My colleague Andrés recently wrote about various theories of personal identity, and how a lack of a clear consensus here poses a challenge to ethics.  From his post:

In Ontological Qualia I discussed three core views about personal identity. For those who have not encountered these concepts, I recommend reading that article for an expanded discussion.

In brief:

1. Closed Individualism: You start existing when you are born, and stop when you die.

2. Empty Individualism: You exist as a “time-slice” or “moment of experience.”

3. Open Individualism: There is only one subject of experience, who is everyone. [Ed: this draws from Daniel Kolak‘s work]

Most people are Closed Individualists; this is the default common sense view for good evolutionary reasons. But what grounds are there to believe in this view? Intuitively, the fact that you will wake up in “your body” tomorrow is obvious and needs no justification. However, explaining why this is the case in a clear way requires formalizing a wide range of concepts such as causality, continuity, memory, and physical laws. And when one tries to do so one will generally find a number of barriers that will prevent one from making a solid case for Closed Individualism.

As an example line of argument, one could argue that what defines you as an individual is your set of memories, and since the person who will wake up in your body tomorrow is the only human being with access to your current memories then you must be it. And while this may seem to work on the surface, a close inspection reveals otherwise. In particular, all of the following facts work against it: (1) memory is a constructive process and every time you remember something you remember it (slightly) differently, (2) memories are unreliable and do not always work at will (e.g. false memories), (3) it is unclear what happens if you copy all of your memories into someone else (do you become that person?), (4) how many memories can you swap with someone until you become a different person?, and so on. Here the more detailed questions one asks, the more ad-hoc modifications of the theory are needed. In the end, one is left with what appears to be just a set of conventional rules to determine whether two persons are the same for practical purposes. But it does not seem to carve nature at its joints; you’d be merely over-fitting the problem.

The same happens with most Closed Individualist accounts. You need to define what the identity carrier is, and after doing so one can identify situations in which identity is not well-defined given that identity carrier (memory, causality, shared matter, etc.).

But for both Open and Empty Individualism, identity is well-defined for any being in the universe. Either all are the same, or all are different. Critics might say that this is a trivial and uninteresting point, perhaps even just definitional. Closed Individualism seems sufficiently arbitrary, however, that questioning it is warranted, and once one does so it is reasonable to start the search for alternatives by taking a look at the trivial cases in which either all or none of the beings are the same.

More so, there are many arguments in favor of these views. They indeed solve and usefully reformulate a range of philosophical problems when applied diligently. I would argue that they play a role in philosophy that is similar to that of conservation of energy in physics. The energy conservation law has been empirically tested to extremely high levels of precision, which is something which we will have to do without in the realm of philosophy. Instead, we shall rely on powerful philosophical insights. And in addition, they make a lot of problems tractable and offer a powerful lens to interpret core difficulties in the field.

Open and Empty Individualism either solve or have bearings on: Decision theoryutilitarianismfission/ fusionmind-uploading and  mind-meldingpanpsychism, etc.

Andrés goes on to discuss David Benatar’s argument for antinatalism— the view that it is ethically preferable for individuals to never be born– and suggests arguments in this space tend to “rely implicitly on personal identity background assumptions. In particular, antinatalism is usually framed in a way that assumes Closed Individualism.” Furthermore, since Closed Individualism is on shaky philosophical ground, Benatar’s argument for antinatalism is likewise questionable.

There’s much more there, and I strongly endorse Andrés’s core theme, that you can’t get ethics right if you don’t get personal identity right, and that most ethical arguments right now assume a theory of identity (Closed Individualism) which breaks in illegible ways if we try to apply it in novel contexts.

At the same time, however, it feels like the distinction between Open Individualism (OI) and Empty Individualism (EI) is merely a distinction in name only; both theories of identity give identical answers for essentially all practical and ethical queries.

The following are some thoughts on one possible way to conceptualize this difference such that OI and EI would definitely point to different things, and give satisfyingly different answers to queries. (Status: exploratory, not strongly held.)


OI and EI are usually phrased in terms of identity— OI says everything is the same thing, and EI says everything is a different thing. But it’s not clear what this claim about identity *means*, how it cashes out in ontological and ethical senses. And it seems like the core issue is with the ambiguity involved in OI’s insistence that ‘everything is one’.  What does it mean to say everything is one, that there is only one subject of experience? It’s not clear that this claim is phrased in a way that ‘ pays rent‘.

Hypothesis:

One way of sharpening the definition of Open Individualism is to consider its core claim not (only) at the level of identity, but on the level of experience. That is: we could take Open Individualism to assert that phenomenal reality is, in the most literal sense, one huge qualia-bundle, and although it seems like this qualia-bundle has partitions or boundaries, these apparent partitions are illusions. Phenomenal binding, on the other hand, *is* real— but on only the *grandest* scale; absolutely everything is bound together. Everything is ontologically unitary, in all important senses. (This brings to mind Wheeler’s One Electron Universe hypothesis, where reality is ontologically unitary, but somehow an illusion of locality and diversity-of-forms arises.)

But why don’t we experience reality as this grand oneness-without-boundaries? Well, perhaps we do and we don’t know it. An important and unsolved question in qualia research is how to handle the reportability of qualia. A move often used by functionalists is that the apparent differences between the subjective experiences of people might *most precisely* boil down to differences in the substructure of the brain processes they use to report their internal states. This move is often used to try to explain qualia away, but perhaps the functionalists *and* the Open Individualists could be right– the universe could be one big global qualia-bundle, but individuals might only be able to report things about the local causal microstructure of the brain/mind. (I note this feels weird and counterintuitive, but not ruled out by the evidence, and as I mention later it might be indicated by certain elegance considerations.)

This definition of Open Individualism gives nearly everybody something that they want— OIs and Buddhists get everything-being-one (in the strongest sense!), functionalists get a heightened focus on the mechanisms in the brain that generate qualia reports (the ‘easy problems’ of consciousness), formalists get qualia formalism as ground-truth (albeit only at the largest scale), and so on.

But nobody gets everything that they want— in particular, functionalists don’t get to eliminate talk about qualia, and formalists have to worry a *lot* about the nature and specific mechanics of qualia reports, and might need to accept there’s no way to get to ‘local ground truth’ about qualia (epistemologically speaking).

Philosophical motivation:

To justify this way of contrasting OI and EI, we can look at their core difference in terms of monads (From Leibniz, “an indivisible and hence ultimately simple entity”).  If monads do exist, then they probably exist at one scale; there probably aren’t two ‘flavors’ of monads. No hierarchy of indivisible things made up of other indivisible things. And so, if Empty Individualism is the case, monads, the-things-that-are-ontologically-unitary, the-things-that-are-indivisible, should be defined in terms of experience-slices. Those would *be* the monads. But  if Open Individualism is the case, there’s exactly one monad, and it’s the universe.

In short, I’m led to the following definitions of Open Individualism and Empty Individualism:  EI says the universe has lots of boundaries, is made of many monads. OI says the universe has no internal boundaries, is made of one monad. Either of these seems more elegant than the current haphazard definition of Open Individualism, that the universe is ontologically unitary but has distinct parts, is one monad made of smaller monads, is one big indivisible thing made up of a lot of smaller indivisible things. (I am still open to the logical possibility of OI with real partitions, but it does seem less elegant than the definition I’m offering, and elegance arguments seem really important here.)

Implications:

This difference between OI and EI would have deep implications for what kinds of knowledge we can expect qualia research to generate, and what sorts of methods might offer reliable data. Basically, EI would be a lot easier to manage; being able to divide and conquer is a key enabling factor for scientific progress. Easier to study the structure of reality if there are many small monads of bounded complexity to study and compare, rather than one big monad with only very fuzzy internal partitions.

In terms of ethics, if this version of Open Individualism is true, it’d be deep justification for utilitarianism: let’s be good to each other, because We Are The Same Thing.

But what’s true?

How do we pick between Closed Individualism, Empty Individualism, and Open Individualism? First, I think it’s important to note there seems to be a significant difference between truth & usefulness regarding theories of identity. It seems as though evolution has biased us toward Closed Individualism as sort of a ‘Goldilocks zone’ of selflessness vs selfishness:

  • EIs have trouble coordinating because there aren’t compelling Schelling points for its momentary-slices-of-experience to cooperate with each other;
  • OIs have trouble coordinating because their natural Schelling point is always cooperate with everything (since I Am You and You Are Me), which fails in the presence of possible defectors, and this ‘game-theoretic naivety’ prevents positive-feedback loops from forming;
  • Closed Individualism provides an effective middle path between coordination and defection: it allows an organism to reliably coordinate with its past and future selves (thus constituting a  temporal superorganism), while at the same time finding it natural to be able to strategically cooperate with or defect against other such temporal coalitions.

I think OIs and EIs who want to build effective coordination mechanisms should take note of how powerful of a strategy CI is. That said, just because something like CI has been evolutionarily  useful as a coordination strategy doesn’t mean it’s metaphysically true as a theory of identity.

What is “metaphysically true”? I suspect we can’t use the traditional method of picking theories (judging them by their predictive power) so instead I think we have to rely on elegance arguments. As Andrés suggests, I think we can already disqualify Closed Individualism here: for CI to be crisply true, there’d need to be a crisp carrier of identity, which seems less and less likely the more we learn about reality. But how do we pick between EI and OI? Essentially we may need to wait until we’ve solved consciousness (determined the precise formalism for qualia) and see which seems simpler, which seems to naturally ‘pop out of the equations’.