How much time will our civilization survive?

A conversation with historian Nasser Zakariya on "doomsday statistics" and other forms of scientific catastrophism

Doominations is a series of interviews with people whose life experiences or work in some way touches on the questions of interest to the post-apocalypse.

For my inaugural interview, I am really pleased to be talking with Nasser Zakariya, an historian of science who is a professor in the Rhetoric department at University of California, Berkeley. Nasser and I have known each other for exactly 20 years, as members of the same graduate program cohort. Nasser is the author of A Final Story: Science, Myth, and Beginnings (University of Chicago Press, 2017), a history of the way in which scientists have sought to create sprawling narratives about the long past, and possible long future, of the universe.

I was talking to him about Doomsday Machines the other day and it brought to mind to him an idea by the physicist J. Richard Gott that is sometimes referred to as the “delta-t,” which, in some forms, involves to trying to quantify a statistical answer to the question: “How much time will our civilization survive?”

I thought this would make for an interesting conversation, so I cajoled Nasser into letting me record an interview with him. What follows is an abridged excerpt below that is edited for readability. My input/questions are in bold. Paid subscribers will have access to an audio recording of of the interview that contains some material not included below.

So, Nasser, the other day, I brought up the end of the world (as I often do), and you were telling me about this thing called the delta-t. Can you, can you tell me what that was again?

In 1989, an astrophysicist named Gott at Princeton in Nature published an argument that in effect says something like: you can, take a guess — a strong guess — of the likelihood of how long any phenomenon will last into the future on the basis of how long it's lived to the present. He claims that he came up with this in the 60s when he visited the Berlin Wall. And he sees the Wall and asks how long it might persist. And he in effect decides that, well, I'm not in any way extraordinary. But, he asks, what constitutes extraordinary? For him that would mean coming at some sort of notable time in the duration of the existence of the wall.

So he then says, let me say that I'm probably somewhere in the midpoint of its existence. And then he says, if that’s true, then it's likely therefore to last X number of years. So he basically establishes a 95% confidence interval — a very standard technique in statistics — to say something like: the duration of the wall from this point forward is anywhere from a few months to a couple hundred years, something like this. Because it turns out to be 1/39th of the duration to the present, up to 39 times that same duration.

If this is in the 1960s, then this is only a few years after the Wall goes up, right? If he was there in its first year, he’d be 1/39th of a year, which is not that long, or it’d be 39 years. Right?

Yes, exactly. And you can see that that interval would get bigger and bigger as the duration itself got bigger and bigger, as time went on.

He said that he had told this to a friend in the 1960s, early on when he was starting graduate school. And then, when the Berlin Wall fell, in 1989, as he saw it his prediction had been confirmed. This is how he states it, at least. He refers to this as an application of the Copernican principle, and he treats it as a kind of wisdom of science generally that what has it taught us, it has taught us at all times to make sure that we don't see ourselves as exceptional.

I can see the logic of this. I can also see that by itself, there are obvious fallacies. I mean, I have lived about 40 years. And so under this approach, I could expect that my future longevity would be somewhere between one more year, or 39-times-40 years, which is a very large number that I'm not going to live to. I mean, we know that there are definite upper limits on the human lifespan.

So this is one of the big criticisms that he gets, is that he treats this idea sometimes as entirely universal. And he gives over the course of years from ‘89 forward a number of presentations of this, and gets a lot of press. There's even a book that's published that features this and these “doomsday arguments,” as they're called, as one of the prominent examples. And he gives predictions for everything: plays, the duration of journals, the duration, obviously, of the human species. That's in fact, one of his big foci ,if you like. And, so, in fact, he uses that as advocacy for space travel, for saying that, in effect, we could anticipate that we are likely to be in a catastrophic scenario before long. That space travel is maybe one of the ways to avoid it.

He sometimes says, you only do this approach, basically, when you're ignorant. And so, when you get into the details of the argument, it's often based on, and some of the debate over it is over the application of the “principle of indifference”: if you know nothing, you can treat basically any time as equally likely.

But a curious thing is, he's not the only physicist who's making an argument like this. And he's not the only one who's doing it in the context of referring to something Copernican.

In the eighties, Brandon Carter gets connected with a philosopher named John Leslie and they come up with a kind of separate argument that in effect says something like: You want to say that we’re not ordinary, with regard to the entire population of the human species, from the from all the way in the past to all the way in the future, right? Then you should — and this sounds a little bit like delta-T — imagine that you're somewhere in the middle, right? Think about what that means with respect right away to the fact that other science suggests to us, that the population has been exploding. Exponentially.

So what does that mean if you right now are in the middle? If the exponential growth continued… even if the exponential growth didn't drop radically, right? Then you'd be somewhere way improbably early. So then what it immediately suggests is that there's going to be some hard drop off. Hence some sort of, from their perspective, universal catastrophe.

So these are all arguments that are based on the idea that we have limited knowledge of the future, that we are assuming our present is some kind of “average” present, and then applying that to this question of the length of the human species. How much time do we have left? What are the kinds of responses to this?

The debates, I'd say, are the two camps are, are these arguments valid or not.

In the early 2000s, Martin Rees refers to it as the “doomsday argument.” And he focuses on Carter. And he makes this kind of point like, “what do we do with these arguments?” They're kind of, they're aprioristic. We can't quite get rid of them and we can't quite not, right?

There are a number of questions here. There's the arguments on their own lights. Is there a way to simply say they're valid or invalid? Do they, in a certain sense, kind provoke our own questioning of our everyday assumptions about certain kinds of scientific language or discourse in the way that you might imagine something like Zeno's paradox has done. Zeno's paradox is for some people dead wrong, resolved, and others not. I'd say for the contemporary moment, people just treat Zeno’s paradox as a little thing that gets resolved by calculus. But you can imagine putting that aside and putting aside those philosophers who might still dispute whether there's any sort of richness here, you could see it as sort of interrogating the sort of limits of our reasoning, right?

So you could say, is that something similar that's taking place with these? Are they really more about questioning “us,” like, how we can go about this? Can one really go so far as to see catastrophe or at the intervals that, for example, something delta-t starts expanding to, is it any way much more than the kind of wisdom that we already have, like, “this too will pass”? Is it saying much more than that?

So where do you sit on this?

I’ll say, I’m very skeptical of the utility of these arguments. I get very confused about some of this. Gott, for example, he applies the argument to so many things. One of them, for example, is Stonehenge. Stonehenge from his perspective is super long-lived. And so by the lights of the delta-t argument, you're going to have a massive integral of time for it to continue existing.

So let’s say that someone had the idea that they wanted to get rid of Stonehenge. Put up the Stonehenge Casino. Put a hotel up. So now the question is, how do you gauge your intentions? Let’s say the company, Stonehenge Casino LLC, like, should it assume it’s likely to fail? Or does it have to say, somehow, we are now in a special position with respect to this, or do we have to do something to accelerate the timeline? It tells me nothing on this, as far as I can tell. Except to ignore it.

And what if I decided that the middle was special, like it was the peak in some way? I could be constructing my forecasts on the basis of whether or not I was likely to be near the start or near the end. And I treated near the start and near the end as not special. I could do that, I could apply the same mathematics and start working it out. What am I gaining from any of this? The metaphysics of it, the philosophical commitments, the kinds of reasoning we’re engaged in, they don’t offer us a good way out.

There are a lot of examples from the Cold War of people going down lines of argument that lead them into wondering, how long will our civilization last? In your book, A Final Story, you talk about these universal histories that scientists construct, starting from the idea of a deep past of the universe and then moving forward into some kind of hypothetical future. Is doomsday always part of that, or is this a more modern concern?

There is no period in which “doomsday” is not part of what we would now call the scientific enterprise. So in the 19th century, a number of different examples come to mind. And they're kind of over the same complex of things, like sometimes like weighing together evolution and energy, and its availability. There are some principled reasons for this, energy and entropy and so on, the laws of thermodynamics. Like Hermann von Helmholtz, he looks at evolution, and says, “we have to realize that just as all species die, our species will die as well.” And he takes as part of the wisdom of science, even if it is kind of a bitter lesson.

And then you have other forms of the catastrophic that seem to hover over the later writings of Charles Darwin, or Francis Galton. That culture itself, by easing the conditions of life, in a certain way spells its own catastrophe. And all of these things are, at least, discursively linked.

I will say, that the main difference that strikes me between these 19th-century catastrophes and the 20th-century ones is perhaps the time scales they have in mind. There's the Helmholtzian “second law of thermodynamics implies that eventually all the energy runs out,” and whatever you believe the time scales are, that's a very long time from us. We’re not going experience that. Whereas the nuclear world seems to be about, what if we reduce that time scale to the an order of potentially decades or even, you know, a few minutes when you actually get into the actual the war of things.

There’s an interesting case of the chemist Harrison Brown who worked on the atomic bomb, and later advocated for a world government as the solution to avoiding nuclear war. And in the 1950s, he published a book, The Challenge of Man's Future, where he's just trying to point out, you have to keep in mind, whatever you're calling civilization, it’s basically householding. You need this much of this kind of heavy metal. You need this much food and water. What do you need to maintain this house? And he's saying, just looking at the statistics, if there isn't more energy around the corner, if there isn't other ways to do this, then there's no way this is going to last.

And he revisits this argument time and time again over the coming decades. And he downplays nuclear destruction — because they survive — and he starts looking at other things like massive distinctions between rich and poor nations and what that means and how that's in a certain sense unsustainable, what that means about dangers for life and what life means and so on. And he extrapolates to, funnily enough, 2020, and he says that if things keep going, we’d get to something like 10 billion people. He says things like, if we’re innovative, there are ways to keep this going. And it’s not a coincidence that he gets linked to people in the next generation who work on things like climate change and so on.

How would you characterize the difference between “doomsday arguments” that scientists have made in the past, versus the ones that are being made more presently?

I feel like one of the big differences which I've been trying to wrestle with between as doomsday arguments in the past and doomsday arguments now, is the rise of statistics. And some of the criticism, some of the fight over the validity of delta-t or other kinds of arguments is: what is a reasonable argument in a statistical frame, and what counts? And this is partly a fight over Bayesianism and Frequentism in another. And can you give even give a Frequentist interpretation of life?

In the 1950s, Wolfgang Pauli co-wrote a book with Carl Jung on synchronicity. And Pauli uses it as an occasion to talk about the play of archetypes, like the kind of image of what counts as, in his context, particular kinds of knowledge. And he focuses on Johannes Kepler, and Kepler’s vision of archetypes, which was explicitly linked to geometry, to the idea that the natural world was given in the language of geometry. And Pauli suggests that the modern archetype is probability, probability in statistics. And he sort of flirts with a vision in the sciences in which the sciences periodize themselves as like, a historical-economic mode of explanation of nature. And built in that is a kind of relativism, a historical relativism. And so what does that mean when we're making massive forecasts into the future, right? Said incorrectly, it becomes a kind of wrong form of hope. Maybe our archetypes and our language are so constraining, that science itself offers ways out that we can't see. From another perspective, it’s sort of the demand for a kind of modesty. Which might take us back to the Copernican principle, which at least might caution us in being overweening with respect to our forecasts.

So just as a final question — if I were pinning you down, just for your view right this moment in time, what would your answer be for the likely lifespan of the human species, or at least the human civilization that we've got right now?

I don't know if one wants to call it optimism, but I tend to side with the idea of the durability of individuals we call humans. When it comes to the question of our modes of living, what I would call for me, from my perspective, “civilization,” even now is threatened to such a degree that I'm not sure that I imagine it extending beyond a few more generations. Like, those things that I would want to call “civilization.”

But we’re already really not uncomfortable living with a lot of death, as long as the death is not quite as proximate as we want it to be. And so we watch the displacement of populations, we watch people being starved, and we watch people being bombed and we treat as though it's not a condition of our civilization. Our vision of current civilization is already consistent with a lot of population death and murder. And that makes me feel like, after a certain number of generations, this will not be the “civilization” that would be recognizable to me. But those future people might call what they have “civilization.” This is where it's hard not to be the historian and have this reflex put forward. But you asked me, right? And then I wind up giving this perspectival answer, which I don't love.

Can I phrase it one more time in terms of weirdly enough, like delta-t or Carter? They constantly look at population dropping and that's all they can say. They can't talk about textured lives. And so then they say by their arguments, that there must be something that we would call “catastrophe.” I actually don't agree with the premises, if I had to be honest with myself. Yet the conclusion, I agree, is a strong possibility —  that there will be something that we would call “catastrophe” even if we judged it by the texture of people’s lives. But perhaps we can imagine something else, like ways in which people choose for themselves somehow, to, for example, to live different or to have different kinds of energy consumption, to have population drops that aren’t just about death and demographic disaster. It would still be as consistent with their predictive models as an idea that tragedy is around the corner.