Graciela Rossi, Shutterstock
John Locke imagined a prince who finds himself inside a cobbler's body. René Descartes worried the world you experience might be a hallucination created by a malicious demon. Thomas Reid mused that a mind might be copied into "two or twenty intelligent beings" claiming to be the same person. All in all, the Wachowskis have nothing on Enlightenment philosophers when it comes to reality-bending thought experiments.
Puzzles of personal identity are not just fun exercises for a college philosophy class, but have real significance for physics and allied sciences today. Quantum physics, cosmology, and artificial intelligence keep coming across situations in which you cannot be certain of who you really are. Personal identity is recurring theme at FQXi meetings and in its members' foundational research. I've been fascinated by these topics for years, dutifully kept notes at those meetings, caught up with many of these researchers by Zoom during the pandemic, and will explore their ideas at depth in my forthcoming book on physics and the mind.
Many of these puzzles are familiar if you've been following foundational physics, but the implications are often even deeper than they seem at first. For instance, it's now a commonplace idea that we may live in a multiverse. If our universe is big or old enough or split into many parallel worlds, there could be copies of you somewhere out there. And that poses tangible problems for calculations, because you have no way to tell which is you. How can the laws of physics predict what you'll observe when you're not sure who "you" are? Some think this ambiguity is the origin of the uncertainty that we observe in quantum physics. Others think the predictive difficulty sinks the whole multiverse concept.
Physicists have also been mulling thought experiments that apply quantum physics to observers themselves. If you manipulated a human observer in the way you do a particle, you could create gaps in personal identity over time, with potentially paradoxical consequences. A realistic experiment of this sort is a long way off, but it still suggests there's something about quantum theory we're not quite getting. A.I. visionaries, for their part, have imagined uploading and duplicating your mind. This technology isn't as far off as you might think. Even if a full brain upload is impossible, all you need to do is program a machine with enough of your memories and personality to convince everyone--including itself--that it is you.
Putting aside these puzzle cases, personal identity comes up in plenty of other contexts. It may, for example, be important in making sense of our experience of time, which is a fascinating and hard-to-parse mix of physics and psychology.
The Other Mueller Report
One physicist who has taken on the continuity of personal identity is Markus Müller of the Institute for Quantum Optics and Quantum Information in Vienna. "You could say, well, this is too far out and I don't want to deal with this question," he told me. "Usually we don't make copies of each other, so let's not bother...But I think, in general, as physicists or philosophers, we should not just push questions aside."
For Müller, identity puzzles reveal a clash between "outside" and "inside" views of the world--the objective perspective that physics traditionally provides and the viewpoint of an embedded observer. Physicists generally assume the objective perspective is primary and ask what an observer will see. But in the development of a theory, we work the other way: We have certain experiences and infer an observer-independent view from them. Müller argues that we never need to do the switch. We can stick with the inside-out view as primary and see where it leads us.
From this vantage point, the fundamental law of physics will not be particle mechanics or field theory, but a principle of reasoning about our experiences. The most rational such principle, Müller argues, is a predictive method developed by computer scientist Ray Solomonoff in the 1960s. It mashes up two ancient principles: Occam's razor and Epicurus's principle of multiple explanations. According to Occam, if two explanations are equally good, you should prefer the simpler. It's not always right, but you can do no better. According to Epicurus, if more than one explanation works, consider them all. Don't settle on one prematurely.
[youtube:9bZjmFEHTJM, 560, 315]
In Solomonoff induction, you make predictions for future observations based on the past, not by attempting to describe a single possible external world, but by considering all allowed worlds and then combining their predictions in a weighted sum that gives primacy to the simplest models. The idea is to do physics the way a machine-learning system does, relying on statistical prediction rather than world-building. "What should you reasonably expect to see next, given what you've seen so far?" Müller said.
You can think of his approach in either a weak or strong way. Müller goes for the strong: He thinks there fundamentally isn't an external world, and we just infer one. For many, including me, denying external reality is a lot to swallow. But you can back off from this metaphysics and say, more simply, that there is a real world independent of your mind and Solomonoff induction is the most rational form of reasoning about it.
Under ordinary circumstances, the Solomonoff formalism is entirely equivalent to the regular approach of physics. But it provides extra insight when you are faced with identity puzzles. For instance, cosmologists commonly worry about so-called Boltzmann brains, which are a naturally occurring brain-in-vat scenario. If the universe cycles randomly through various configurations, copies of your brain will naturally arise. Over the vastness of eternity, you are more likely to be a copy than the original, so your observations are more likely to be illusory than real.
Müller notes a crucial feature of this scenario: that the veil could lift at any moment and betray your view of the universe as a big lie. If you have to allow for the possibility of such a dramatic revelation, your description of nature is way more complicated than assuming that what you see is what you get. Thus, Müller's formulation assigns this scenario a weight that is low and getting lower with every moment. "It's very irregular in a specific mathematical sense, and so it's very unlikely," he said.
It's gratifying to know that we're probably not Boltzmann brains. For cosmologists, though, Müller's analysis is a mixed blessing. They routinely use these brains as a theoretical sieve, reasoning that, since we're pretty sure we're not Boltzmann brains, such brains must be vanishingly rare, and any theory suggesting otherwise can be safely discarded. But Müller says they can't assume that. We will always pretty be sure we're not Boltzmann brains, as long as we're reasoning rationally. The relative numbers of these brains is immaterial. So we can't extrapolate from our own experience to the cosmological situation. Just because we think we're not Boltzmann brains doesn't mean the brains aren't out there somewhere. "Cosmologists cannot rule out cosmological models--declare them absurd--on the basis of having a Boltzmann-brain problem," he said.
Boltzmann brains are a weird edge case, but the broader lesson is that we have to be careful when drawing conclusions about reality from our direct experience. Often that is justified, but there are cases where our experience is heavily filtered by our thought processes. As Immanuel Kant famously argued, the laws of physics are a knotted tangle of the nature of the world and the nature of our own thinking. The task of physics is to tease the two apart.
The Flow of Time
Our view of reality is especially jumbled when it comes to time. Some aspects of temporal experience are physics, some are psychology, and researchers still debate which is which. "I think of this problem of time as one of the last great mysteries, up there with the problem of consciousness," said philosopher Craig Callender of U.C. San Diego.
[youtube:tjVPE9gTvZU, 560, 315]
An especially perplexing aspect of time is our sense that it flows. Who hasn't felt in their gut that time is slipping away, that the accelerating pace of events hasn't swept them along, or that a staff meeting is lasting forever? This presumes other aspects of temporal experience, such as the distinction among past, present, and future, but is an extra sensation. Some physicists conclude that time really does flow, but most consider that a non-starter. Time is the measure of motion; it can't itself move. "There's such a gulf between the experience of time and time as you intellectually understand it," Callender said.
He and other philosophers such as Jenann Ismael argue that personal identity is the key to our experience of flow. We live and learn; we love and lose. Through all these changes, we feel that one thing endures: our selves. We can look back into our personal narrative and see a continuity between our childhood and our present selves. "You have these second-order ways of thinking about your self that's structured through time," Callender said. "What's moving, then, is this self." This becomes our reference frame. Like a passenger on a train moving through the landscape, we feel the landscape is moving past us. "If I then switch from an ego-moving cognitive frame to a time-moving cognitive frame...it will then seem natural to speak of time as dynamic, flowing by," Callender said.
[youtube:WNH-sHlllD8, 560, 315]
On this view, the flow of time is a high-order experience that most creatures lack. An amoeba or robot lives entirely in the present. Without an autobiographical self, it doesn't realize its memories are accumulating, and it doesn't feel that one moment follows on the next. Humans and perhaps a few other animals have the impression that time passes, because we have a sense of personal continuity. We are not reborn every moment.