Permutation City
Greg Egan, 1994

the twenty-eighth book in the visitor recommendation series;
suggested by Aris Katsaris

The only thing I knew Greg Egan for before reading this book was a short story called "Reasons to Be Cheerful", about a sci-fi cure for anhedonia, which I happened to read a month before my own physical anhedonia abruptly became total.  That was at the end of 2012.  Five and a half years earlier, in an article on Time's Arrow, I had written that "I will in all likelihood never feel better physically than I do today, and never as well as I did yesterday" — which proved more true than I would ever have imagined in 2007 — before adding, "Unless the singularity does something about it. Maybe living in a hard drive feels really awesome."  At first, it seems like this is what Permutation City will be about — it opens in a world in which we have learned how to upload copies of ourselves into computers, where they live, fully conscious, in virtual reality environments, with all the memories their human originals had at the time their brains were scanned.  This is a big buy-in.  We don't know how consciousness arises in the brain, and the notion that you'd get the exact same consciousness from interconnected electronic memory registers that you'd get from interconnected neurons isn't too far removed from the idea two hundred years ago that because electricity can make muscles twitch, a really big jolt of electricity could bring a corpse back to life.  But, hey, it's fiction — you get a buy-in!  In any case, it turns out that living in a hard drive doesn't feel awesome — the "Copies" tend to describe it as feeling like being buried alive, and given the chance they quickly abort their new existence.

One of these Copies, a digital translation of a man named Paul Durham, isn't given that option.  He reluctantly assists the original Paul Durham out in the real world in running experiments to learn more about the nature of Copies' consciousness.  Now, before I get into this, I should say up front that a lot of what went on in this book went over my head, to the point that I kind of had to laugh that it was recommended to me — seriously, a book that's all about math and physics and computer science?  Why not put it in untranslated Farsi while you're at it?  But I think that what happens is something vaguely like this.  My understanding is that when I click the Save icon in my word processor, I cause a set of electrical impulses to alter a number of tiny circuits on my hard drive called memory cells; by convention, the two stable configurations of these circuits are called zero and one.  This sentence is stored as a series of such zeroes and ones; it begins with a capital T, which the American Standard Code for Information Interchange (ASCII) encodes as the number 84, which in binary is 01010100, and the h that follows it is 104, or 01101000.  But the circuits representing those zeroes and ones are not necessarily physically next to each other in the hardware.  The T that begins this sentence might be physically far away from the h that follows it, and the e might be somewhere in between — but it doesn't matter, so long as the program that displays those letters knows where to find them.  (Well, there might be some loss of speed in loading the file due to the need to traverse the distances between one memory cell and another, which is why there are defragmenting programs to place pieces of the same file next to each other.)  It turns out that the consciousness of Copies works in much the same way.  We are told at the beginning of the novel that the Copies live in "slowdown": to run the number of brain state changes that a human undergoes in one second, the computer needs seventeen seconds.  But the Copies don't feel any different from how they remember their human existence feeling.  And what the two Paul Durhams discover is that the same holds true even if the brain states are run in a different order.  The video feed from the real world normally looks to the Copy like it's on fast forward; during the experiment, it looks scrambled.  But he feels like he's just a man in a room counting, "One, two, three, four, five…"  Therefore, he concludes, consciousness is the product of mathematical patterns that produce consciousness and exists not only independent of the brain, not only independent of computer hardware, but independent of space and time.  The real Paul Durham responds to this proposition by turning off the Copy, whose next conscious experience is waking up in a flesh-and-blood body, told by his girlfriend that he had never been a Copy at all, but had had the VR environment projected into his real brain in order to learn what life as a Copy was like.  But he contends that this is proof that his theory was right: that he had been the Copy, a mathematical pattern, and "running" him on a computer merely interwove a few strands of that pattern with the VR environment; when that process was terminated, the pattern persisted — for erasing "2 + 2 = 4" from a whiteboard doesn't make that mathematical property cease to be true — and his consciousness picked up in another place where its pattern was interwoven with an environment where it could register experiences, which happened to be in a parallel universe.

Or something like that?  Again, much of this was beyond me.  I remember that once — I think it was at the beach house that I later put into Endless, Nameless! — I happened across a book that tried to wave away the distinction between space and time.  I understand the standard analogy: given a set of values in three dimensions, you can either call them length, width, and time, in which case you get a movie, or length, width, and a third spatial dimension called height, in which case you get a statue.  Similarly, this book at the beach house said, if you just think of time as a fourth spatial dimension, the equations all work fine and the math is much more intuitive!  But despite all the 4D-in-2D charts the book provided, I don't know how you go about imagining a fourth spatial dimension.  I guess the standard answer to that one is that the math is how you imagine it, but since I can barely do trig nowadays, for me that pretty much ends the discussion.  It also means that when the science headlines trumpet that the universe is a hologram or a six-dimensional crystal or whatnot, the articles that follow are virtually meaningless to me — and I get the sense that Permutation City was written for people who not only understand those articles but spend a lot of time mulling over the various theories.  Then there's the question of consciousness, which I actually have thought about a lot, mostly when I was younger and fearfully flailing around for some theory to support its persistence after death.  Persisting after they die is of course what the one-percenters in the novel are after by having Copies made of themselves, but consider what the scans actually preserve: traits, beliefs, interests, memories.  How important are these sorts of things in establishing a sense of self?  Very important, if we go by all the Tumblr sidebars announcing, "I'm an INFP! I'm a 420-friendly pescetarian sex-positive feminist! I'm into rock climbing, fine wine, and Star Wars fanfic! Here is the story of my life as captured in daily selfie photosets!"  And I used to find those sorts of things very important myself, but I've grown increasingly skeptical.  For instance, I've always been interested in the psychology of personality, and from what I've read, fundamental personality traits do tend to be fairly stable over the course of a person's lifetime… but they're also highly context-dependent, and how we behave in a given situation depends more on the situation than on us.  And how many people really think that a particular commixture of introversion and impulsiveness is what defines them?  Now beliefs are something else again: countless people over their centuries have indeed willingly laid down their lives for their religion or their politics or other ideological affiliations.  I would never have gone that far, but when I was younger I certainly kicked up a lot of drama about all my deeply held values and convictions.  And then, like everyone else, I found that as time went by about half of these core beliefs of mine turned around 180 degrees, or at least 90.  As for interests — even as a kid I could see how quickly those come and go.  But what I couldn't have imagined for most of my life was the extent to which the same is true for memories.  Up until I was around thirty I remembered most everything that had happened to me, complete with date stamps, and I thought of that mental autobiography as myself.  But then my memory deteriorated with a quickness.  If mentally filing away the events of each day used to be like collecting money from an ATM and putting it in my wallet, now it's more like standing in one of those cash grab booths frantically trying to hold onto at least a couple of the higher-denomination bills as they whirl past.  (And when I read a novel I have to take notes as I go because by the time I get to chapter four I've usually forgotten what happened in chapter one.)  It's very frustrating.  But it's not as depersonalizing as I would once have thought.

Egan addresses the question of what constitutes a person in the novel's B plot, which revolves around a Copy of a man named Thomas Riemann.  This Copy deals with his endless existence by launching a program to reincarnate him every eighty years or so, giving him new interests, new surroundings, sometimes even a set of false memories.  It doesn't give him a new personality, because he already changes his traits from moment to moment as he sees fit.  So if the point of getting copied was to make Thomas Riemann immortal, all of this raises the question, what of Thomas Riemann has survived?  For most of the novel, the answer is that he has one key, life-defining memory that he clings to: he murdered someone, and has been haunted by that fact for decades of his natural life and millennia of his unnatural one.  But eventually he even lets that go.  Which leaves him in roughly the same spot I ended up many years ago, after I had more or less discarded the idea that the self is a function of character or memory.  All that left, so far as I could tell, was this tautological sense of "me-ness".  Which I could sometimes convince myself was encouraging.  After all, there are billions of loci of "me-ness" on this planet alone!  If the only thing that matters is the sense that "there is a sentient entity called 'me', registering experiences", then you'd have to be a total solipsist to doubt that that sense will indeed live on past the death of any particular locus!  But ultimately I could never get past the fact that somehow the only experiences I ever have registered have been those that have happened to this particular, bizarrely arbitrary body — whatever a body is, given that none of its constituent atoms stick around for very long — which seems at odds with the notion that consciousness is transpersonal.  That is, the sense of "me-ness" doesn't help me any if I'm not the one registering it, so the fact that I can't experience anyone else's "me-ness" in the present strongly suggests that I will never experience anyone else's "me-ness" in the future.  I.e., the fact that I somehow found myself registering experiences after fourteen billion years of oblivion once doesn't mean that, at the moment of my death, the particular "me-ness" currently attached to this body will suddenly find itself attached to some other organism — say, an alien fetus in the Andromeda Galaxy — with the "me-ness"'s previous parameters wiped, and the intervening additional fourteen billion years of oblivion having seemed to pass in an instant.  This is why the idea of Copies as a route to immortality struck me as fundamentally flawed right from the start: Copies preserve everything about a person except that sense of "me-ness"!  What lives on is a separate consciousness!  Getting scanned is therefore like retroactively turning yourself into an identical twin, as if your twin's immortality would somehow make you less dead.  But Permutation City ends up positing the notion that a universe could be inhabited by a single consciousness, experiencing all lives sequentially even though some are actually happening simultaneously, if consciousness is indeed independent of space and time.

To which a natural response is to say, "Uh, yeah, but it's not. That's just the premise of a sci-fi novel."  But recently I've run across a lot of people talking and writing along these lines.  One article I encountered suggested that reality may bear no resemblance whatsoever to our experience of it, offering up this analogy:  The operating system of my computer encourages me to think of my hard drive as a sort of file cabinet full of documents sitting in manila folders.  It does so not because this is remotely true — as noted, my hard drive is just a set of tiny circuits, and it is not even the case that when I move a file from one folder to another a chunk of contiguous data physically moves from one location on a silicon wafer to another.  Rather, it does so because of market forces.  Operating systems that represented files this way became popular and survived; those that didn't became unpopular and died out.  Similarly, if I take a physical piece of paper and move it from one physical manila folder to another, I may think that I'm moving matter through space over a span of time, but perhaps this bears no relation to reality either.  We are products of natural selection, so perhaps concepts like "matter" and "space" and "time" are useful in helping us survive without necessarily being real.  Maybe our entire universe is a simulation being run on a computer in an outer universe, which may itself be a simulation, ad nearly infinitum.  Apparently a whole bunch of Silicon Valley billionaires, Elon Musk among them, are convinced that this is the case.  ("There's a one in billions chance we're in base reality," Musk says.)

In the second half of Permutation City, Paul Durham puts his theory about consciousness to the test by having a computer run a mathematical model of another computer, whose instructions would be to launch a mathematical model of another computer, and so on forever — counting on the notion that by sending a Copy into the original computer's VR environment, he would weave the mathematical pattern of his consciousness into the infinitely expanding mathematical pattern of the program, which would thereby continue to exist after being shut down on the original computer.  Running this program would actually be the Big Bang of a new universe, what he calls the "TVC universe", "extending itself from moment to moment by sheer force of internal logic, 'accreting' the necessary building blocks from the chaos of non-space-time by the very act of defining space and time."  But while the computer of the TVC universe will be running a VR environment for Durham and other Copies to live in, Durham wants to try another experiment at the same time.  He has kept tabs on another simulated reality known as the Autoverse, which instead of approximating an environment (faking scenery by playing animations in the windows, for instance) actually models it all the way down to the atomic level — and no farther, because the Autoverse is necessarily greatly simplified.  There are no protons and electrons, no forces other than gravity, and only thirty-two chemical elements.  Learning that an Autoverse expert named Maria Deluca has recently succeeded in modeling the evolution of one species of bacterium into another, Durham hires her and sets her the task of creating the conditions that would allow intelligent life to evolve in the Autoverse.  He calls it a thought experiment, because Autoverse environments tend to be the size of a petri dish, and Durham wants an entire solar system, which would require a computer of incalculable power.  But, he eventually reveals to Maria, the TVC universe can run it.  And after three billion years of Autoverse time, intelligent life does evolve, in the form of smart gnatlike creatures (the "Lambertians", after Autoverse creator Max Lambert) who have no technology but who communicate scientific ideas by flying in particular formations, sort of like the "dances" bees use to communicate the location of food sources.  Durham wants to see how long it will take for these lifeforms to discover that the initial conditions that led to their existence and the ancestor bacteria from which they evolved could not have arisen through any natural process — that they must have been Intelligently Designed by creatures from an outer universe.  Then he and Maria can show up and reveal themselves as their gods!  What fun!  (Maria isn't so sure about the "fun" part.)  As it turns out, there's just one problem.  The Lambertians do come up with a set of equations that explains the functioning of the Autoverse without needing to postulate an outer universe.  These equations are wrong — we know that Max Lambert created the Autoverse, that Maria Deluca created the Autoverse universe, and that Paul Durham created the TVC universe within which the Autoverse universe runs.  But they work.  They correctly predict the state of the Autoverse at any given moment.  And there are a lot more Lambertians than human Copies — more conscious beings that believe the brilliant but incorrect guess than believe the truth.  So if consciousness is independent of space and time — if space and time derive from consciousness — then the Lambertian equations are not a falsehood but a redefinition.  A redefinition that does not include the TVC universe where the human Copies live.  And so the TVC universe begins to fall apart.

Again, I don't know how much stock to put in the idea of "subjective cosmology" outside the confines of this story.  I've read pop science articles citing respected scientists in arguing that what common sense tells us about the nature of reality is all wrong, that matter is a function of mind and that the behavior of fundamental particles cannot be fully and accurately described without references to consciousness.  And I've read other pop science articles scoffing that this is all New Age claptrap and that phenomena like the effect of observation on the path of a photon just means that observation interferes with the photon, not that it somehow defines it.  I don't have the slightest idea who's right — you might as well spring the "blue dress or white dress" problem on a blind person.  So let me end on a tangent.  A long time ago I happened across a pretty obnoxious cartoon that missed the point of an old saw I'd been hearing since high school.  The biologist sneers at the psychologist: "Psychology is just applied biology."  The chemist sneers at the biologist: "Biology is just applied chemistry."  The physicist sneers at the chemist: "Chemistry is just applied physics."  In the cartoon, the joke is that the physicist says, "It's nice to be on top," and then a mathematician sneers, "Oh, hey, I didn't see you guys all the way over there."  Which is like laughing at the ancients for thinking that the center of the universe was the earth, and then explaining that it's funny because of course everyone knows that the center of the universe is the sun.  Er, no.  The universe has no center.  And similarly, in this little parlor game, there is no top.  Because math is just applied philosophy — the work of mathematicians is to add to a specialized set of logical propositions.  Which doesn't put philosophy on top, because what you believe is generally a function of how your heredity and environment have shaped your thought process, meaning that philosophy is just applied psychology.  And then we're back to "psychology is just applied biology", and the whole thing ends up forming a circle.  I bring this up because it occurred to me that if you want to study the nature of consciousness at a university, you will find the appropriate classes listed under biology and psychology and philosophy — particularly philosophy, if you want to most directly grapple with what current thinking tells us about the nature of consciousness, the nature of reality, and how the two interrelate.  And while Permutation City does name-check philosopher John Searle (whose class on the nature of mind I have audited!), implicit in the novel is the notion that consciousness really falls under the purview of physics and math.  Which in turn suggests a view of these fields that, like the cartoon, is hierarchical rather than circular.  (Whether the characters in the novel are the prose equivalent of stick figures I will let others judge.)

comment on
reply via
this site
return to the
Calendar page