On causality

Cause and effectCausality is one strange concept. It is absolutely essential to our understanding of the so-called “manifest image” of the world, i.e., the world as perceived and navigated by human beings. (The distinction between the manifest and the scientific image was introduced by philosopher Wilfred Sellars.) It is crucial for us to distinguish between events that happen because (i.e., are caused by) other events, vs things that appear to be the result of chains of cause-effect but really aren’t. We think smoking, statistically speaking, causes cancer, meaning that there are physical events that make it more likely that if you are a smoker you will get cancer. But when a few years ago someone showed a statistically significant correlation between number of births in London and frequency of storks flying overhead, nobody cried out for a revision of human biology textbooks…

When it comes to the “scientific image,” i.e., how science tells us the world is, things are more complicated. Talk of causality is all over the place in the so-called “special sciences,” i.e., everything other than fundamental physics (up to and including much of the rest of physics). In the latter field, seems to me that people just can’t make up their minds. I’ve read articles by physicists, and talked to a number of them, and they seem to divide in two camps: those who argue that of course causality is still crucial even at the fundamental level, including in quantum mechanics. And those who say that because quantum mechanical as well as relativistic equations are time symmetric, and the very idea of causality requires time asymmetry (as in the causes preceding the effects), then causality “plays no role” at that level.

Both camps have some good reasons on their side. It is true that the most basic equations in physics are time symmetric, so that causality doesn’t enter into them. But it is also unquestionably true that we have to somehow explain the arrow of time and the fact that things do very much appear to happen one after the other. While we move freely back and forth the three spatial dimensions, we definitely don’t do that along the fourth, temporal, dimension.

Three possible solutions to this conundrum are: I) to say that causality is an “illusion,” part and parcel of the manifest image, but not really a scientifically viable concept; or II) to claim that causality somehow emerges from basic physics (whatever “emergence,” a philosophically controversial concept, means); or III) to argue that causality is fundamental and that there is something incomplete about quantum mechanics and general relativity, and that’s why it needs to be “added by hand,” so to speak, in order to describe how the world actually works.

This, in turns, depends on how one conceives time — the element that, after all, is needed for causality. For instance, Brad Skow adopts the “block universe” concept arising from Special Relativity and concludes that time doesn’t “pass” in the sense of flowing; rather, “time is part of the uniform larger fabric of the universe, not something moving around inside it.” If this is correct, than “events do not sail past us and vanish forever; they just exist in different parts of spacetime … the only experiences I’m having are the ones I’m having now in this room. The experiences you had a year ago or 10 years ago are still just as real [Skow asserts], they’re just ‘inaccessible’ because you are now in a different part of spacetime.

It isn’t entirely clear what this view does with respect to causality, and it doesn’t seem to explain why we feel like time is something very different from space. Moreover, it doesn’t explain, say, the manifest image-level difference between causation and correlation. None of this means that the block universe concept of time/causality is wrong, but it does mean that there are serious pieces of the puzzle still missing.

Lee Smolin has a very different idea of time, and therefore of causality, as I have explained in detailed in the past. For him quantum mechanics and relativity are indeed incomplete (on this everyone seems to agree, including string theorists, who vehemently reject Smolin’s approach), time is fundamental, and so is causality. Indeed, he goes as far as saying that the laws of nature emerge from the specifics of causal interactions at the fundamental level, not the other way around.

In philosophy too, causality has always been a messy business. Famously, according to David Hume, it is something we add onto our perception of the fabric of the universe, and that may not be inherent in it. As the excellent Internet Encyclopedia of Philosophy article on Hume and causality puts it: “Whenever we find A, we also find B, and we have a certainty that this conjunction will continue to happen. Once we realize that ‘A must bring about B’ is tantamount merely to ‘Due to their constant conjunction, we are psychologically certain that B will follow A,’ then we are left with a very weak notion of necessity. This tenuous grasp on causal efficacy helps give rise to the Problem of Induction — that we are not reasonably justified in making any inductive inference about the world.”

However, it is not at all clear whether Hume thought that this is all there is to causality, or rather simply all that an empiricist approach to causality allows us to say, and Hume scholars disagree on this point.

Modern philosophers have developed a number of different theories of causation (and of time), that attempt to take into account what we have learned from science, and particularly physics, and make sense of it. It’s not an easy task, to put it mildly.

One of my favorite modern ways of thinking about causality (though, of course, it has its critics and drawbacks) is the co-called conserved quantity theory of causation. Here are the two major versions, according to the Stanford Encyclopedia of Philosophy (if you keep reading that article, you will also see a number of standard objections raised against it, the proposed responses, etc.):

P. Dowe’s version (1995, p. 323):

CQ1. A causal interaction is an intersection of world lines which involves exchange of a conserved quantity.

CQ2. A causal process is a world line of an object which possesses a conserved quantity.

W. Salmon’s version:

Definition 1. A causal interaction is an intersection of world-lines that involves exchange of a conserved quantity.

Definition 2. A causal process is a world-line of an object that transmits a nonzero amount of a conserved quantity at each moment of its history (each spacetime point of its trajectory).

Definition 3. A process transmits a conserved quantity between A and B (A ? B) if it possesses [a fixed amount of] this quantity at A and at B and at every stage of the process between A and B without any interactions in the open interval (A, B) that involve an exchange of that particular conserved quantity.

Here is a list of universally conserved properties in interactions between elementary particles:

  • energy
  • linear momentum
  • angular momentum
  • electric charge
  • baryon number
  • electron-muon-tauon number
  • lepton number

All of this, of course, has profound implications for both science and philosophy, but also for the way we should think about the world, i.e., these considerations affect both our scientific and our manifest images of the world.

Recently, I’ve began to think of causality as somewhat similar, in its manifestations, to physical forces, such as gravity. While gravity is universal, meaning that it acts in every point of the universe, so that in theory we are subject to the gravitational pull of every body in the cosmos that has mass, in practice we only need to be concerned with the gravitational effects induced by sufficiently massive bodies laying close enough to us. Our everyday life is affected by the gravity of Earth, the Moon, and the Sun, and little else. You need not worry about the gravitational pull of, say, the Andromeda galaxy because, even though it’s huge, the thing is so far from us that its orbital period is billions of years, so that it has no measurable effect on your existence. You also don’t need to concern yourself with the gravitational influence of people around you, because while they are nearby, their mass is just too small to do anything of consequence to you.

Perhaps causality is like that: while it makes sense to think of cause and effect as a universal phenomenon, with everything connected to everything else, for any practical purpose we are free to take into account only local causal interactions, all the other ones being dampened or overridden so to become irrelevant. It remains to be seen what such view would do to radical metaphysical notions like universal determinism (and consequent reductionism), or to controversial ones such as top-down causation (and consequent anti-reductionism).

You would think that this is an obvious area of inquiry where scientists and philosophers should come together. It isn’t, in my opinion, simply a matter of letting science tells us how things really stand. For one thing, because I’m confident that a fundamental physicist, a non-fundamental one, a biologist, and a social scientist would have very different views of what “science tells us” (indeed, as I mentioned above, even fundamental physicists vehemently disagree among themselves, so…).

Nor, of course, is it a question of calling the philosophical cavalry to explain to the naive scientists how they ought to think about the matter. That would be presumptuous to the level of silliness.

But why isn’t the question of time, or that of causality, a straightforward scientific issue? Why do we need philosophers to begin with?

One answer would be because philosophers have spent literally centuries thinking about these issues, much more so than scientists, and so there is likely something to learn from the best proposals they have put forward so far.

But that’s not actually it, or at the least, it isn’t the whole story. I think time and causality are a perfect example of the power of “sciphi,” if you will, because the issue isn’t just one of discovering facts about time and causality, it is to develop an understanding of these concepts that allow us to keep pursuing Sellars’ overarching objective: “to formulate a scientifically oriented, naturalistic realism which would ‘save the appearances'” [1]. The more I think about it, the more it seems to me that the (or at the least a major) goal of philosophy is precisely to articulate a mapping function that connects the scientific image — which only science can provide us — with the manifest image, which we simply cannot do without as cognitively limited biological creatures of a certain kind (and that includes scientists, obviously).

[1]  “Autobiographical Reflections (February 1973),” p. 289 in Action, Knowledge, and Reality: Studies in Honor of Wilfrid Sellars, H-N. Castañeda (ed.), Indianapolis: Bobbs-Merrill, 1975: 277-93.

125 thoughts on “On causality

  1. brodix

    “So we are all OK with a mind or an intention being an ontological thing that exists and can cause stuff?”

    Why wouldn’t it be “an excitation of the field?”


  2. synred


    This Arkani-Hamed, Lance Dixon at slac and some others have created an alternative way of approximating QFT calculations:

    Arkani-Hamed thinks the trouble with Feynman diagrams is that they are expressly local. They show particles as interacting with one another at specific locations in space and time. The diagrams look reassuringly like the trails particles leave in a detector such as the party tumbler in my basement; indeed, this is why physicists were drawn to Feynman’s approach. Yet the calculational quagmire brings this feature of the diagrams into disrepute. Locality is directly responsible for the algebraic bloat. “You’ve insisted that the theory be local,” Arkani-Hamed says. “You then suffer through ten thousand terms.” By taking every point in space as strictly independent of every other, the Feynman technique overstates the amount of complexity in nature. Much of what appears in the diagrams doesn’t exist in the real world, such as “virtual” particles and “ghost” fields. Theorists must impose special rules to ensure that these unwanted guests don’t stay for dessert.

    Musser, George. Spooky Action at a Distance: The Phenomenon That Reimagines Space and Time–and What It Means for Black Holes, the Big Bang, and Theories of Everything (pp. 40-41). Farrar, Straus and Giroux. Kindle Edition.

    In effect it doesn’t sums up many Feynman Diagrams at once. It can be thought of quite differently than the usual virtual particle picture.

    I think the ‘locality’ here is not the same as that violated with entangled particles. It just ‘excitations’ rather ‘particles’ – they are intrinsically spread out, but do not propagate faster than light.


Comments are closed.