My two minutes hate: hyperobjects
As a given semester progresses and I become more caffeinated and more sleep-deprived, I inevitably drift further and further toward being a ranting, flanderized caricature of myself. (My most recent post already displays considerable development in this direction.) Given the inevitability just mentioned, the only thing I can do about this process is try to enjoy it.
In that spirit, I give you this post, which is one long sneer at the concept of “hyperobjects” as expressed by lit professor Timothy Morton in this interview I came upon when surfing the internet this morning. I don’t know why I even care about this. I guess it’s because 1) Morton’s ideas are dumb, 2) Morton’s ideas are dumb in a way that encroaches on my academic turf in a few ways, 3) Morton is a tenured professor who has published nine books and seventy-six papers, so we should expect better of him, dammit.
Here’s how the interview begins:
Q: To begin, what are ‘hyperobjects’ and how are they explained through the five characteristics you attribute to them – namely, that hyperobjects are viscous, molten, nonlocal, phased, and interobjective?
Hyperobjects are simply gigantic objects, gigantic from some other object’s point of view. (Here object is taken to mean any entity whatsoever.) So, for instance, from the point of view of humans, the biosphere is a gigantic object that surrounds us and penetrates us and lasts for billions of years. Global warming is an object that emerges within this biosphere as a result of fossil fuel burning. It lasts up to 100,000 years as igneous rocks slowly absorb the remaining traces of excess carbon dioxide.
So there are things, and some of them are big. Deep, man.
No, no — that was a cheap shot. He’s going to develop this concept further, right? It’s going to have involve all these attributes with bizarre names — “viscosity,” “moltenness” (?), etc. The real content is in those words. So let’s see how Morton defines them. First, viscosity:
Hyperobjects are viscous. This simply means that they stick like goo to whoever or whatever they touch. This stickiness is both physical and conceptual. For example, the more we know about the biosphere, the more we realize we are stuck to it. Say you decide to move to Mars because of global warming. In some sense global warming is still stuck to you, because that’s why you move to Mars. And on Mars you have an even bigger problem: You have to create a livable biosphere!
Two comments, one semantic and one substantial. Semantics first. “Viscous” is not a apt word for the concept Morton is trying to express. In fluid mechanics, the viscosity of a fluid is, roughly, the strength of its internal friction. A viscous fluid is one that strongly resist flowing, one that is “thick.” This is not the same thing as “stickiness” — viscosity can be thought of as how much the fluid sticks to itself, but that’s not the same thing as how well it sticks to other things. The latter sort of behavior is more accurately denoted by the word adhesion. Viscosity and adhesion are not unrelated (the no-slip condition for viscous fluids, etc.), but it seems from Morton’s remarks that he is only interested in describing adhesion itself — “stickiness” — rather than resistance to flow. (His choice of the word “goo” makes it seem like he’s using an analogy to some goopy fluid that’s both viscous and adhesive, but he goes on to refer only to how hyperobjects “stick to” things, so it’s not clear whether the [actual] viscosity is an intended part of the definition or whether it’s just piggybacking on the analogy.)
My substantive objection to this concept is a lot simpler: what isn’t sticky in this sense? The world is a big network of interacting causes. Everything in the world is constantly getting its causal fingerprints all over everything else. If I go to Mars, what exactly are the things that haven’t “stuck to me”? The whole world participated in creating the environment that led me to be who I am, and to go to Mars; is it all viscous?
Hyperobjects are molten. They are so long lasting and so massive that they physically refute the idea that space and time are firm, consistent boxes (either physical or conceptual) in which things just sit like balls in an executive toy. They are living examples of why Newtonian mechanics is wrong. Think about Earth, surely a hyperobject for its inhabitants. It was recently established, using a host of tiny gyroscopes, the most accurate ever made, that there really is a spacetime vortex around Earth. This vortex is an emergent feature of Earth itself. It doesn’t contain Earth. Earth produces it. Einstein was correct in this regard.
Oh dear. Morton notes, correctly, that experiments about the motions of gyroscopes in orbit around earth agree with the predictions of General Relativity, as compared to Newtonian mechanics. This different is very small and takes extremely accurate measurements to notice, as Morton observes; the effect for an object smaller (technically: less massive) than the earth would be pretty much invisible. From this he concludes that “big” objects are the ones capable of disproving Newtonian ideas about space and time. Fine, I guess, if by “big” you mean “containing lots of mass” (note that earlier Morton spoke of “gigantic” objects as though they were those that last for a large amount of time!), and if by “disproving” you mean “disproving given our current instruments.” (As far as we know, all objects “physically refute the idea that space and time are firm, consistent boxes” — but the effect is usually tiny, so it’s extremely hard to notice even for something as massive as the earth.)
Once you try to extend this idea beyond the specific example Morton provides, it falls apart. General Relativistic calculations are necessary for, say, accurately computing the orbits of satellites, but they’re utterly irrelevant to, say, global warming; the tiny effects involve just don’t matter on the relevant scales. (In other words, global warming is, ironically, too “big” — in a certain sense — to feel the effects that Morton says are relevant to big things!)
Hyperobjects are nonlocal. They are so massively distributed that they confound prejudices we have about objects as located in specific regions of time and space. For example, global warming causes hazardous weather such as tornadoes. You feel the tornado: It rips your house apart. But you don’t feel global warming. But global warming is the mother of the tornado. It’s a necessary condition for the tornado. Something you can’t feel becomes more substantial than a tornado tearing through your neighborhood! Electromagnetic waves and gravity waves propagate throughout space: They are nonlocal in this sense. Evolution is a gigantic wave of replicating molecules expressing as viruses, spider webs, arms, mucus, birdcalls, and my answers to this interview. (Iain Hamilton Grant’s concept of a megabody is quite similar to this.)
The stuff about the tornado seems like the beginning of a potentially interesting observation about the way that the most important forces in our world are “invisible” ones, ones that can’t be straightforwardly inferred from the more immediate events they cause us to experience. But then he screws it up with his statement about waves. I have no idea whether Morton is aware of this, but “nonlocal” has a very well-defined and interesting meaning in math and physics. A “nonlocal” process is one that isn’t “local,” where a “local” process is, roughly, one that is equivalent to some sort of spatially extended mechanical assembly of interlocking parts where each part only directly feels the effects of the parts right next to it. (Like a line of interlocking gears: each gear only feels the effects of gears far down the chain through the mediation of the gears it directly meshes with.) Electromagnetic waves are, in fact, a textbook example of a local process in this sense. (James Clerk Maxwell, the founder of modern electromagnetism, actually spent huge amounts of time trying to construct these sorts of mechanical analogies; the purely mathematical formulation that is currently in use was only developed later.)
Semantics again. But I’m not just pointing this stuff out to be a pedantic science nerd. Using terms in technically inaccurate, “slanted” ways is fine with me if you’re doing it for a good cause. But Morton isn’t: he’s appropriating the word “nonlocality,” which properly refers to something cool and interesting, and applying it to the almost unbelievably banal concept of “propagating through space.” By that definition, why, just about anything is nonlocal! I’m nonlocal (I just got up to put my laundry in the dryer). Your mom is nonlocal. (Yo mama’s so fat, she confounds prejudices we have about objects as located in specific regions of time and space!) Et cetera.
(In the remainder of this answer, Morton talks about the idea that very small phenomena in physics are ” ‘everywhere’ in some sense,” which is a bit closer to the real concept of nonlocality. But there’s no indication that Morton understands the difference.)
Hyperobjects are phased.
I think the last time I heard the word “phased” as an adjective was when I was a 12-year-old playing Magic: The Gathering. (Wait, no, that was “phased out.”) Okay, okay, fine: there’s no shortage of dorky-sounding terminology in science and math, and I don’t complain about it there. So let’s see what it means to be “phased”:
The reason they are nonlocal and molten is that they occupy a higher dimensional phase space than other entities can easily cope with. One tornado can be seen as one point on a plot of a huge weather algorithm that spreads out in a high dimensional phase space.
So far, so good. (Terminology primer: a phenomenon has N “degrees of freedom” if it takes N independent numbers to specify its state. For instance, if we choose to describe humans only in terms of their height and weight, then we’re talking about something with 2 degrees of freedom. A point on a 2-D plot also requires two numbers to specify its state — an X coordinate and a Y coordinate — and hence we could represent human height-weight pairs on a 2D plot, with one axis for height and one for weight. If we plot a bunch of points showing one person’s height and weight at different points in time, the result would look like a curve in 2D space. This construction — a 2D plot with curves representing the system changing over time — is a “2D phase space.” You need an N-dimensional plot for something with N degrees of freedom, and a high-dimensional phase space is just one where N is really big. Things that are extended in space, like tornadoes, tend to have big phase spaces because there are quantities [like their velocity] that can vary from point to point along their extent, and since the value at each point is technically independent of the value at every other point, you need a dimension per point, at least. In fact, if you want to take seriously the idea of space being a continuum, you’ll have to say there are infinitely many points in any finite spatial region, and hence that your tornado-or-whatever’s phase space is infinite-dimensional. Part of the fun of math is getting to say things like trippy-sounding things like “infinite-dimensional” a lot [so I guess Morton and I share, after all, our love of dorky terminology]).
When Edward Lorenz looked at weather patterns in 1963 he discovered a very strange entity in the high dimensional phase space: a weird figure of eight pattern that is now called the Lorenz Attractor, the first strange attractor ever discovered.
This is a very special sort of statement: it’s not just wrong, but backwards. Lorenz didn’t “look at weather patterns,” at least not real ones. Real weather patterns evolve over time in a very high-dimensional phase space. Lorenz simulated a mathematical system on his computer, a system that was an extreme simplification of real weather systems. One key way in which it was simplified is that it had a very low-dimensional phase space! (3 dimensions, to be exact.) Before Lorenz, people thought that the difficult-to-predict quality of things like real weather came from their high phase space dimension (in Morton’s terms, they were unpredictable because they were “phased”). But Lorenz’s 3D system displayed a remarkable amount of unpredictable behavior despite its low dimension. That is why Lorenz and his 1963 paper are so famous — they mark the discovery that even low-dimensional systems can be unpredictable. The Lorenz Attractor isn’t “a very strange entity in the high dimensional phase space.” It’s an object sitting in 3D space, one that’s roughly a 2D surface (to be exact, it’s 2.06-dimensional; yes, really; the fractional dimension thing is a long story, and about as weird as it sounds). You can see it, just like any object in 3D space.
This is not advanced, esoteric stuff. It’s stuff that is clearly outlined in any good popular presentation of chaos theory, of which there have been many, some of them bestsellers. Any layman who’s read James Gleick’s Chaos (and paid attention) knows everything I just said. Why doesn’t Morton?
Hyperobjects appear to come and go in our regular 3-D space, like the way the Sun seems to appear from behind a cloud, or a storm seems to arise in a cloudy sky. But if we were four or five dimensional beings, perhaps we would see them as gigantic vortices or tubes spreading out around us.
Here, Morton seems to mean that the appearances of hyperobjects in our 3D space are like Poincare maps, which are lower-dimensional cross-section of an object in a higher-dimensional phase space. (Think of taking a 2D cross-section of a 3D object and just looking at the cross-section. Think of how we might appear to beings in Flatland — one slice at a time.) But the reason that most high-dimensional systems in real life are so high-dim is the one mentioned earlier: they’re spatially extended and have at least one dimension per point. This means that, although in phase space they are high-dimensional things that we can’t visualize, in physical space at any given time they’re just objects, ones we can see like any other. We can see zillions of phase-space dimensions every time we look at anything, insofar as we see zillions of points, and each point has at least one phase space dimension attached to it.
Hyperobjects are interobjective. They are formed as interactions between more than one entity. For instance global warming emerges from the Sun, the biosphere, humans burning fossil fuels, carbon dioxide and so on.
Is there anything that isn’t “formed as interactions between more than one entity”? My eyes and my brain and my hands and my stomach and my heart and my spleen are all interacting right now. They are, in turn, composed of interacting cells of different types. (Molecules . . . atoms . . . quantum fields . . . )
What sorts of things are Morton really talking about? Earlier, he defined “hyperobjects” as objects that are large with respect to others. He seems to be talking about objects larger than humans, now — so I’m not a hyperobject relative to me, of course, though I am relative to my cells. The kernel of the idea seems to be that when things are way bigger than you, they powerfully affect your life and yet obey mysterious rules that are different from the rules that tend to govern your scale. (Different physical rules do tend to obtain at different scales of time, space, mass, etc., though this a complex physical fact that is not done justice by Morton’s simplistic and inaccurate notions about phase space, “moltenness,” and “nonlocality.”)
The basic idea that big things rule your life without playing by your rules seems to me to be basically true, for us as for our cells. But rather than stating this idea outright, Morton makes it sound much more complicated and deep than it is by wrapping in it miles of misunderstood science and pseudo-precise terminological bullshit. Maybe if Morton was an essayist or a novelist rather than an academic whose job requires him to present the appearance of esoteric expertise, he would be capable of producing an interesting and well-written development of this (honestly) interesting concept. Alas, that’s not the case.
(What the hell am I doing, close-reading an interview? Everyone sounds stupid when they’re put on the spot. I know nothing at all about Morton’s actual work — this is just me riffing endlessly on a little peripheral scrap of text I’ve latched onto for some reason. Oh well.)