This is a revised version of an IU Philosophical Society blog post.

At the Philosophical Society’s first spring meeting, I opened with a methodological point: Semantics matters. Misunderstanding is everywhere, and it is dangerous. If we don’t clarify what we mean, then we’ll never pinpoint where exactly we non-verbally disagree.

But the importance of semantics doesn’t mean we should fetishize which particular words we use. Just the opposite: In analyzing what we mean, we frequently discover that the world doesn’t neatly break down into the shape of our linguistic categories. We may have one word (“monkey”) where there are really two or three things, or two words (“electricity” and “magnetism”) that pick out the same phenomenon in different guises. Thus we talked about the value of “Tabooing your words”, of trying to find paraphrases and concrete examples for terms whose meaning is unclear or under dispute.

This is of special relevance to discussions of the self. People mean a lot of different things by “self”. Even if in the end those things turn out to be perfectly correlated or outright identical, we need to begin by carefully distinguishing them so that we can ask about their relatedness without prejudging it.

For example: David_Perry noted that many classical Buddhist texts denied the existence of a self. But what they actually denied was what they called ātman, which some people have translated as “self”. Even had they written in English, for that matter, it wouldn’t necessarily have been obvious which ideas of “self” they had in mind — and, importantly, which they didn’t have in mind.

What are some of the concepts of “self” that we came up with? I lumped them into five broad categories.

1. Thing

When we say “That’s an ugly coat of paint, but I like the house itself,” we don’t have the same thing in mind as when we say “I have a self”. It may seem trivial to note that objects in general can themselves be called “selves”; but this has real relevance, for example, to the Buddhist critique of “self”, which really does generalize to all objects — for early Buddhists, humans lack a “self” for much the same reason chariots lack a “self”, because they aren’t things in quite the way we normally take them to be.

Some things, of course, may be more intuitively “selfy” than others. The idea of discrete organic selves, or organisms, is applied to everything from viruses to humans. In this biological sense, I am my body, even though my body can change drastically over time.

Two troubling questions arise here, and they’ll recur for our other ideas of “self”. First, can my concept of myself as an organism be trumped by other (say, more psychological) conceptions? If not, then if my brain were turned into a sentient machine, or if my body perished while my soul lived on, I would not survive! Some ghostly or robotic impersonator would survive, while the “real me” perished with my body. Could that be right? Or is the “real me” something more abstract? And why does the question of which “me” is “real” feel like it matters so much?

2. Persona

By “self” or “person” we sometimes mean the specific things that make you who you are. We mean someone’s personality, character, life-experiences, social roles, and so on. As the Stanford Encyclopedia article on selfhood notes:

We often speak of one’s “personal identity” as what makes one the person one is. Your identity in this sense consists roughly of what makes you unique as an individual and different from others. Or it is the way you see or define yourself, or the network of values and convictions that structure your life. This individual identity is a property (or set of properties). Presumably it is one you have only contingently: you might have had a different identity from the one you in fact have. It is also a property that you may have only temporarily: you could swap your current individual identity for a new one, or perhaps even get by without any.

3. Subject

We may also have a more generic idea in mind — a “self” as a subject of experience. But this too conflates several ideas.

First, there’s the idea of an experiencing subject, an experiencer. At a minimum, this could be whatever directly brings experiences about. But does this causal notion adequately incorporate our intuition of a self that “undergoes” or “has” its experiences? What would we have to add to turn an experience-generator into an unconscious self? And if some brain region or ectoglob can be “me”, where do we draw the line between the parts of the world that are me and the parts that aren’t?

Jonathon, for one, voiced skepticism about there being any fact of the matter about the dividing line between Me and Everything Else. Some philosophers even reject the very idea that a self exists “outside” or “behind” experience:

But even so, there remains the distinct idea of an experienced subject. Our self isn’t just hidden behind our experiences; it’s also indicated within them. Thus we can speak of experiences that are “self-aware”, in different ways and to different extents. This ranges from the self-awareness of explicit thoughts like “I am getting rained on!” to primitive perceptual impressions that a certain hand is Me while a certain chair is Not Me.

At the outer edge of this category, David_Perry raised the idea of a bare “phenomenological” subject, which I took to be the perspectivalness or subject-object structure in experience. Here our discussion became very murky, and David_Beard expressed some skepticism about the possibility of disentangling this idea from the very idea of consciousness.

In general, we had a number of difficulties reconciling the philosophical method of phenomenology, or describing how things appear from a first-person perspective, with the method of third-person science. Most basically, Neeraj asked, can the fact of first-person experience itself be accounted for in objective, scientific terms? As Briénne put it: Supposing I were an intelligent zombie or automaton, could you explain to me what this thing you call “consciousness” is? This brought us to another way of conceiving a self — behaviorally.

4. Agent

“Self” can be defined in behavioral terms. We generally say that humans and animals can perform actions and deeds, while beaches and kaleidoscopes, metaphors aside, cannot. So agency is an important way of distinguishing persons from non-persons.

Of course, “action” is a vague category. It’s easiest to tell persons from non-persons apart when we’re dealing with intelligent agency, i.e., behaving in a skillful, adaptive, goal-oriented way. We debated whether intelligent behavior can occur in the absence of conscious thought, and if so how we could ever identify subjects of experience based on how they act. Sam noted that we very readily ascribe agency, and perhaps even awareness, to beings based merely on their superficial resemblance to humans and other animals — suggesting that our agent-detecting intuitions are prone to leading us astray.

We might also distinguish deliberative agency, which makes decisions, from rudimentary animal behaviors that possibly lack real “choice.” Even more narrowly, we can ask what gives deliberative agents (or agents in general) free agency. Does social or political freedom, as Nathaniel suggested, inform our concept of “person”? Does psychological or metaphysical freedom help determine whether something is a self in the first place?

This brought to the forefront the important fact that our idea of “self” is not merely descriptive; it is also prescriptive. What things we call “person” is bound up with our values, preferences, and principles. Thus we have to ask how the above ideas relate to moral agency, a being’s responsibility for its own actions. A storm can make bad things happen, but it’s not the storm’s fault. What sorts of things can be at fault?

5. Patient

Just as an agent is something that acts, a patient is something that’s acted upon. Thus, along similar lines, we can ask what beings are moral patients — beings that can be harmed or benefitted. And we can ask whether there is a special, narrower category of personal patients — whether, for example, humans or intelligent agents have their own special rights above and beyond those of other sentient beings.

But the normative concepts of self aren’t just about morality. We also need to know what it takes to count as a prudential patient. Or, to ditch the jargon: What does it take for something to be the proper object of my self-interest? What sorts of things can be me, when it comes to my looking out for my own welfare?

The question seems so basic as to be bizarre. But in fact it’s not a trivial matter to figure out why I should care about myself — or, given that I do care about myself, what it takes for a thing to qualify as “me” — or how to go about discovering which things so qualify!

More generally, we can distinguish two questions:

1. What does it take to be a certain kind of self? What makes Bob, say, an agent?

2. What does it take for two things to be the same particular self? What makes Bob at 3:00 am and Bob at 3:45 am the same agent? Why aren’t the two hemispheres of Bob’s brain two different agents?

Thus far, we’ve only even begun to address the first of these two questions. And we’ve barely scratched the surface of the normative concepts of self, and of the relationships between the above concepts of agent, patient, subject, and persona. But we’ve made real progress, and we can use the distinctions we’ve drawn as tools for beginning to make headway on the remaining riddles.

For those interested in further reading on these two questions, I recommend John Perry’s A Dialogue on Personal Identity and Immortality, a rousing and very accessible introduction to the philosophy of self.