§1.0.0 – The Types of ‘Identity’ That Matter

It is a huge exaggeration to say, as Korsgaard does, that 'an obligation always takes the form of a reaction against a threat of a loss of identity'. I could remain me, both in the evident banal sense and in every pertinent non-banal sense, if I gave nothing to help the distant dying who oppress my conscience. I just wouldn't feel very good about myself. And I might even say, in morose reflection: 'how typical of me, to be so bloody selfish'. And I might lose my grip on myself if I suddenly found myself being very philanthropic. So, this is not Korsgaard's point that my identity is solid enough to withstand a measure of wrongdoing that contradicts it. This is the different point that plenty of what I do that I regard as wrong does not challenge my identity at all.

— G.A Cohen, in response to Korsgaard, in Korsgaard’s ‘The Sources of Normativity’ (1996), page 177

§1.1.0

I assumed that something like the view I am about to elaborate would have already been said, and that I would pretty quickly run into the view if I read a bit of the secondary literature on Kant. However, if someone like G.A. Cohen is saying something like this as late as 1996, then maybe the sort of view I hold is not as widespread as I thought.

Cohen describes that upon reflection on a selfish action, he might be annoyed at himself, thinking: ‘how typical of me, to be so bloody selfish’. Cohen views this as an example of how doing something wrong is not in contradiction with his identity—in fact, something he views as a wrong behavior can be a “typical” sort of action for him and thus apparently a part of his identity. Therefore, he thinks it is incorrect for Korsgaard to say that ‘an obligation always takes the form of a reaction against a threat of a loss of identity’.

§1.2.0

The important difference here is that Korsgaard and Kantians here understand ‘identity’ to mean something like ‘identity as a free-willing, rational agent’, while Cohen is taking ‘identity’ to be ‘a matter of loyalty and identification’ about things both general and particular. (175, italics are my added emphasis) Practical identity, to Cohen, can include loyalty to general moral principals, but it can also include loyalty to particular people, like specific friends or family. It can include one’s identification with being a free-willing agent, but Cohen says that his identity is also importantly defined in terms of his identification with being Jewish, and with being a Fellow of All Souls College at Oxford. Cohen does not think that it makes sense to say that general moral principals or identification as a free-willing agent can necessarily take priority over the more specific loyalties/self-identifications when defining one’s practical identity: “...no credible characterization of what practical identity is, in general terms, would yield a general priority for principled over particularistic identifications”. (175)

I think, however, that there is a ‘credible characterization’ of practical identity as giving moral primacy to one’s ‘identity as a free-willing agent’. I will make that argument now.

§2.0.0 – A Sellarsian Account of the Concept of ‘Self’

My argument is of the sort that Sellars makes in Empiricism and the Philosophy of Mind (1956) when describing how even the concept of something’s being a color is dependent on a few other concepts. For example, <being green> only seems to make sense to talk about when it is possible to frame it in reference to other basic concepts such as <looks green>, <standard observers>, and <standard viewing conditions>. Similarly, I think the way we use the term <self> has sense by existing in relations to other concepts such as <freedom from nonstandard influence/coercion>, <observations>, and <actions>.

Who are the ‘standard observers’ who can correctly say that something ‘is green’ when it looks green to them in standard viewing conditions? What are the standard viewing conditions? —As for one’s conception of ‘self’, what counts as nonstandard influence or coercion? These things are in large part socially determined, and in other ways, shaped by a long, historical, evolutionary process.

§2.1.0

Here is how I think the concept of <self> exists in relation to those other concepts of <freedom from nonstandard influence/coercion>, <observations>, and <actions>. When reflecting upon our ‘self’, I think we recognize that we make observations, and also seem to have some sort of direct control over the actions of our body. That is, there seems to be a process which takes in inputs and translates them into our outputted actions. This process is our ‘self’, and some boundaries to this process which we identify with seem to be defined in part in contrast to things in the world that we consider to be nonstandard influences / coercion.

§2.1.1

Sometimes the process of translating inputs to outputs seems incorrect in one way or another. In an extreme example, we might imagine an evil brain surgeon sending unwanted electrical signals directly into someone’s brain to make their arm move. Their body is taking in the inputs of the surgeon’s electrical signals, and is outputting actions—and this seems like it is not attributable to the person’s true self. The body is receiving inputs, and outputting actions, but the method of receiving inputs seems improper. An example of a ‘proper’ method of input, for contrast, might be light entering someone’s eyes. Therefore, this evil-brain-surgeon-process is not seen as a valid case of the normative, moral self’s decisionmaking process. What counts as proper or improper inputs seems largely socially constructed, although why things are socially constructed in this way are probably largely influenced through biological and cultural evolution.

Or, one’s decisionmaking might be influenced by one’s addiction to alcohol or some other drug. In these cases, a person might reasonably say that they are “not themselves” when acting based on reasons motivated by addiction.

Similar statements might be said by people regarding other mental conditions or states which they do not associate with their true selves. Even being extremely tired might make someone say, “Please don’t think the way I am acting now represents who I really am.”

There are also cases where people are misled or misguided by others, and upon reflection they may say something like: “If I were placed into that situation without that coercion or misguiding influence, I would have acted better. Please don’t think I actually am the sort of person who you saw back then out of context.”

There is a sense, of course, in which the person is the same physical being who performed those actions. Nevertheless, it is reasonable to say that there is also a sense of ‘self’ or ‘personhood’ which takes into account what we consider the standard ways of getting input (e.g. not from an evil scientist’s brain electrodes or other misguiding influences); and standard ways of making decisions based on inputs (e.g. not under the influence of addiction, or being in a standardly accepted state of mind for proper decisionmaking); and standard ways of acting (e.g. just because I thought about opening a door to the room I am in, and the door opened, doesn’t mean that I was the one who acted to open the door). I am not denying that there is a range of acceptable variation in these standards. I am simply claiming that there are some ways of ‘interpreting the world and translating that into actions’ which clearly seem non-standard and unrepresentative when we want to identify ourselves or another person in moral, normative conversation.

§2.1.2

This helps us clarify that a person’s normative self—including one’s own—necessarily seems to possess the quality of being ‘a process of receiving stimuli and translating those stimuli into actions, in a way which is free from unacceptable influences or coercion’. This, I believe, is usually the concept of ‘self’ or ‘identity’ we refer to in normative conversations: a free-willing agent.

Note: There are likely also other qualities which are important to our concept of a normative self/personhood, like the quality of being able to ascribe ‘intention’ to a person’s actions. That is, a person should seemingly act with a ‘reason’, or ‘toward an end goal or goals’. With this additional quality in mind, we might understand ‘free-willing agent’ to more precisely mean a ‘ free-willing, intentional/rational agent’.

§2.2.0

This sort of ‘true, moral self’ usually appears when someone has the time and calm mental state to reflect—i.e. in those situations in which they are free from nonstandard influences on their decisionmaking process. Thus, we get authors’ emphases on reflection, and the common identification of ‘a person in a reflective state’ as being mostly representative of their true ‘moral self’.

Korsgaard points out in her ‘Reply’ to Cohen and others that:

Kant first establishes that in so far as a person is active—'in respect to that which may be pure activity in himself—he regards himself as a member of the intelligible world. The moral law is the law of the intelligible world, and Kant argues that it binds the person because 'it is only as intelligence that he is his proper self.’ … The idea of identifying normatively with a certain conception of one's nature—the conception of oneself as active and rational—therefore plays a central role in Kant's view, just as the idea of identifying normatively with the conception of oneself as human does in mine. (237-238)

I think there is great potential here to examine possible overlaps between Kant’s discussion of “the intelligible world” and Sellars’ discussion of “the space of reasons”. I expect this has already been done to some degree. I hope my discussion here of Kantian ‘identity’ in the framework of Sellars’ ‘space of reasons’ have shown how fruitful this framework can be.

§3.0.0 – Threats to Identity & Reactions Those Tend to Engender

I have argued that in normative conversations, we usually consider a person to necessarily possess the quality of being a free-willing agent. Other, more particular qualities of a person, like their identity as part of a social group or their loyalty to friends, are less necessarily a part of their identity within moral conversations. One may say that their identification with a social group is not relevant to their decisionmaking process, but it seems impossible to deny that a person’s being a free-willing agent is a necessary assumption of their moral decisionmaking process.

§3.1.0

Viewed in this light, I think the example Cohen gives of being ‘typically bloody selfish’ is actually a pretty perfect example of what Korsgaard is trying to show: that moral obligations arise to protect a person’s identity. Upon reflection, a person annoyed with themselves for their selfishness is recognizing that, in those past actions, they had subscribed to selfish ethical principals. They are annoyed with themselves for doing so because in this reflective state, they do not identify with or endorse such selfishness as relevant to their moral decisionmaking. This is even possible while they recognize that, outside of this reflective state, selfishness is a common occurrence for them—that it is ‘typical’ of them.

§3.1.1

The person is essentially annoyed and ‘morose’ because their broader actions and reasons for acting had not aligned with the principals held by their ideal, reflective self. They appear to lack free-willed control over their actions: apparently something is improperly getting in the way between the ‘reasons/goals/maxims of their ideal, reflective self’ and ‘acting on those maxims in the rest of their lives’.

As I argued above, being a free-willing agent is a core part of people’s identity as a moral being, essentially by-definition. One cannot be ‘moral’ in any practically understandable sense unless one has the capacity to freely choose to act morally instead of immorally. Insofar as someone wants to talk of themselves in moral terms, it seems that they must identify as a free-willing agent in some meaningful sense.

§3.1.2

The moral obligations / duties we impose on ourselves (e.g. not to be so selfish) are attempts to take the maxims held by our ideal, reflective selves, and to impose those throughout the rest of our lives. For example, on reflection, a person may come to the conclusion that they have been acting in ways that they believe are unacceptably influenced by a desire for alcohol and/or their regular consumption of it. They may then say: “This addiction is not who I am!” and go on to set normative duties for themselves to take actions to free themselves from the improper influence of addictive cycles. By acting in accordance with one’s self-imposed duties, a person affirms their identity as a free-willing agent in every moment of their life—and not merely in their moments of self-reflection. The obligation one feels to these duties is essentially one’s desire to affirm one’s identity as a free-willing agent.

The person who is morose or annoyed with their own selfishness does not morally identify with their excessively selfish actions. They reject such extreme selfishness as something that is not a part of their proper decisionmaking process which they identify with in their reflective state of mind. Clearly, they would not think they are obligated to be selfish. They may then set for themselves the obligation/duty not to act so selfishly, because doing so would be in accordance with their not-so-selfish normative identity. This duty is ‘a form of a reaction against a threat of a loss of identity’, as Korsgaard claims it must be.

§3.2.0

I also want to address Cohen’s point that many particular things comprise his identity. While I still think that a person’s identity as a free-willing agent is necessary to their moral identity in ways that most other things are not, there are still some other aspects to a person’s identity which might be morally relevant to them upon reflection. Cohen may give examples here of his Jewish identity, or a person’s identity as a husband to his wife. Upon reflection, when I am most myself, I might still find it important to me to identify as a person who is deeply interested in learning and contributing to philosophy. A violinist might identify as a person deeply interested in music and developing their skills in playing the violin. These parts of their identity may be a source of obligations to them. On a day when they have free time, the violinist might feel lazy, but still feel some sort of duty to themselves to practice the violin or otherwise act in ways that are consistent with improving their violin skills in the long run.

These obligations which come out of contingent parts of people’s identities are not as important to a person’s moral decisionmaking as those obligations which come from necessarily being a free-willing agent. These contingent obligations are important nonetheless, and usually not in conflict with the necessary obligations. If it is discovered that these duties are in conflict, then when faced with certain situations, a person’s ethical system would be suggesting two different, incompatible actions. To morally decide what to do in such situations, it seems like the person would have to jettison one of the parts of their identity as being ‘less morally relevant’ than the other. It is impossible to jettison the necessary part of one’s identity as a free-willing agent, so that will always take priority over the other contingent parts of their identity. It would make no sense to say that you are making moral choices as a violinist, but not also as a free-willing agent—that you can choose how to act at all implies that you are a free-willing agent.

§3.3.0

Finally, the last paragraph of §3.2.0 raises concerns to me about choices which are seemingly made by us, but not as free-willing agents. This seems impossible—if we weren’t free while acting, then it wasn’t our choice. I agree that it is impossible, but I think there is an important related ambiguity that deserves discussion. This section, §3.3, is a brief discussion of that.

There are processes which are arguably a necessary part of ourselves, which take in inputs and result in actions, but which may not appear to be free choices. For example, when I touch a hot stove, my hand instinctively flinches away. I might be able to overcome that with forethought, but other similar evolutionary presets are less simple. Cognitive biases and optical illusions seem to necessarily be a part of our selves—they seem unavoidably part of the process which translates inputs into outputted actions. However, it also seems reasonable that someone might reject these parts of their decisionmaking process as unrepresentative of who they consider themselves to be morally.

§3.3.1

Similar to how, upon reflection, someone might not identify with ‘excess selfishness’; a person might also not identify ‘confirmation bias’ as a proper part of their moral thinking. Nevertheless, it seems like instincts to confirmation bias might be more ingrained than selfish personality traits. It seems reasonably possible to reject and unlearn selfishness by imposing duties on oneself, but the same might not be possible with things like confirmation bias.

However, there seems to be a difference between the necessity of being a free-willing agent and the necessity of things like ‘having cognitive biases’ or ‘being affected by optical illusions’. It seems logically necessary that we are free-willing agents insofar as we are moral. In contrast, it merely seems materially/evolutionary necessary that things like cognitive biases and optical illusions are parts of one’s decisionmaking process.

§3.3.2

If it is impossible to rid oneself of these aspects of the decisionmaking process, then they may just so happen to take priorities within one’s decisionmaking similar to that of the priority of one’s identity as a free-willing agent (as described in 3.2.0). However, if technological advancement allows oneself to be free of these unwanted, evolutionarily built-in styles of decisionmaking, and a person does not morally identify with those evolutionary presets, then they may take actions to free themselves from those processes which they consider to be improper influences of evolutionary history. One might, for example, establish counteracting institutions or feel a duty to follow certain decisionmaking heuristics as to curb the ultimate effects on one’s actions of unwanted cognitive biases. At the most extreme, we might even imagine people inventing cybernetic brain implants to override unwanted cognitive biases. People might even identify more with these technologies than with their own brains.

Technological advancements seem to have the capacity to shift what appear to be materially necessary influences on our decisionmaking to be closer to merely contingent influences, thus creating more capacity for free choices. It might be better said then that these influences were always contingent, but that in most peoples’ judgements, the effort needed to free oneself from those influences were too great, and the outcomes too uncertain, for them to dedicate most of their life to attempting to overcome those influences. A person may want to practice violin more than they want to research heuristics and cybernetics, and that seems perfectly acceptable. However, in doing so, they might be practically admitting that they are accepting such material influences on their decisionmaking as valid parts of their moral self—at least, to the degree that they are not rejecting them as much as they could be.

§4.0.0 – Korsgaard’s ‘Reply’

The most relevant sections for this post in Korsgaard’s ‘Reply’ are Sections 2 and 3 (234 – 242).

Near the end of Section 2, she describes how we identify ourselves morally. Kant’s view is that people normatively see themselves as active and rational, and Korsgaard’s view is that people normatively see themselves as ‘human’. These are aspirational and identificatory, and a source of justification for normative self-legislation.

I also already discussed in §2.2 how Sellars’ ‘space of reasons’ seems like it can be fruitfully used to explicate Kant’s notions of what is ‘intelligible’. The sorts of reasons we morally appeal to in Kant seem to have their place in Sellars’ space.

In Section 3, Korsgaard discusses how we might handle the sorts of desire that are in-part largely influenced by our instincts. “Our contingent practical identities are, to some extent, given to us—by our cultures, by our societies and their role structures, by the accidents of birth, and by our natural abilities—but it is also clear that we enter into their construction.” (239)

She says that, “Part of my intention in invoking the concept of practical identity is to break down Kant's overly harsh, and even in his own terms oversimplified, division between natural impulses that do not belong to my proper self and rational impulses that do.” (240) This nuance, I think is also accomplished in my discussions in §3.2 - §3.3 which allow for someone to be driven, for example, by a partially impulsive desire to enjoy developing their skills in the violin. It is not a totally rational thing to want to learn violin as a hobby, but it nonetheless can be an important and valid part of one’s normative identity.

The violinist of §3.2 actively chooses to practice the violin on a lazy day in a way that represents their normative self—they do not simply passively enjoy playing the violin whenever they have the basic desire to play. Korsgaard describes something similar too: “But I think that it is essential, if we are to get this right, to distinguish our attitudes towards contingency from our attitudes towards passivity. For contingency itself is something that may either be actively embraced or passively endured, and this makes all the difference: the mature attitude is the one that actively embraces it, not the one that passively endures it.” (242)

I mention in §3.1.2 that “By acting in accordance with one’s self-imposed duties, a person affirms their identity as a free-willing agent in every moment of their life—and not merely in their moments of self-reflection. The obligation one feels to these duties is essentially one’s desire to affirm one’s identity as a free-willing agent.” I think this is getting at a very similar idea as to when Korsgaard says that “...at the moment of action I must identify with my principle of choice if I am to regard myself as the agent of the action at all.” (241) A person acting to follow a duty—to act according to their normative self—is affirming their identity as a free-willing agent. Otherwise, they are still acting, but not according to the sense of self with which they identify with in reflective moments. Furthermore, upon reflection, they might go on to say they lacked free-willed control over their actions in the way I described in §3.1.1, and that they weren’t truly the agent of those actions in the normative sense that matters.