People Are Data

What is the relation between you and the data stored about you in the servers across the globe? Is it your property? Does it represent you? What should happen when people mistreat that data?

These are the sort of big questions that are only going to become more relevant. Just a couple of days ago Twitter was abuzz with a funny example: someone had been leaking private information about Coleen Rooney, a celebrity and footballer’s wife, to the UK tabloids. I will let you discover the amazing story for yourself, but it suggests the question: what exactly was the wrong the leaker committed?

Coincidentally, last week I found an answer to this question that I strikes me as pretty convincing. It was in the unlikely form of a master’s thesis in English literature by one Carrie North of the University of Louisiana that I had for some reason stored on an old external hard drive (it was published, and stored, in 2016, but I have no recollection of doing so and hadn’t read it).

This week I tried to track North down on the internet to tell her how much I loved it. I have so far failed in this task (maybe she dropped out of academia, got married, and changed her last name?), so I decided to write up some of the main ideas that I took from the thesis. If I get her permission, I will update this post with actual quotations from it, but in the meantime let me tell you about the philosophical theory she develops, according to which, as my title suggests, people are data. All credit is to her; when I expand upon her ideas in ways not found in the thesis I make it very clear.

The Theory

North’s theory has, I think, two main components: a sort of philosophical underpinning that motivates the whole project, and case studies that exhibit those underpinnings. That’s ridiculously vague, so let’s jump in.

The Philosophical Underpinning: Economic Constructionism

“Economic constructionism” is an unfortunately polysyllabic word for a cool and interesting theory. It’s the idea that the concepts in terms of which we understand reality are not pure representational vessels, but rather are shaped by power relations and, in particular, by economic relations. Her theory, kind of amusingly, is well-characterized as ‘postmodern neoMarxism’ (the bugbear of the Intellectual Dark Web) despite the fact that ‘postmodern neoMarxism’ really makes no sense.

It’s explicitly postmodern, appealing to theorists like Judith Butler and especially Deleuze and Guattari, from whom she takes over the idea that the aim of philosophy is the creation of concepts (as they argue, apparently, in their book ‘What Is Philosophy?’ As an analytic philosopher I can’t make much sense of D&G although I think their project at least sounds interestingly connected with that of conceptual engineering which I have written about here). But it also has Marxist elements, as she seems to think that culture, including philosophy, is determined by economics.

Again, that’s all a bit abstract, so let’s get concrete. She thinks that our concept of personal identity is i) temporally relative ii) constructed out of the economic ideas around in at the time. She presents this with a series of case studies that are to my mind interesting and worthy of attention.

The first person she considers is John Locke. Locke is famous in philosophy for (at least) two things: for his theory of personal identity, and his theory of private property. North’s interesting idea is that these two seemingly very different theories deep down stem from the same source.

To see this, first consider the personal identity question. What is a (human) person, as opposed to say a slab of meat or a God or a beloved cat? One way to answer this is to find out what the difference between a slab of meat (say) and a person is. So consider: imagine I am alive on day one, and dead on day two. In both instances, the slab of meat is there (first as what we call a body, then as what we call a corpse). But it seems on the second day, the person is no longer there; so it seems that people aren’t slabs of meat. Using these sorts of arguments about persistence through time lets us differentiate concepts.

So: what are people, if not slabs of meat? Well, what is it for a person to exist through time? What makes me the same person now as I was yesterday?

Locke’s answer — roughly, and following North’s gloss, which I think some scholars might disagree with — is that for me now to be the same person as some past stage of me is for the past stage and I to have roughly the same memories of the bit of our lives up to the time when the past stage existed.

An example might make this clearer. Imagine that last night while I was sleeping my brain was removed from my head and put into a new body. I wake up today in the next body. Which is me? The person with my brain or the person with the remainder of my body?

The answer we’re meant to like — and which I think is pretty plausible — is that the person with my brain is me. It’s the person who wakes up remembering my yesterday — remembers going to the gym yesterday, eating pasta, watching Succession, and so on. We are where our memories are, and so personal identity is determined by memory. This, very roughly (and, again, maybe inaccurately) is a Lockean idea.

Locke also has a political philosophy that seems at first glance entirely separate from his philosophy of personal identity. And a key part of that philosophy is a theory of private property, stating the conditions under which one can come to own property. Again scholars dispute, but Locke thinks that under some conditions one can morally acquire previously unowned property and that one then has rights to it.

So far, so philosophy 101. North’s interesting thought is that the notions of identity and property are, at bottom, sort of the same idea. Just as our thoughts and memories are private, for us alone, so is our land. In support of this surprising connection, she cites passages from Locke in which he seems to include our own life as part of our property. Life is just another type of private property.

In order to see this in more detail, we need to introduce another aspect of her philosophy: the event. The notion of event looms large in the work of contemporary French philosopher Alain Badiou. For Badiou — as I understand him, which isn’t very much — events mark a sort of rupture, the beginning of something new that can’t be understood in terms of the concepts of the time. Events can be political, philosophical, aesthetic, even romantic. A paradigm event, in this technical usage, would be a political revolution, a scientific paradigm shift, or a new movement in art.

For North, one such event was the enclosures that took place in the 17th century English in which Locke lived and which were codified in the considerably later enclosure acts. Starting in 1604, about thirty years before Locke was born, what was formerly common land for the use of everyone started to be enclosed and privately owned. For North, this marked the beginnings of capitalism itself, and was an event in Badiou’s sense: something brand new that warped the time and forced a rethinking of the very concepts in terms of which life was understood. Locke’s political work clearly engages with this, but North argues that his philosophy of identity does too. Realizing that land is private, he came to think identity was too.

There is plenty one could criticize here, but I think it’s a very interesting idea. At the very least, it encourages us to attend to the material conditions under which philosophy and culture is produced, and offers the helpful lesson that we should be open to the possibility that those conditions affect that production.

Let’s continue, albeit more briefly, with North’s thesis. The next thing she considers is “L’Homme Machine” by Julien de la Mettrie. According to it, as the title suggests, man is a machine, operating according to mechanical laws. The key point for North (she doesn’t discuss it anywhere near as much as she does Locke) is the date of publication: 1748. It is, she points out, just after the 1712 invention of a commercially and technically (somewhat) viable steam engine by Thomas Newcomen (information here). Given the massive importance of the steam engine in the industrial revolution, North argues that this was another Badiouvian event, and that de la Mettrie’s work was responding directly to it. Something new had come along to jolt the economy forward, and the philosophers, as if pulled along, needed to and did create new concepts apt for the era.

The third concept-event pair North considers takes us up into the 20th century. Here she focuses on the computational theory of mind which, as its name suggests, thinks of the mind as the software run by cerebral hardware. The classical reference here, of course, is Alan Turing, notably his 1950 paper “Computing Machinery and Intelligence” which asked the question whether machines can think and proposed as a way to test it the famous Turing test. The key point is that this inaugurated an era of thinking of humans as information-processing machines, implementing instructions that guide behaviour in the same way that machines implement code that guides how they act.

The event that North pairs with this innovation, interestingly, is not a clear economic breakthrough (she considers some options of possible events, such as the moon landing, but dismisses them), but is the publishing of Turing’s other famous paper, in which he lays out the ideas behind the Universal Turing Machine, namely his 1936 paper “On Computable Numbers, with an Application to the Entscheidungsproblem”.

It might seem strange that an academic paper in mathematical logic should go along with the enclosure acts or the invention of the steam engine, but North thinks that this illustrates an important point: that events are becoming — to use her word — ‘rarified’, more abstract, and less physical. Economic development was first the actual physical closing off of land; then the mere invention of a machine, a set of instructions physically realized, and then the invention of a more universal, abstract machine that doesn’t perform some particular but can be programmed to perform any task. It’s hard not to think, in this era of bots trading wildly complicated derivatives on stockmarkets round the world, that there’s something to this idea of economic rarification.

(Let me just note for completeness that I have ignored a whole other important type of event in her thesis. She thinks that there are two different types of events that are relevant: conceptually-constructive events, and conceptually-destructive events. The ones I’ve presented here are examples of the former, but she also has examples of the latter. She talks about, for example, the midcentury trifecta of women entering the workforce in World War 2, Roe vs Wade, and birth control as examples that helped with the movement of ‘woman’ from the category of unperson to person (it destroyed the idea of women as unpeople), although she thinks this is far from a completed project; and more more equivocally, she mentions the Whiggish interpretation of the US civil war as doing the same for black people. While very interesting, to get into this would make this post much too long.)

So far, so interesting (I hope). But now we can finally come to the title of this article, and where I take it she would say her most important contribution lies. This is that the next stage in the development of the idea of personal identity will be assimilating people to data.

And she thinks this simply because she thinks that’s where economic events are heading, and that that’s where the ever-increasing abstraction in the models we use to think about identity are heading. The Universal Turing Machine, although more abstract than a steam engine, is still in a sense a machine (if a merely hypothetical one). What comes next will move away from the last pretense of physical, bodily trappings.

From an economic perspective this doesn’t seem ridiculous. We all know how important data is, how our socializing (social media), dating (Tinder), and for many our jobs (Uber) are mediated by software concerned with sucking in, processing, and routing data. Especially companies like Uber and Airbnb seem dead set on moving as quickly away from the physical substrate — cars, buildings, stably-hired employees — that previously defined their industries. As before, North thinks, where economics leads philosophy will eventually follow.

Let me use the — kind of dark — thought experiment that North uses to try to convince you that people are data. Imagine an immediate future in which a loved one grows up in an environment in which all data about people is recorded and stored. Imagine that loved one dies, and that data is collected up and stored on a USB stick, the only copy. It’s important to think about the magnitude of data here, she tells us. Do not analogize this to a photo album; it is ‘toto coelo’ (which means, like, ‘very’) different, orders of magnitude more data about a person than we are used to having.

Now imagine someone gets hold of it, and stomps on it, destroying the last traces of your loved one. What sort of wrong has that person done?

North argues that while it isn’t murder (obviously) it is nevertheless something, and something very important. It is a species of violence which she calls ‘info-personal violence’ and it is deserving of great censure, and she thinks that this intuitive judgement that we will all share is indicative that something deep is changing in how we think about people in our new data-rich world.

Unfortunately, she doesn’t have much to say about ‘info-personal violence’. In fact this bit, which is just at the end of the thesis, seems a bit rushed. So I want to end by considering ways to develop that thought that make use of some ideas that wouldn’t have been quite so apparent to her when she was writing about 5 years ago.

Getting People From Data

Before doing that, one other thing. She wants to argue that people are data. But … something just seems off about that. People are things with which you can interact, with which you can do things. People react when you interact with them. Data is ‘inert’, lifeness, inactive. So people aren’t data.

I think what she has to say here isn’t very impressive. She again pushes on the idea of the magnitude of data that is available. She tells us to imagine the USB stick with the loved one’s data on it and think that one could go through it for years always encountering new information about the person, and seems to think that that would suffice, or be emotionally satisfying. Because the data would be in effect boundless, she thinks, it would suffice for something like an interpersonal relation for a person — say the person whose loved one dies — to merely go through that data, relive the days of the deceased person.

This seems questionable to me, and moreover creepy: would you want to go through all the personal data of someone close to you? It sounds gross.

But I think she was very close to a right idea. When she was writing, big data-driven AI hadn’t quite made the popular press as it has now. But now we know that by using neural networks or other machine learning tools to churn through a lot of data, we can make some pretty decent predictions.

So imagine this. Imagine we have all the data (say, every single relevant piece of information about a person up to the age of 20), and imagine that some fancy new machine-learning algorithm is capable of using that data and producing predictions about what the person would do in any conceivable circumstance. It would be just like any other bit of machine learning: we’d train it up using examples and then let it figure out what the important connections between the input data (the Tweets, the emails, the Facetimes, the bank records) and people’s behaviour is.

This, I take it, is barely even science-fictional: it’s pretty plausible that if we could get and store such data in the first place, we could make software that would make such predictions.

So imagine there’s software that could emulate your deceased loved one, say in a virtual reality. In the virtual reality you could interact with the person in a way that would feel real and, I think, be satisfying: you can do with them whatever you previously would have, and their reactions are as unique and multifarious as any real life human’s. And now imagine our previous meanyhead person stomping on the drive containing the only available copy. What do you say then? Really think about it, about how you’d feel.

While it still feels a bit off to me to call this ‘murder’, nevertheless it seems pretty close, and that this person has done something very bad and deprived you of something good. But that judgement, it seems to me, is tracking a way that you would think about your data in this scenario: as close to person-like. And so this, it seems to me, is a coherent, futuristic, but not wildly implausible scenario, in which we treat data as somehow similar to people, and so a scenario that bears out North’s now five year old thesis.

That’s about the sum of North’s ideas. Even if you quibble with some of the details (and I think it’s a good idea to do so), I think they shed a very interesting light on what it’s like to live today, and pose questions that we will, eventually, have to answer.

A Confession

Reader, I’m sorry. I misled you. Carrie North doesn’t exist, and she didn’t write a master’s thesis about people being data. She is the protagonist of my philosophical novella Coming From Nothing, an extract of which you can read here. (She did fictionally write a thesis on people as data, and the book is, in part, about that.)