original formulation can accommodate this anyway with a thicker notion of awareness (e.g. to be “aware of deterioration” requires not just awareness of present badness, but also the *specific kind* of old goodness) or disavowal (e.g. to “disavow change” requires not just disavowal of the resultant present badness, but also the departure from the *specific* older set of norms).

    As for whether personal identity can come apart from moral responsibility, I have an awareness of something like my own intuitive commitment to these being closely linked. In particular, I’m drawn to seeing personal identity as a necessary condition for the application of many moral and legal concepts, like responsibility (notable exceptions do exist of course, e.g., vicarious liability/responsibility). In Phineas Gage kinds of cases I think we’re owed some explanation by an account that posits non-identity AND responsibility (although I wonder whether some sorts of responsibility remain intuitive, even in the face of non-persistence intuitions).

    Your example is a case in which we intuit identity but not responsibility – I agree this certainly appears possible and it also seems to me an important and accurate insight. The middle of your post above raises a very difficult question about to which aspects of decision making this would apply: should such a patient make decisions about the care of his stepdaughter; about his own treatment? Surely there are still some decisions he may make? Drawing this line is a complicated problem. I suppose some candidates to explore are degrees of responsibility (rather than all-or-nothing), more finely individuated concepts of responsibility, or maybe some more complex account using the concepts you’ve mentioned (e.g. moral knowledge) and some others.

  12. Jennifer Rowe

    Hi Kevin,
    Thanks so much for your response! I’ll reply below, but first I want to congratulate you and the other commentators on a terrific job. I found your paper, the other commentaries and your replies, as well as the comments here, to be really absorbing.

    You wrote that “it seems less intuitive that persistence is preserved when one deteriorates massively, but the present person is aware of such and disavows it for *totally new and distinct* reasons. Maybe the original formulation can accommodate this anyway with a thicker notion of awareness…or disavowal…” I like the way you framed this case and agree with you that identity seems less intuitive here – I had not thought about a case of disavowal for the wrong reasons, or for totally new and distinct reasons, as you put it. I’ve definitely missed something in only having thought of a person’s disavowing a deterioration for the (obvious/right) reasons (as in the Burns and Swerdlow case), and thus thinking that this would stand as evidence for his continued adherence to the norms of his culture, then thinking about a drastic change in moral knowledge in separate terms (in the cult case, I was assuming that the agent *avowed* the changes for new reasons). Despite having been thinking about such a change, I had not worked it out in the disavowal language yet until you brought it up in these terms. The way you’ve described the case makes clear that there is disavowal of a deterioration, but it does *not* inform us that the agent still adheres to our norms. And you’re right that this requirement of adhering to our norms would have to be captured somewhere in order to account for such a case. Thank you for pointing this out and prompting me to think about it! (In an earlier draft of my commentary, I had used the phrase “awareness of the change as such” and for some reason later thought the qualifier unnecessary. The case you gave and your suggestion that a thicker notion of either awareness or disavowal could accommodate it makes me think that I needed it there after all. This is really insightful and helpful!)

    On the question of whether personal identity and moral responsibility can/should come apart, you write that “I think we’re owed some account that posits non-identity AND responsibility (although I wonder whether some sorts of responsibility remain intuitive, even in the face of non-persistence intuitions).” I think we might have such cases where an agent who was bad before has now unquestionably improved, but is lacking in awareness of any change. We may feel as though she is a totally different person; that we don’t “know her at all anymore” (but perhaps here it will be expressed with approval, as in “she’s like a new person!”). But this new person is a *better* person – in fact, if we had to choose who should be making ethical decisions, we’d frankly choose the new, better person over the pre-change person. The intuition to allow an improved person to make ethical decisions (as, for example, in your John vignette, improvement scenario) seems relatively uncomplicated by intuitions about a lack of identity when we believe that this person’s decisions will be more in accord with the decisions we would make – the behavior, including decision-making behavior, will better accord with our cultural norms, and we like that outcome, identity or not.

    Your work seems to show that the tendency to favor improvements as more identity-preserving than deteriorations is present in a wide variety of cases – not only in the empirical data about identity judgments in your target paper here, but also in (Tobia, K.P. 2015 Personal identity and the Phineas Gage effect. Analysis 75(3): 396-405). I was interested to read in that paper that you had tested Parfit’s Russian Nobleman case – a case in which there is slow (and presumably voluntary) change over time, very different from the Phineas and John cases – and got similar results. This bias toward the “improved” person as maintaining the same identity seems quite strong – strong enough, maybe, that we loosen our standards in ways like allowing “he’s finally his true self” sorts of stories to paper over a drastic improvement so that it doesn’t threaten identity . Trusting improved John to make ethical decisions seems as natural as trusting the positive Captain Kirk over the “impostor”, even when we still might have – or should have – lingering questions about their identities.

    I share your intuition of being committed to seeing personal identity and moral responsibility as closely linked. My concern is that we might indeed be so biased toward what we deem improvements that we give a pass to an improved agent, allowing him to make ethical decisions even where there is a genuine challenge to his identity, such as his being completely unaware that his behavior is different, or even, say, the sheer magnitude of it. No such pass is afforded a deteriorated agent; a challenge to his identity, such as his being unaware of any change, might easily overturn it. So I think your work shows that the direction of change effect and the bias it suggests are possibly strong enough to support conflicting intuitions about identity and responsibility in certain cases. If experimental results were to show this, though, I’m not sure I would take it to indicate that responsibility should or really does come apart from identity; rather it might indicate that something is wrong with the judgments produced under the influence of such a bias. All the more important that your research has drawn this bias out for discussion.

    Thanks, Kevin, for a really interesting and helpful discussion, and for your thoughtful responses to us all!

    [In my previous post I mentioned a paper by Strohminger and Nichols that you cited and I inadvertently reversed the authors’ names. My apologies for the inaccuracy! The citation is (Strohminger, Nina and Shaun Nichols. 2014. The essential moral self. Cognition 131: 159-171).]

Comments are closed.

Back to Top