Ed Feser, physics, and the Dunning-Kruger effect

Leah Libresco keeps recommending Ed Feser’s books, though I can’t fathom why. But it gave me the idea to try to write something about his books. So I opened my ebook copy of Feser’s Aquinas… and was quickly reminded that all of Feser’s “arguments” always strike me as blatant non-sequiturs, just like the original versions in Aquinas’ writings. There is, however, one thing that’s Feser-specific I can comment on: the way Feser makes a hash of modern physics.

Here’s Feser trying to defend one of Aquinas’ arguments against the possibility of a regress of causes (which I’ve already discussed at the previous link):

Aquinas’s example from the First Way of the staff which is moved by the hand is a standard illustration, and we can add to the example by supposing that the staff is being used to move a stone, which is itself moving a fallen leaf. Here the motion of the leaf depends essentially on the motion of the stone, which in turn depends essentially on the motion of the staff, which itself depends essentially in turn on the motion of the hand. For if any member higher up in the series ceases its causal activity, the activity of the lower members will necessarily cease as well. For instance, if the staff was to slip away from the stone, the stone, and thus the leaf too, will stop moving; and of course, if the hand stops moving, the whole series, staff included, will automatically stop. In this case the causal power of the lower members derives entirely from that of the first member, the hand. In fact, strictly speaking it is not the stone which is moving the leaf and the staff which is moving the stone, but rather the hand which is moving everything else, with the stone being used by it as an instrument to move the leaf and the staff being used as an instrument to move both stone and leaf.

…Causal series ordered per se are paradigmatically hierarchical with their members acting simultaneously, as in the staff example where the movement of the leaf occurs precisely when the movement of the stone occurs, which is precisely when the movement of the staff occurs, which is precisely when the movement of the hand occurs… since the lower members of a causal series ordered per se have no causal power on their own but derive it entirely from a first cause, which (as it were) uses them as instruments, there is no sense to be made of such a series having no first member. If a first member who is the source of the causal power of the others did not exist, the series as a whole simply would not exist, as the movement of the leaf, stone, and staff cannot occur in the absence of the hand.

What Aquinas is saying, then, is that it is in the very nature of causal series ordered per se to have a first member, precisely because everything else in the series only counts as a member in the first place relative to the actions of a first cause. To suggest that such a series might regress infinitely, without a first member, is therefore simply unintelligible. The leaf is “moved” by the stone only in a loose sense; strictly speaking, the leaf, stone, and staff are all really being moved by the hand. Thus to suggest that this series of purely instrumental causes might regress to infinity is incoherent, for they would not in that case be the instruments of anything at all (CT I.3). As A. D. Sertillanges put it, you might as well say “that a brush can paint by itself, provided it has a very long handle” (quoted by Garrigou-Lagrange in God: His Existence and His Nature).

This all sounds very intuitive, but it’s wrong. To see why, first imagine replacing the staff with a spring. And let’s suppose you’re pushing the stone along a smooth, low-friction surface, so that the force of your hand on the spring and the spring on the stone can both be constant. As you push the stone, the spring is compressed, but all the forces are steady so the spring isn’t going to make the stone go flying off. Forget about the leaf.

Now suppose you are very fast and slick, so that you can be pushing the stone along one instant (using the spring), and the next instant have taken your hand away so quickly it may as well have been instantaneous, leaving the spring momentarily in contact with the stone. What’s going to happen? Well, the spring is compressed, so it’s going to push off the rock and go flying. And according to Newton’s third law, the rock is going to get pushed. In fact, for a split second, the rock is going to get pushed exactly as if you’re hand was still there. It’s just not going to last, because the spring is going to go flying.

And if you understand physics, you’ll realize the situation with the staff and the stone is qualitatively identical. It’s just harder to see that that’s what’s going on, because a staff is a lot less elastic than a spring, which isn’t the same thing as being not elastic at all. But we know the two situations are really analogous, because the laws of electromagnetism say that the force that the staff exerts on the stone depends entirely on the relative positions of the protons and electrons that make up the two objects. The hand only matters because it will affect the future positions of the protons and electrons that make up the staff.

“But!” Feser will insist. The laws of physics can’t possibly matter here, because we’re talking about metaphysics! Except they do matter. Feser just doesn’t understand physics well enough to see that. And worse, his lack of understanding of physics leads him to imagine that as a philosopher, he knows that philosophical arguments are the way to determine whether he needs to know any science to do his philosophy, and the philosophical arguments have proven he doesn’t need to know any science, so he can be sure that anyone who tries to explain to him that he has his science wrong (and that this matters) is philosophically ignorant. That’s not a hole Feser can dig himself out of without learning more science, but he’s convinced there’s no reason for him to do that.

It seems like a rather nice example of the Dunning-Kruger effect.

  • http://vexingquestions.wordpress.com Daniel

    In fact, for a split second, the rock is going to get pushed exactly as if you’re hand was still there. It’s just not going to last, because the spring is going to go flying.

    Right, because the hand isn’t there to exert a causal influence! So the effects play out from previous causal influence, but causal activity is no longer present in the system as it was previously. Moreover, if the causal activity of the first cause (God) were to cease, not even previous effects would subsist to “play out” their various motions and interactions. For while the hand caused the spring and rock to move, it did not sustain them in existence. God, on the other hand (pardon the pun), sustains the whole system. If God ceased the activity of sustaining the hand, spring, and rock, then all three are annihilated. Of course, given that God does sustain creation, the spring and rock will continue to move about once the hand ceases to exert a causal influence on the system, as we would expect. These further interactions do nothing to disprove Aquinas’ point. In fact, you admit that the spring and rock will interact quite differently once the hand is removed. This change in interaction is extremely significant. There is a hierarchy such that the remote cause orders the system to a higher degree than the proximate cause. Once the remote cause is removed from within a system, the causal story will be altered and the previous order is destroyed! A fortiori if God ceases to cause those proximate causes of our universe, the previous order is destroyed and no new order comes to be. So, I think you’ve missed Aquinas’ (and Feser’s) point, IMHO.

    • Patrick

      That’s not a defense of Feser and Aquinas, its a brutal refutation. You’ve just explained why their analogy is trash, and why intuitions based on arguments akin to that analogy are not legitimate reasons to believe in their conclusions. Thanks?

    • Chris Hallquist

      Except this isn’t just a matter of previous effects playing out. The interactions between the rock and the spring are, at first, exactly the same.

      It’s worth mentioning that Feser is specifically contrasting how he imagines the hand/staff/rock case to work with cause-effect sequences that are ordered in time. So it’s the instantaneous effects that matter here.

      • anodognosic

        I have to agree with Daniel here, Chris. Maybe you’re right, in a nitpicky sort of way, that their movement is not, strictly speaking, simultaneous. But I don’t see how that touches the substance of the argument in any way. The point–that the movement of the hand is causally prior to that of the stick and leaf–seems to survive intact. I think nonscientific thinking does cause Feser problems, like, as Dorfl says below, when he treats cause-and-effect as a fundamental concept, or perhaps giving too much importance to categories such as “instrumental cause” (although I can’t say for sure without reading a little more context). Still, I don’t see how you’ve substantively challenged his point here.

        • Chris Hallquist

          I didn’t want to make the quote any longer than necessary, but yes read in context the issue of whether the causes are simultaneous, and the concept of an “instrumental cause,” end up being absolutely fundamental because Feser is trying to argue some kinds of causes can’t go back to infinity, but he concedes that’s *not* true of causes ordered in time.

        • JHendrix

          There is no such thing as absolute simultaneity. This was debunked about a century ago, by Einstein’s Special Relativity.

          Theologians and apologists like William Lane Craig like to dress it up a bit by adding unfalsifiable, non-detectable in principle yet supposedly real, absolute reference frames to the theory in order to try and get what they want, but to justify believing in THAT over the normal interpretation of SR, they appeal to a pre-existing belief in god.

          It’s crappy circular reasoning.

  • http://exconvert.blogspot.com Kacy

    Exactly! Feser doesn’t know science, doesn’t care to know science, and he uses his horribly archaic Aristotelian understanding of science to draw bad conclusions about reality.

    Feser treats biology the same way he does physics, as I learned when reading a blog article in which he defends monogenism, not based on actual scientific evidence, but on its metaphysical fittingness. (http://edwardfeser.blogspot.com/2011/09/modern-biology-and-original-sin-part-i.html)

    • http://Vexingquestions.wordpress.com Daniel

      I think Feser’s point is that the apparent conflict between Church teaching and modern biology can be resolved once one understands that “species” has one meaning in A-T metaphysics and another meaning when used by Neo-Darwinians. I can affirm that the “species” homo sapien sapien descended from about 10,000 individuals, and that the species-form human had two primary instances in an “Adam” and “Eve”. Feser is not rejecting modern science in favor of archaic Aristotelian science, he’s showing that modern science can cohere with Aristotelian metaphysics. I don’t see how this is problematic, or demonstrates a lack of scientific knowledge on Feser’s part. Could you elaborate?

      • http://exconvert.blogspot.com Kacy

        Feser writes:

        “After all, the question of human origins is not a matter to which biological considerations alone are relevant. Metaphysical considerations are at least as important — indeed, they are more important, as we shall see ”

        This is a clear example in which Feser admits that he has a higher regard for A/T metaphysics than he does for science, something that Chris was criticizing with regard to Feser’s understanding of physics. In this particular article Feser admits as much with regards to biological origins.

        This gets especially wonky with Feser’s view of the soul because he describes the transmission of souls working like something akin to DNA, or to use his terms “material transmission.” (As an aside, most of the intellectual operations he describes are known to be material operations.) He also tentatively suggests that the “ensouled” and “unensouled” could have intermarried.

        If anything, this shows the lengths someone will go to rather than admit that their theory is flawed. He’d rather claim that Adam and Eve and their offspring were sexing it up with the dumb monkey-men around them (wouldn’t that count, technically, as bestiality under Catholic morality? Or was that God’s plan all along?) than admit the flaws in his metaphysics.

        • MNb

          “he has a higher regard for A/T metaphysics than he does for science”
          The way he treats kinematics shows exactly the same. He writes it somewhere on his blog. All theological and philosophical problems can be solved by proper understanding of Aristoteles and Aquino. If science contradicts something then science is wrong. Screw empiry.
          This summarizes Feser’s entire position.

        • Bob the Chef

          Well, of course he has a higher regard for metaphysics as metaphysics is a more fundamental science than biology! But that’s not why he says what he says. He says morality hinges on metaphysics because morality is metaphysical in nature. Morality is not the object of biological study. Biology can inform ethics, for instance, and metaphysics as well, but the proper place to really get down to brass tacks is metaphysics. Can’t you understand that? Both A. and T. emphasized the primacy of experience as the ground of knowledge, making them forefather of what is called empirical science (I hesitate calling them empiricists because of the way the term is used by certain schools, and because today the term is often associated with particular methods instead of the range of methods hinging on experience as a starting point). Now while I think philosophers should keep abreast of or become involved in science (some, such as Heller and Jaki, do this), the problem in t his particular case lies in your court. The reason is that you don’t appear to have a very good grasp of philosophical and empirical methodology as well as how they relate to one another. Some say it’s a good thing there’s some discussion going on about these things, but I often doubt that this discussion is intellectually sincere (there’s a greater willingness to demolish and criticize the opponent in some catharsis of resentment or simply to get a cheap feeling of superiority, over understanding the position thoroughly). I say this because I’ve studied methodology and after one has put a serious effort into *understanding* it (as opposed to a flimsy, pseudo-erudition that sophists and lawyers may use), it’s blatantly apparent how lacking in this department his opponents generally are. I’m sure serious philosophers could raise (and do raise) objections to some of his arguments, but this is not what this is. There are also very deep reasons why Aristotle uses the method that he uses, for example. Some accuse him of anthropocentrism, but there are very good reasons for the position he takes (I believe Gilson, and existential Thomist, explains it in one of his books). Science also needs to be understood from this position such that we know what science is actually giving us, and not just something magical coming from SCIENCE!

          • Nox

            Some of the mistakes of Aristotle and Aquinas can be excused for their time. Some wrong things they said were completely consistent with the data they were working with. That sort of works as a defense of the men themselves. The fact that someone in the 13th Century couldn’t have known any better does not make their statements true (and in the case of Aquinas, straight up dishonesty is just as much of a problem as lack of data).

      • Laurence

        It’s a total ad hoc explanation and would not be considered remotely rational unless you are already committed to the idea that there is original sin. So while it might not logically incompatible, his explanation is not a good argument and is really just sophistry to save a theory that has been destroyed by the theory of evolution by natural selection. I could come up with all kinds of explanations for all kinds of things if I really wanted to and some of them might even seem plausible, but, at the end of the day, they would still just be stories and nobody would even remotely think the would be rational to believe.

  • anatman

    you can use the schrodinger’s cat story to neatly kill the infinite chain of causation idea. it is about as certain as anything in science that radioactive decay is completely random and there is no hidden variable (i.e. cause) involved. it is the decay (or not) of the atom that actuates the setup that kills (or doesn’t kill) the cat. so obviously, the the cat’s potential death results from a finite chain of causation. by the way, the experimenter doesn’t ’cause’ the cat’s death unless you say that handing someone a gun causes them to commit murder. if you say that a so-called first cause makes the atom decay, experimentally the first cause’s effects are completely random. in other words, god doesn’t play dice with the universe, god IS dice.

    • MNb

      And this, dear friends, is philosophy based on Modern Physics.
      Anyone who contradicts Heisenberg’s Uncertainty (better Probability, but OK), on which Schrödinger’s Cat is based, denies that the nuclear bomb works. The victims of Hiroshima and Nagasaki can witness that the cosmological argument is flawed in its very core: causality.
      Until someone shows up to formulate a causal theory that describes all these (sub)atomic processes as accurately and preferably better. There have been attempts (Zeh, Bohm), but they are far less satisfying.

  • Patrick

    You missed Feser’s absolute ignorance about basic Newtonian physics.

    If my hand pushes a rock, then my hand stops pushing the rock, the rock keeps moving. FOREVER. Or at least until acted upon by another force.

    Feser is taking friction as if it were a given because he’s taking his cues from a long discredited Aristotelian notion of “impulse.”

    • MNb

      Pardon me for repeating your point underneath. You’re spot on and need way less words than me.

  • Dorfl

    I am studying physics right now, and I’m used to thinking of ’cause-and-effect’ as a helpful approximation that can be applied in some situations. Those situations are very noticeable, since our brains are wired to seek out cases where that approximation applies, but much of the time there is no meaningful way in which causes and effects can be picked out from the time evolution of the entire system.

    That means I’ve always felt a bit baffled whenever philosophers* have talked about ’cause-and-effect’ as though it were an absolutely fundamental concept. Especially when they’ve asked what caused the universe – since space-time is a part of the universe, it should be a priori obvious that the concept of ’cause’ won’t apply in that situation.

    *or maybe it’s mostly theologians. I’m sorry if I accidentally denigrated your entire profession.

    • Andrew G.

      Philosophers have the fundamental problem that their discipline has no good way to weed out bad ideas permanently.

  • Nox

    Philosophers are concerned with cause and effect in a very different way than physicists. The fixation with first cause is particular to theologians.

  • MNb

    “the way Feser makes a hash of modern physics.”
    Glad you noticed. A bit of nitpicking though: Newton is not modern physics, but classical physics. What Feser presents is Aristotelian physics (he basically says that motion needs a cause, which contradicts Newtons First Law). These have been outdated since at least 400 years.

    “we’re talking about metaphysics”
    You lose me here. Since when is motion metaphysics?
    The way I read Feser is that he tells us that his philosophy of religion is more reliable to describe natural phenomena than kinematics. He screws Galilei’s experiments on falling subjects (actually rolling from a slope, but whatever).
    What makes his “philosophy” even more stupid is that it is actually possible to use Newtonian Mechanics for a cosmological argument. The key is that according to Newton not motion, but a change in motion (more precise: change in velocity) needs a cause. That cause is called force. That’s what the Three Laws say and what Newtons’ Universal Law of Gravity means (plus a few other laws, but this should be enough for my argument).

  • http://mllamberth@gmail.com LordGriggsSkepticGriggsyCarneadesHume

    Da niel, Aquinas begs the question in stating that should one take away the Primary Cause, she takes away the intermediate causes. Aquinas defeats himself with his superfluity argument! As Percy Byssshe Shelley implicitly states it:” To suppose that some existence beyond, or above them [ the descriptions-laws- of Nature,L.D,] is to invent a second and superfluous hypothesis to account for what already is accounted for.” And no, ti’s no case of mistaken category as that would beg the question!
    Logic is the bane of theists.” Fr. Griggs
    htttp://fathergriggs.wordpress.com mountains of information against the Deity
    http://ignosticmorgansblog.wordpress.com Why the Deity is just another square circle or married bachelor.
    http://ehrmanlabmb.blogspot.com like the next one
    http://buy-bull.posterous.com

  • http://verbosestoic.wordpress.com Verbose Stoic

    I think that all of these counters are so obsessed with trying to find the analogy/argument incompatible with physics that they’re missing what the analogy or example is trying to show. From your own precis of his point:

    “Here’s Feser trying to defend one of Aquinas’ arguments against the possibility of a regress of causes …”

    Now, you reply to that with the “spring” example, attacking the example directly. But recall that even to you the whole purpose of that example is to demonstrate that there cannot be an infinite regress of causes. In the case you cite … there’s still no infinite regress of causes. The same thing applies to the “momentum case” cited by Patrick and MNb, and the atomic decay case. They all, effectively, have a first cause, in that there is an event that starts the whole chain off. The friction case is actually the WEAKEST comment, because it is clearly the case that we have the hand starting an action and no infinite regress. In your case, it can be credibly argued that the cause of the continued motion is the “springing back” of the spring, which was caused by the compression of it … which was done by the hand itself, and so even under your altered scenario considering only this interaction (from the hand down) the argument still comports with modern physics. In the radioactive decay case — as well as in QM — there is clearly no infinite regress of causes because the claim is that there is, in fact, no cause. It might work against positing God as a formal First Cause of everything, but it doesn’t work against this analogy.

    Using physics against an analogy or example is valid is only when the example itself violates physics, by predicting something that wouldn’t actually happen … and that’s only true if the example or analogy is intending to make a prediction as opposed to simply clarifying a point. Your examples never really get at showing that, nor do they establish that if it is that the analogy itself becomes useless to clarify the point. So, no, he doesn’t really have to worry about physics here, since the analogy is merely to clarify the position. We can work out the hard parts later, using full physics and everything else at our disposal to figure out what’s really going on/

    • Ray

      “In the radioactive decay case — as well as in QM — there is clearly no infinite regress of causes because the claim is that there is, in fact, no cause. It might work against positing God as a formal First Cause of everything, but it doesn’t work against this analogy.”

      I’m tempted to take this as an admission that first cause arguments do not prove anything after all, and leave it at that. That said, I think you are being unfair here:

      “The friction case is actually the WEAKEST comment, because it is clearly the case that we have the hand starting an action and no infinite regress.”

      Near as I can tell, no one is claiming that the motion of the hand is uncaused, so this still leaves the question of whether the chain of causes for the hand’s motion can go infinitely far back. The whole point of the Feser passage is to set up a distinction between causal chains per accidens (which can go infinitely far back) and causal chains per se (which cannot go infinitely far back.) Without the claim that the motion of the leaf will cease instantaneously once the staff stops pushing the stone, from whence comes this distinction?

      • http://verbosestoic.wordpress.com Verbose Stoic

        “I’m tempted to take this as an admission that first cause arguments do not prove anything after all, and leave it at that. ”

        Depends on what you mean by that. I think that the infinite regress reply is both physically and philosophically provably incorrect, and that therefore all causal chains have to have a beginning somewhere. And from that, that there must be causes that are themselves uncaused (unmoved movers). Now, what specifically those are or if there is only one of those is an open question.

        “Without the claim that the motion of the leaf will cease instantaneously once the staff stops pushing the stone, from whence comes this distinction?”

        And the answer to that, in light of the hand motion having its own causes is … it’s an analogy, not an argument. It’s meant to explain how the concept works, not prove it in and of itself. And if it wasn’t, then again the friction example is a bad one because it presumes that the causal chain starts with the hand … and your counter here would be a completely different argument.

        • Dorfl

          If you have a physical argument against the possibility of infinite chains of causation* I really, really would like to hear it, because I’m not aware of any such thing.

          *Although, again, I’d be inclined to think less in terms of an infinite chain of discrete things, and more in terms of time evolution over an infinite amount of time.

          • Martin

            The argument does not so much seek to prove that an infinite regress of causes is impossible, but rather to show that there must be something with a power of self-motion, since nothing else has it.

            The question being asked is: what is moving the stone? The stick?

            Not the stick, because sticks can’t move themselves any more than stones can. So the hand, then?

            Not the hand, because hands can’t move themselves either. So the muscles?

            Not the muscles, because muscles are useless without motor neurons. So motor neurons are moving the stone?

            No, because they can’t move themselves either, and must be actuated by electrons.

            And so on.

            So the answer to “what is moving the stone?” is: something that pushes without having to BE pushed by anything else. An unpushed pusher.

          • Dorfl

            I’m not sure what you mean by “power of self-motion”. Any physical system which is not at a local energy minimum will change, of its own accord, until it is. Most systems you encounter in the real world will already have done that, which is why they tend to stay more or less the same until something external disturbs them enough to push them out of their local minimum. This is probably why our intuition assumes that things will stay unchanged until something outside them changes them. But again: that’s just an assumption that our brains make because it tends to work. It cannot be treated as some fundamental law that the universe is required to follow.

          • http://verbosestoic.wordpress.com Verbose Stoic

            Dorfl,

            Science can’t prove anything absolutely impossible, and so neither really can physics. They can prove contradictory to the evidence, or that the best evidence points to a certain conclusion. I concede that my loose statement might have implied the former, but in thinking about it I don’t think that’s quite true, but do think the latter is reasonable given QM: The fundamental building blocks of all of reality, according to physics, are uncaused if current physics is right. I think, given that, that you’d need a really, really strong argument to show that there IS something that actually has an infinite regress … and you’d still run into the philosophical problems with that if you tried.

          • Ray

            “The fundamental building blocks of all of reality, according to physics, are uncaused if current physics is right.”

            Are you talking about quantum randomness as in the radioactive decay case, or are you simply pointing out that the truth of the axioms of quantum mechanics can loosely be thought of as being uncaused. If the former, this really only holds in the Copenhagen interpretation of QM. The Many worlds interpretation is completely deterministic. If the latter, well, there’s more than one way to axiomatize quantum mechanics, so which set of axioms are you going to regard as uncaused? If you pick a set of axioms arbitrarily, it seems to me you may have proved something about your description of the universe, but you have proved very little about the universe itself.

            Or do you have a third type of uncaused thing in mind?

          • http://verbosestoic.wordpress.com Verbose Stoic

            Ray,

            My rather vague understanding of QM is that the indeterministic one is still the standard interpretation and the one that is purported to be best supported by the evidence, and that QM interactions are generally considered to be uncaused and so in at least some cases “just happen”. Is that incorrect?

            Note that, looking at it now, the Many Worlds interpretation is less parsimonious than the Copenhagen interpretation, and so should lose out scientifically in the case of a tie …

            Now, returning to the original question, either you can settle which case it is — is it deterministic or not — or we don’t know enough about QM to say anything, really, about whether it is or isn’t. This, then, doesn’t impact the analogy at all, and so cannot be an argument against it … but note that the person I was replying to insisted that it wasn’t caused as far as I can tell and so counted against the analogy. Taking that interpretation, then, was perfectly reasonable to carry on the discussion, even if the science isn’t as settled as I believed.

          • Ray

            MWI and Copenhagen are both popular interpretations among professional physicists (In fact I have an easier time naming modern physicists who prefer the former.) As far as Ockham’s razor goes, MWI proponents would argue that MWI is more parsimonious ( see e.g. here ) There are also a number of “shut up an calculate” proponents who tend not to think that “which is the right interpretation of quantum mechanics?” is even a meaningful question.

            That said, there is a fundamental asymmetry here. Those who defend first cause arguments are the ones arguing that something *must* be the case. So all that is required for a refutation is to demonstrate that there is a reasonable interpretation of modern physics where the premises of the argument are wrong. Copenhagen is certainly held by enough modern physicists to count as a reasonable interpretation of quantum mechanics.

          • Ray

            I tried to post a reply before, but it didn’t seem to go through.

            A good summary of the state of opinion regarding interpretation of QM (at least as it stood in 1997) is here:
            http://arxiv.org/abs/quant-ph/9709032

            In particular, it should be noted that there is no consensus regarding whether MWI or Copenhagen is more parsimonious. MWI proponents argue theirs is the more parsimonious interpretation, for example here: http://www.hedweb.com/manworld.htm#ockham%27s

            There is also no consensus that which interpretation of QM is correct is even a meaningful question (the “shut up and calculate” proponents for the most part would tend to argue it isn’t.)

            The bottom line, though: It is proponents of first cause arguments that are claiming something *must* be the case. MWI and Copenhagen are both perfectly respectable interpretations within mainstream physics, so if either of them contradicts what first cause arguments claim to prove, then the first cause arguments do not live up to their claims.

          • Dorfl

            @ Stoic
            “The fundamental building blocks of all of reality, according to physics, are uncaused if current physics is right.”

            I don’t think I’d agree with that. We don’t currently know what the fundamental building blocks are, so we can’t really say anything much about them with a reasonable degree of certainty. There does seem to be a strong feeling that space and time are likely to be emergent phenomena from some more fundamental physics. If this turns out to be true, then we pretty much have to throw out causality entirely when discussing that physics.

            “I think, given that, that you’d need a really, really strong argument to show that there IS something that actually has an infinite regress”

            I’m not sure what argument to give, since I’m not completely sure what you’re asking. I can say that with most physical systems, nothing very interesting happens if you let the time go to infinity. The mathematical description will certainly not break down if you do that.

            “and you’d still run into the philosophical problems with that if you tried.”

            What are those philosophical problems?

          • Martin

            Dorfl,

            “This is probably why our intuition assumes that things will stay unchanged until something outside them changes them. But again: that’s just an assumption that our brains make because it tends to work.”

            The argument does not rest on intuitions or assumptions, but on argumentation. It says that because a possibility is merely possible, it cannot make itself actual. For example, possible ice (actual liquid water) will not make your drink cold. Only actual ice can do that. So a possible cannot make itself actual, either.

            However, if we want to say that things like atomic decay are spontaneous, that no little particle hits the atom to make it decay, then we would probably say that it is just in the nature of some atoms to decay spontaneously. But in that case, then, you are just talking about final causality and you end up going right back into the argument (the Fifth Way) and end up with the same conclusion.

          • Dorfl

            “It says that because a possibility is merely possible, it cannot make itself actual. For example, possible ice (actual liquid water) will not make your drink cold. Only actual ice can do that. So a possible cannot make itself actual, either.”

            I can’t follow this analogy at all. If the ambient temperature is below zero, liquid water will turn into ice, without you needing to do anything to it. If the temperature is above zero, ice will turn into water.

            If you think this is cheating, since the environment itself is cooling the ice or heating the water, imagine an amount of liquid water in an empty vacuum, not interacting with anything. It will then quickly turn into very cold steam, without anything external acting on it.

            So, to repeat what I said earlier but using your terminology of possible and actual: possible things can and do turn themselves into actual things, of their own accord. The reason you don’t see this happen very often is that it’s generally already happened before you arrive to look at the things in question, which is why it feels reasonable to say that possible things cannot turn into actual.

          • Dorfl

            “cooling the water and heating the ice”, that is.

          • Ray

            Martin:

            The fifth way is a terrible argument, which is why almost no one ever brings it up first when actually trying to convince someone of something. It fundamentally rests on the assumption that whenever something behaves in a consistent fashion, it does so either because it is intelligent or because it is directed by an intelligence. This premise cannot be established a priori, since there is no agreed upon definition of intelligence, and it cannot be established by experience, because the only things known to fit the premise are humans and their technology, and zillions of other things behave in a regular fashion. The best you can say in these cases is that I can’t prove beyond all doubt that an intelligence didn’t do it, but that argument works just as well if I replace “an intelligence” with bigfoot or the invisible pink unicorn.

            Stoic:

            MWI and Copenhagen are both popular interpretations of quantum mechanics among experts, as is “shut up and calculate”, which often goes so far as to deny that it’s even meaningful to ask which is the correct interpretation of quantum mechanics.

            If looked at from the perspective of number of equations, MWI is actually more parsimonious than Copenhagen — since it requires the Schroedinger equation and nothing else. The Copenhagen interpretation also requires some sort of rule for defining when a measurement has taken place — the exact rule is a very contentious issue, but as it turns out there are almost no experimental consequences for choosing the rule differently as long as your measuring device is bigger than a very small breadbox, as it were.

            Regarding whether you can just assume Copenhagen is right and use that to prove infinite regresses are unlikely:

            1) Just because there are uncaused events doesn’t mean there aren’t infinite regresses of causes too. Each step of the regress may arise from a confluence of causes, some of which are uncaused and some of which aren’t.
            2)The argument you’re trying to make attempts to rule out any possibility that an infinite regress may occur in nature. As long as any of the three mainstream interpretations of quantum mechanics contains an infinite regress of the sort you claim can’t exist, you have not ruled out the possibility of an infinite regress.

          • http://verbosestoic.wordpress.com Verbose Stoic

            Dorfl,

            “I’m not sure what argument to give, since I’m not completely sure what you’re asking. I can say that with most physical systems, nothing very interesting happens if you let the time go to infinity. The mathematical description will certainly not break down if you do that.”

            Basically, it would be the point me to the causal chain that clearly just runs BACKWARDS to infinity. I’m not too concerned about what the mathematical model says here because it’s a model and so may not accurately reflect the views of the universe [grin].

            “What are those philosophical problems?”

            The simple one is this:

            Take any causal chain of events. Therefore, for event B to occur, then event A has to occur. If event A does not occur, then neither does event B. So, is event A caused or uncaused? If it is caused, then we must find another cause C that occurred so that A can occur. If it is uncaused, then we can stop. Now, assume that we never find an event C that can occur without another event C’ occurring. What event in the chain actually occurred to start the causal chain? By causation, one of them has to happen or else none of them do. But we argued that none of them can happen unless one of the others does. Thus, a causal chain like this, it can be argued, could never instantiate. So we need a first event in all such chains (ie dependent ones).

            Ray,

            “MWI and Copenhagen are both popular interpretations of quantum mechanics among experts, as is “shut up and calculate”, which often goes so far as to deny that it’s even meaningful to ask which is the correct interpretation of quantum mechanics.”

            Fair enough, but as I said in the context of what I was replying to it seems reasonable that Copenhagen was the one being used for the atomic decay example, or else none of them were, so it really doesn’t impact what I said at all.

            “If looked at from the perspective of number of equations, MWI is actually more parsimonious than Copenhagen — since it requires the Schroedinger equation and nothing else. The Copenhagen interpretation also requires some sort of rule for defining when a measurement has taken place — the exact rule is a very contentious issue, but as it turns out there are almost no experimental consequences for choosing the rule differently as long as your measuring device is bigger than a very small breadbox, as it were.”

            But why should we judge it on equations and not entities, which is what Ockham’s Razor actually applies to? Many Worlds posits many, many unobservable universes, while Copenhagen doesn’t. That the math is simpler doesn’t to me suggest that parsimony is on its side, especially if parsimony is supposed to roughly be Ockham’s Razor.

            “1) Just because there are uncaused events doesn’t mean there aren’t infinite regresses of causes too. Each step of the regress may arise from a confluence of causes, some of which are uncaused and some of which aren’t.”

            The argument there was the the fundamental QM interactions would be uncaused. Given no other evidence of uncaused series, I do think it reasonable to say that most of the initial interactions would be QM and that there’s no reason to think that there’s anything else required. But I’m comfortable throttling down that claim if you want to just it makes it perfectly reasonable to think that there aren’t any until evidence can be given for their existence.

            “2)The argument you’re trying to make attempts to rule out any possibility that an infinite regress may occur in nature. As long as any of the three mainstream interpretations of quantum mechanics contains an infinite regress of the sort you claim can’t exist, you have not ruled out the possibility of an infinite regress.”

            I corrected that, didn’t I, to say that implying that it ruled out any possibility was far stronger than I wanted to go?

          • Dorfl

            “Basically, it would be the point me to the causal chain that clearly just runs BACKWARDS to infinity. I’m not too concerned about what the mathematical model says here because it’s a model and so may not accurately reflect the views of the universe [grin].”

            Take any causal chain you like that runs forward to infinity. Flip the sign of the time coordinate, from t to -t. You now have a causal chain running backwards to infinity.

            I realise that this answer probably looks very silly, but the fact is that time does not have any built-in directionality. The fact that time does seem to ‘flow’ from the past to the future is an emergent phenomenon due to our universe having much lower entropy at times closer to the big bang. This means that you cannot make any argument about causal chains continuing forwards indefinitely that would not also apply to chains continuing backwards indefinitely. The most you can do is the say that a causal chain can only have infinite extent in one direction. Are you willing to change you position from “causal chains cannot extend infinitely far into the past” to “causal chains cannot extend infinitely far in both directions in time”?

            “Take any causal chain of events. Therefore, for event B to occur, then event A has to occur. If event A does not occur, then neither does event B. So, is event A caused or uncaused? If it is caused, then we must find another cause C that occurred so that A can occur. If it is uncaused, then we can stop. Now, assume that we never find an event C that can occur without another event C’ occurring. What event in the chain actually occurred to start the causal chain?”

            I think the error in your reasoning happens right here. By assuming that there was any event that started the chain, you are already assuming what you’re trying to prove, that there was any first event.

            Say that we have an infinite chain of events, and choose to look at some particular one of them, which we denote 0. You might ask what caused event 0, in which case I’ll say it was caused by event -1. You might then ask what caused event -1, and I’ll say it was caused by event -2. We continue like that indefinitely and for any event -n I say that it was caused by event -n-1. You will probably tire of it at some point and ask for which event -N is there no previous event -N-1. When I answer that there is none, that every event has a preceding one, you will probably get quite annoyed at me. But you will never be able to derive any logical contradiction from the assumption that the causal chain can always be extended one step further. You will be able to show that it leads to any number of very counter-intuitive effects, but no actual contradictions. And that something is counter-intuitive shouldn’t really impress us all that much, given that we are studying a situation where we wouldn’t expect our primate brains’ built-in intuition to be of much use.

          • http://verbosestoic.wordpress.com Verbose Stoic

            Dorfl,

            “The fact that time does seem to ‘flow’ from the past to the future is an emergent phenomenon due to our universe having much lower entropy at times closer to the big bang. This means that you cannot make any argument about causal chains continuing forwards indefinitely that would not also apply to chains continuing backwards indefinitely.”

            And my reply is … who was talking about TIME when I talked about BACKWARDS [grin]?

            Look, it’s either the case that we define what it means for X to cause Y at least in part by saying that X precedes why — including time — or we don’t. If we do, then flipping the multiplier isn’t valid because that changes what we meant by “precedes”. If we don’t, then it devolves to the philosophical discussion I gave below — based on something like dependence — and then the direction of time (or even if there is time at all) is irrelevant.

            “But you will never be able to derive any logical contradiction from the assumption that the causal chain can always be extended one step further. You will be able to show that it leads to any number of very counter-intuitive effects, but no actual contradictions. And that something is counter-intuitive shouldn’t really impress us all that much, given that we are studying a situation where we wouldn’t expect our primate brains’ built-in intuition to be of much use.”

            And my reply here would be exactly what I already said: because for the chain to occur at least one of these events must have occurred which is the one that occurred to actually start the chain? Your mathematical example, I’d argue, is talking about a series that has already been generated, and we just pick a point in it and talk about it from there. Fine. But unless the causal chain itself is independent, I’m asking how a specific chain was actually generated. Or, roughly, which number did you write down first that let you generate the next terms? What’s f0? Let’s say that you call it A. Well, if it is the case that A can only be generated if C happened, then we couldn’t have started with A; we needed C. So either we needed another event — call it E — to cause A and then we have our infinite chain, or else A just happened all on its own. If we appeal to E, then we can ask the same question about E. If A happened on its own, then we can stop.

            The philosophical argument I’m making here is basically this: this analysis is basically recursive. Start at the effect we’re at, and ask if this thing is caused by something else or if it isn’t. If it isn’t, then we have an uncaused cause. If it is, then move to that cause and ask the same question. I argue that in order for the chain to kick off at all we need to eventually hit a case where we stop asking these questions, because as the effects are dependent on each other they won’t happen until something else in the chain happens. If you extend that to infinity, then NOTHING happens because nothing in the chain can happen until something else happens but the infinity argument claims that none of them have happened (because, essentially, they’re still “waiting” for the event they’re dependent on to happen). So, somehow, they have to stop waiting, and the only way they can stop waiting is for at least one of them to actually happen. At which point we can point to that or those events and say “See, those are our First Causes”.

          • eric

            VS, I think the real issue here is that you agree with Aquinas’ assumption that “there is no sense to be made of such a series having no first member.” To put it in your words: “for the chain to occur at least one of these events must have occurred which is the one that occurred to actually start the chain.”
            The answer to that is: no. Not mathematically. It may seem absurd to your intuition, but formally it appears to be true that no start to the chain need be necessary, even for a universe with a starting point.

            Now, this doesn’t mean there must have been an infinite regress. Maybe there was, maybe there wasn’t. But what it does mean is that you cannot simply assert that an infinite regress is impossible/absurd and leave it at that. Aquinas could say that because calculus and set theory had not come along yet. But when you say it, you just sound like a philosopher who has not kept up with the last 300 years of mathematical discovery.

          • Dorfl

            The problem is that once you’ve flipped the direction of time, you’ve also flipped all dependencies:

            We don’t normally say “I dropped the egg because it smashed on the floor”, but as far as the laws of physics are concerned there is nothing wrong with that statement. The reason that we normally take the egg-dropping as being the cause and the egg-smashing to be the effect is mostly epistemological. With just a basic understanding of mechanics, you can start with an egg suspended in mid-air, and predict that it will accelerate towards the floor, ending in some event of a general egg-smashing nature. On the other hand, you would need near omniscience about the state of everything in the kitchen, combined with near-infinite computational power, to predict the mechanical waves traveling through the floor, the sound waves of a “!hsamS” travelling through the air, converging on the sticky mess, assembling the shreds of yolk sac and the shards of shell into a whole egg and flinging it up at my hand.

            Hence my earlier grumbling about cause-and-effect not being as fundamental as it seems. We can talk about a chain of events – with some arbitrariness in what we consider to constitute each event – but which ones are causes and which ones are effects essentially ends up being a matter of convenience.

            To clarify what I mean: picture a perfectly elastic ball bouncing along a perfectly flat surface, so each bounce takes it to the same height and moves it the same direction parallel to the surface. We now consider each particular bounce to be our events. Whether the ball has been bouncing along forever or not doesn’t matter for our purposes, we only need to look at some specific interval along its path.

            It would now be perfectly reasonable to say that bounce G was caused by bounce F, when the ball had position so-and-so and left the surface which such-and-such velocity, and that bounce G then caused bounce H, because at G the ball left the surface at position this-and-that with velocity something-or-other.

            However, say that we flip the time coordinate. This makes no real difference – looking at the system there is no way to tell that things are going backwards, except that the ball is now moving in the opposite direction. Now that we have interchanged past and future, it instead becomes more reasonable to say that bounce H caused G, which caused F, and so on.

            All we can say is that a particular bounce could not have happened as it did, without violating the laws of mechanics, unless the bounces next to it happened as they did. And by extension, every bounce requires every other bounce in the series. But which bounce is caused by which and causes which ends up being just a matter of sign convention.

            For more complicated systems, flipping the time coordinate is likely to give increasingly larger asymmetry between the computational tractability of deriving the effects of what you are currently considering to be the causes. But as far as the laws of physics are concerned, it’s still just a matter of convention.

          • Andrew G.

            Dorfl, have you read

            http://lesswrong.com/lw/qr/timeless_causality/

            and if not, does it change your position regarding the relationship between time and causality?

          • Ray

            Re: Timeless Causality
            I can’t speak for Dorfl, but I don’t think that article should change his or her position.

            1)By Eliezer’s own admission, his rule for determining the direction of causality doesn’t work if the laws of physics are deterministic and information preserving — considering Eliezer likes MWI quantum mechanics, which is deterministic and information preserving, this is a big deficiency.

            Relevant quote: This trick used the assumption of probabilistic generators. We couldn’t have done it if the series had been generated by bijective mappings, i.e., if the future was deterministic given the past and only one possible past was compatible with each future.

            2) Eliezer’s probabilistic model of causality makes the additional assumption that timelike separated effects of a single cause are uncorrelated. This is famously not true (c.f. the EPR paradox) in actual physics.

            Relevant quote: “On each round, the past values of 1 and 2 probabilistically generate the future value of 1, and then separately probabilistically generate the future value of 2.” So once we have L1 and L2, they generate M1 independently of how they generate M2.

            3)Eliezer is too quick to dismiss the notion that the way we actually decide the direction of causality is based on the second law of thermodynamics (which is the standard answer.) In particular his simulation example: “To compute a consistent universe with a low-entropy terminal condition and high-entropy initial condition, you have to simulate lots and lots of universes, then throw away all but a tiny fraction of them that end up with low entropy at the end. With a low-entropy initial condition, you can compute it out locally, without any global checks,” follows from the fact that the simulations are being run in a universe where the second law holds with overwhelming probability, with no need for additional assumptions about the direction of causality.

          • Dorfl

            Hmm…

            I hadn’t read that before. I admit I’ve never studied Bayesian statistics, but I think I only need to modify my position this far:

            If the laws of physics are deterministic, either because the many-worlds interpretation or one of the non-local deterministic interpretations is correct, then everything I said still stands: As long as you have limited knowledge of the state of a system, you will have the appearance of causes followed by effects. Rocks throws down cliffs produce crashes, but crashes don’t necessarily lead to thrown rocks. But as the resolution with which you know the state of a system and the computing power available to you increases, the distinction between cause and effect eventually blurs. You will become able to identify the specific kinds of “!hsarC”-sounds that will converge on rocks and throw them up cliffs. So at the most fundamental level, there is no real difference between a cause and an effect.

            If the laws of physics are not deterministic, then I need to modify my position a bit: Even as your resolution and computing power approach infinity, you won’t be able to predict that the “!hsarC”-sound will throw that rock up that cliff. You have to stick to rocks thrown down cliffs producing crashes. But even in this case cause-and-effect remains an emergent property. For a small enough system, you would not be able to distinguish causes from effects. For a system with maximal entropy, you also would not be able to.

            So at best, cause and effect can only be non-arbitrarily defined for certain kinds of systems, for some limited interval of time.

          • http://verbosestoic.wordpress.com Verbose Stoic

            eric,

            “The answer to that is: no. Not mathematically. It may seem absurd to your intuition, but formally it appears to be true that no start to the chain need be necessary, even for a universe with a starting point.”

            It’s not that it’s absurd to my intuitions. It’s that it’s absurd to what I say it means to cause something, which I outlined in detail around that statement. So, I get the math. I accept the math. I accept that it might even be possible to have infinite series in the physical universe. I don’t accept that you can have physically existing infinite CAUSAL series, for the reasons I gave. Thus:

            “But when you say it, you just sound like a philosopher who has not kept up with the last 300 years of mathematical discovery.”

            Since I accept the math, but say it doesn’t apply here because of the properties of the elements and the specific relation, and even go into translating the math into the case I am talking about, this seems a rather odd statement to make. On the contrary, when you say that, you just sound like a mathematician who doesn’t understand what problem he is trying to use mathematics to solve.

            Dorfl,

            “The problem is that once you’ve flipped the direction of time, you’ve also flipped all dependencies”

            My definition of “dependence” isn’t dependent — pardon the unintentional pun — on time at all. Thus, you can do all the manipulations you want to time and all of the dependencies — and the relevant descriptions — will stay the same.

            This, then, returns us to my two original objections. Either your definition of “cause” is time-dependent or it is not. If it is not, then if you describe it with time “running” one way and then flip that, you aren’t describing the same set or series of events anymore, and then we can ask which one better reflects reality. On the other hand, if your definition of cause is NOT time-dependent, then whatever you do with time will not impact the description at all, and so won’t impact any notion of whether or not we have a First Cause, and my analysis will still apply.

            You can escape this by saying that there is no sort of “objective” cause for an event, which is what I think you’re aiming at here. The problem is that I think it objectively true that, say, I was born at least in part because my parents engaged in sexual relations, and that it is muddled descriptions to argue that my parents engaged in sexual relations because I was born. Perhaps, as eric aims at, this is a case where our intuitions lead us astray, but I’d need a lot more than that to accept so radical a proposition.

            “For more complicated systems, flipping the time coordinate is likely to give increasingly larger asymmetry between the computational tractability of deriving the effects of what you are currently considering to be the causes. But as far as the laws of physics are concerned, it’s still just a matter of convention.”

            But I’d counter that the question we are asking is a philosophical one, not one of physics (necessarily). Physics defines the term “cause” in the way that best suits their questions and purposes, and so does philosophy. Thus, if physics defines “cause” in a certain way, that does not mean that that is the way the term is or should be used in the philosophical argument. It might be nice to use the physics definition that says that it’s all conventional, but philosophically we can see that this more side-steps the question than answers it — or, at least, I argue it does because that doesn’t answer my question AND leads to what I would consider absurdities. If physics could actually settle the question in a way that satisfies the philosophical concerns, then we’d have progress, but so far it look more like it side-steps it because of an accident of how it looks at the world as opposed to actually taking it on directly and showing us how the world really is. The same thing is the case for the “Something from nothing” cases; sure, that’s what “nothing” means in physics, but when the answer to that is “That’s not what we mean by nothing” it’s not a good argument to simply say “Then you have an overly complicated/wrong definition of nothing”. You need to engage the philosophy to actually prove that, yes, your definition is overly complicated/wrong.

          • Dorfl

            @Stoic

            Can you define precisely what you mean by the statement “A is the cause of B”?

            In my argument above I used an implicit definition of “A is the cause of B iff B would not have happened if A had not happened.” For non-deterministic physics we can amend that to “-probably would not have happened”. If you are using some other definition, I don’t think we can get much further unless I know exactly what that definition is.

  • eric

    VS:

    I think, given that, that you’d need a really, really strong argument to show that there IS something that actually has an infinite regress … and you’d still run into the philosophical problems with that if you tried.

    There are many things that actually have an infinite regress. Add a half, plus a quarter, plus an eigth, plus…

    IMO its not just physics that philosophers like Feser are unsophisticated about, its math too. There’s two types of infinite series; convergent and divergent. For some reason, philosophers who make the “there must be a first cause argument” assume that any past series of physical causes must be divergent. On that assumption, they then claim a prior series of infinite events would be impossible or absurd. But divergence is an unwarranted assumption. Its perfectly possible that there has been a infinite convergent series of causes in our past, with earlier causes smaller than later ones, ending at some finite time in our past.

    • http://thebronzeblog.wordpress.com Bronze Dog

      Heh. Remember some fun stuff about fractals, like convergent ones that would approach a finite area as you approached infinite iterations. I also recently reread Problem Sleuth at MS Paint Adventures where they referenced that sort of thing with self-containing objects, warped space and such, like a candy mech with a house on top, with a model inside that contained the mech, the house, and itself. Convergence was game mechanic since divergence triggered explosions.

    • http://verbosestoic.wordpress.com Verbose Stoic

      eric,

      And of course what I meant there was, in fact, some actual physical causal chain that has an infinite regress, as per the original topic, and therefore citing mathematical series that have an infinite regress doesn’t in any way impact the argument.

      And note that philosophers do not ASSUME that it cannot be the case, but argue for it. And against it. Kant famously did both. We are well aware of the mathematical series. We just don’t think — philosophically — that that sort of method works here. And in terms of your suggestion:

      “Its perfectly possible that there has been a infinite convergent series of causes in our past, with earlier causes smaller than later ones, ending at some finite time in our past.”

      Can you explain what that means, in the sense of asking this question: Can I find a “first” member in the past on that causal chain? Then that’s a First Cause. If the causal chain runs that way, ending at some finite time in the past implies that that is the “first” element as per most causal analysis in cases like this.

      Note that, of course, none of this answers what I said, which is that I wanted a really good argument for thinking that there ACTUALLY IS ONE. Suggesting that it’s possible somehow doesn’t quite cut it.

      • eric

        What it means is: no, in such a series you cannot find a first member in that causal chain. Ending at some finite time in the past does NOT imply that there is any “first” element. You can have an end at a finite time in the past AND no first element.
        Second, suggesting how its possible does indeed cut it, because I am not the one ruling out a possibility; you are. So if you want to rule out the possibility of an infinite series of causes, you have to give me a good argument as to why I should do that. Your assertion that there cannot be any such series does not get priority; if you want to claim there cannot be any such series, it is up to you to tell us why.

        • http://verbosestoic.wordpress.com Verbose Stoic

          eric,

          The problem is that by most conventional analyses of causal chains, anything that “ends at a finite time in the past” IS the first event in that causal chain, as I said. I think you’re aiming at the same sort of argument as Dorfl, but he’s explained it far better if I’m right about that. But then I still say you have the same problems as he has: if the definition of cause depends on time, you can’t simply reverse time and claim to have the same description, and if it doesn’t whatever you can do with time won’t impact the description and so won’t get you out of the issue.

          As for the last part, if you quote me asking for a really good reason to think these things are possible, then it indeed does not cut it to simply say that they are physically possible. And as I have indeed already said that I’m not totally physically ruling it out — that was in the comment that you quoted this from — but am simply saying that it seems best supported by the evidence, insisting that I’m still “ruling out the possibility” is a bit off, don’t you think?

          • http://verbosestoic.wordpress.com Verbose Stoic

            “As for the last part, if you quote me asking for a really good reason to think these things are possible …”

            Sorry, meant “actually exist” here …

          • eric

            The problem is that by most conventional analyses of causal chains, anything that “ends at a finite time in the past” IS the first event in that causal chain.

            Then those conventional analyses are wrong. Their understanding of infinite series is about 300 years out of date. There can be an infinite series of causes that ends a finite time in the past. The existence of converging infinite series prove it. At this point in history, with our knowledge, anyone saying or implying that this can’t be is just using a fallacious argument from incredulity.

            but am simply saying that it seems best supported by the evidence, insisting that I’m still “ruling out the possibility” is a bit off, don’t you think?

            You have provided zero, zilch, nada evidence that a finite series of causes exists in the past. To do that you would have to provide empirical evidence of a causa causans, and you have none.

          • http://verbosestoic.wordpress.com Verbose Stoic

            “Then those conventional analyses are wrong. Their understanding of infinite series is about 300 years out of date. There can be an infinite series of causes that ends a finite time in the past. The existence of converging infinite series prove it. At this point in history, with our knowledge, anyone saying or implying that this can’t be is just using a fallacious argument from incredulity.”

            Seriously, your replies are really making me feel that the argument is going something like this:

            A – You can’t have an infinite series of negative numbers that contains a positive number.
            B – Of course you can! You can have infinite series that contain both positive and negative numbers, so that proves that you can have an infinite number of series of negative numbers that contains a positive number.
            A – But an infinite series of NEGATIVE numbers cannot contain a POSITIVE number by definition.
            B- You just don’t understand math.

            From my side, the problem is not with infinite series. I accept how they work. I even think I mostly get how they work. What I’m saying is that an infinite CAUSAL series can’t have that, because of what it means to be a CAUSE. From my view, the debate is not over what infinite series can do but over what it means to be a cause. And nothing you have said has, in my view, addressed that at all.

            Here, again, is my objection:

            1) In any causal series, in order for an element to be instantiated at least one term must be instantiated so that the others can be instantiated.
            2) The “first cause” in any causal chain is defined as the element that was first instantiated in order to cause the rest to be instantiated.
            3) If you cannot find a “first cause”, then there is no element that was instantiated in that chain.
            4) Therefore, no element in that causal chain was instantiated.
            5) Therefore, if you cannot find a “first cause” the causal chain was never instantiated and so doesn’t exist.
            6) Therefore, if we can find an element in a causal chain that was instantiated, then there must have been a “first cause”.

            If you think that infinite convergent series impact this argument or the definition, feel free to spell out how in detail.

            “You have provided zero, zilch, nada evidence that a finite series of causes exists in the past. To do that you would have to provide empirical evidence of a causa causans, and you have none.”

            The Copenhagen interpretation of QM says that we have things that are uncaused, and since those events cause other events we thus would have examples of terminating causal chains. So we’d know THEY exist. Even with the qualifiers of Many Worlds and that this doesn’t prove the infinite causal chains impossible, that’s a heck of a lot more evidence that you’ve presented.

          • eric

            VS: your objection (steps 1-6) ignores feedback loops and iterative processes. Also, your premise(3) is the same as your conclusion (6), so your argument is circular. Try giving an argument for why there must be a first cause without making some variation of the assertion “there must have been a first cause” a premise. Right now your argument reads as ‘the universe would not exist without a first cause (3), so therefore the universe would not exist without a first cause (6).’

            [eric]“You have provided zero, zilch, nada evidence that a finite series of causes exists in the past. To do that you would have to provide empirical evidence of a causa causans, and you have none.”
            [VS in reply] The Copenhagen interpretation of QM says that we have things that are uncaused…

            Hey, that’s great! We are in agreement then, that the only empirically supported candidate for a causa causans is the laws of QM. Yes?

          • http://verbosestoic.wordpress.com Verbose Stoic

            eric,

            So, is it safe to presume, then, that you DON’T think that the problem with my reasoning is that I don’t understand infinite convergent series and so bringing them into the picture isn’t going to do anything to resolve the dispute?

            “VS: your objection (steps 1-6) ignores feedback loops and iterative processes.”

            Explain how.

            “Also, your premise(3) is the same as your conclusion (6), so your argument is circular. ”

            Nope. 6 follows from 4, actually. 3 says that if you can’t find a first element, then no element in the list was instantiated. 4 then points out that if that’s true, then no element WAS instantiated (obviously) and so 6 then says that if you CAN find an element in a causal chain that was instantiated, you had to have a first cause (following from the rest of the argument). They aren’t the same statement at all.

            “Try giving an argument for why there must be a first cause without making some variation of the assertion “there must have been a first cause” a premise. Right now your argument reads as ‘the universe would not exist without a first cause (3), so therefore the universe would not exist without a first cause (6).’”

            Well, aside from my not saying anything about universes … 3 follows from 1 and 2, and so the only assertion is probably 1, and I argue that that follows from the definition of cause.

            “Hey, that’s great! We are in agreement then, that the only empirically supported candidate for a causa causans is the laws of QM. Yes?”

            Is this your attempt to get me to say that God doesn’t have to be the First Cause of the universe? Because I’ve never claimed that He would, and wouldn’t claim that, so I fail to see what reason you have for saying that this is great or that we are in agreement … noting that you are now claiming to be in agreement over something that you claimed I had never presented, when I presented it from the beginning.

          • eric

            Well, aside from my not saying anything about universes … 3 follows from 1 and 2, and so the only assertion is probably 1, and I argue that that follows from the definition of cause..

            Hmmm…well I guess my problem is really with those then, not 3. Specifically, (2). For one thing, you are treating the instantiation of your first cause as a non-causal event. Why? Second,(2) seems arbitrarily limited to one element. Why do that? Why can’t we use an alternate (2) that says: “The “first causes” in any causal chain is defined as the set of elements that was first instantiated in order to cause the rest to be instantiated.” A set could contain infinite members, or one, so my (2) is broader than your (2) while encompassing it. Saying “first element” assumes a non-infinite chain, which is one of the things I think you are trying to prove. Saying “first set of elements” does not.
            So, maybe I placed the circularity with the wrong numbered statement, but you do seem to be building some specifics of your desired conclusion into your premise statements. (2) builds in the assumption or requirement that there is a singular first cause an the assumption that this first cause’s instantiation is not itself a cause.

            [eric] “Hey, that’s great! We are in agreement then, that the only empirically supported candidate for a causa causans is the laws of QM. Yes?”
            [VS] Is this your attempt to get me to say that God doesn’t have to be the First Cause of the universe? Because I’ve never claimed that He would, and wouldn’t claim that, so I fail to see what reason you have for saying that this is great or that we are in agreement

            Surely you understand that the cosmological argument is overwhelmingly used by theists to support a claim that God, in a traditional sense of a conscious, powerful agent, exists. I think it was definitely worth a few moments to determine whether we agree that this is an invalid use of the cosmological argument. To take a few moments to see whether we agree that the argument proves no such thing, because the cause it argues for could be a ‘mere’ law of nature. And it appears from your last post that you do agree with me on that. Yes? That when theists and theologians trot out the cosmologial argument as an argument for a sentient God, they are making an invalid argument. Yes?

          • Ray

            VS
            Your problem is that 3 does not follow from 1 and 2.

            Compare:
            1) In order for a number to be an integer, at least one number less than it must be an integer. (this is true, e.g. if X-1 is not an integer, neither is X.)
            2)The “first integer” is the lowest integer, which can be used to generate all other integers via the successor function.
            3)If there is no first integer, then there can be no integers.
            4)But, there is no first integer among the elements of Z. Thus there are no integers.

          • http://verbosestoic.wordpress.com Verbose Stoic

            eric,

            2 follows from 1, and so is essentially a “Let X describe this element” statement, so it doesn’t assume that at all. If I assume it anywhere, it’s in 1 … but I argue that 1 follows from what it means to be a cause, so that will simply move the debate out there. As for it being limited to one element, recall that 1 says that it must be at least one, so there’s really no problem in extending it, except that it means that you’d end up adding a lot of elements that were not instantiated by an element in the causal series (that’s what 1 means, you realize). If you extend that to all elements in the causal series, then you’d have a causal series where none of the members are caused by any of the other members in the series, making one wonder why this should be called a causal series at all. For every element that you add that is not instantiated by the term before it in the series, the more potential you have for having multiple causal series as opposed to one, changing the entire description. Obviously, that’s not really a good move for you.

            Ray,

            I don’t think that’s equivalent to my argument. In order for it to be equivalent, you’d have to put it this way:

            1) In a series of integers each element in the series must be generated by a successor function starting from at least one defined integer.
            2) The “first integer” in any series of integers is the integer that was defined to start off that successor function.
            3) If you do not have a “first integer”, then you could not have generated any element in the series.

            I think that this argument is valid in the sense that if 1 and 2 are true then 3 must be true. The problem in this case is that 1 is false for integers. As I have said before, the debate here is over what it means to be a cause, not over the math or, in my opinion, the logic.

          • eric

            VS:

            2 follows from 1.

            Nope. Since 1 says “at least 1,” it would be entirely consistent with statement 1 to define ‘first cause’ as the ‘set first intantiated.

            For every element that you add that is not instantiated by the term before it in the series, the more potential you have for having multiple causal series as opposed to one, changing the entire description. Obviously, that’s not really a good move for you.

            At the risk of starting a ‘grue’ argument, your “first element” is a set. So when I modify 2 to recognize that, I am not really making it worse for me because you start with an set that is instantiated outside the causal change too.
            But, let’s go ahead and grant for sake of argument that that my set is more complicated than your single item. So what? Is there some metaphysical rule against multiple instantiated things? I guess you can cite Occam, but is that it? Are you saying that your proof of a causa causans depends on the assumption that the universe will have no more than one causa causans?

          • Silentbob

            @ eric

            There can be an infinite series of causes that ends a finite time in the past. The existence of converging infinite series prove it. At this point in history, with our knowledge, anyone saying or implying that this can’t be is just using a fallacious argument from incredulity.

            With the disclaimer that I am neither a physicist, nor a mathematician, nor a philosopher, I think you are making a mistake in assuming that if a concept can be expressed mathematically it can exist in nature.

            The convergent infinite series you have described assume that it is possible to subdivide a quantity an infinite number of times and also that each element in the series has an absolutely precise value. Modern physics tells us that in nature these things are not possible. Any quantity in nature can only be subdivided a finite number of times before a quantum – or minimum indivisible quantity – is reached. Also there is uncertainty in any physical quantity and physics tells us that this is a fundamental property of our universe that cannot, even in principle, be overcome. Since there is always a finite, non-zero uncertainty in any quantity there is no such thing a quantity that is defined with absolute precision.

            Your convergent series can only contain an infinite number of elements if it is possible for those elements to be infinitesimally small and to have infinitely precise values, so I don’t believe they can exist in nature, even in principle.

          • Ray

            @SilentBob

            While the question of what can be realized in nature is certainly relevant, I don’t think it creates problems for Eric’s construction. Quantum mechanics as currently understood uses continuous parameters for both space and time.

            Regarding the uncertainty principle. It is true that in order to make measurements that are precise to within a given time interval, you need an amount of energy of roughly hbar divided by the relevant time interval, and infinitesimal time intervals will therefore create problems for making measurements distinguishing between the two ends of the interval in practice. That said, Eric is making claims about what is, not what can be measured in practice, so I don’t think this creates problems for him either.

            There’s also the issue that the wavefunction cannot have very large spatial or temporal derivatives at reasonable energy levels, which is roughly what gets you out of having to admit the famous Banach-Tarski construction as being physically realizable (more properly, due to their infinite kinetic energy, any physical realizations of the sets would not last long enough to translate them anywhere, but I digress.) Nonetheless, there’s no rule that I know of restricting how similar to one another adjacent causes in a causal chain can be, so we don’t need the wavefunction to have large time derivatives in order to realize Eric’s hypothetical causal chain either.

          • http://verbosestoic.wordpress.com Verbose Stoic

            eric,

            I think we’re really into different arguments here. You seem to desperate want to fight against there being only one independent cause, let’s say, so that you can argue that the argument doesn’t necessarily support the idea that that cause is God. As you probably should have remembered from way back, I have no interest in that sort of argument (I think you’d need to show intentionality to get that off the ground) but was simply arguing against the idea that you can have an infinite chain of causes. You seem to have abandoned that argument now despite arguing most strongly for it, since introducing more than one of these “first causes” doesn’t seem to help you get to that point.

            Now, under my actual argument, adding multiple first causes really has little impact on it. One element, a set, it doesn’t matter as long as at least one of the elements in the series was instantiated by something outside of that series. The complications for you, though, if you are going for the “infinite causal series” argument, is that the more of those we introduce, the more reasonable it is to splinter the series into multiple series and, thus, reduce it to what I say in 2 anyway, as I said. But i didn’t express that that clearly, so let me do that here:

            Imagine that you have this causal series: A->B->C->D->E, where -> denotes “causes”. Now, off the top, our candidate for a first cause is A, as in this hypothetical series it is not caused by anything else in the series. Now, let’s say that you point out that C was ALSO not caused by anything else in the series, and so is another first cause. My reply to that is to ask why I can’t look at it this way then:

            A->B C->D->E

            Or, in short, as two different causal series. And if we do that, I have two causal series all with only one first cause … and then 2 remains entirely true, and you’re even further away from having an infinite causal series with no first cause.

        • Ray

          VS

          No. 3 Still does not follow from 2. The argument you are hinting at appears to have the structure of an inductive proof:
          i.e.
          1) If at one stage of the definition process, no integers are defined, there will be no integers defined at the next stage. (inductive step)
          2) There is some stage of the definition process at which no integers are defined. (basis case)

          leading to

          3) No integers are defined after the basis case.

          The problem is, nothing in your premises 1,2 establishes the basis case.

          Returning to causality: In order to prove what you want, you need to assert or prove that there is a stage in your causal chain where nothing has been caused, i.e. a starting point. This is pretty clearly begging the question.

          • http://verbosestoic.wordpress.com Verbose Stoic

            It’s not an inductive argument. I think your reply here is basically saying “Why would you think that 1 is true?”. I argue that follows from my definition of cause, which we can talk about separately. If not, then I’m afraid that I don’t know what your claim is. To address the charge of my begging the question, my answer to that is that I am not assuming that there is a starting point, but simply that in order for us to say that A caused B we have to say that B is instantiated by A, and that an element that is not instantiated cannot instantiate anything else. Everything follows from that.

          • Ray

            I don’t see how
            “an element that is not instantiated cannot instantiate anything else.”
            leads to a the claim that all causal chains must have a starting point.

            There are two plausible way of parsing this premise:

            1) The unordered way: “the elements that are instantiated” refers to a single set of elements, S. In this case the claim just implies the existence of an ancestor function, so that f(B) = A means that A caused B. In this case, the premise is not violated as long as for all x in S, f(x) is also in S. This allows not only infinite regresses, but causal loops as well.

            2) The ordered way: “the elements that are instantiated” refers to a set whose membership changes with some parameter, t — which for simplicity I will refer to as time, but which could refer to some abstract concept of ontological priority or whatever you’d like.

            Then your premise would read — if x is in S(t1), then there exists some t2 less than t1 such that x is not in S(t2), but f(x) is in S(t2.) i.e. x is instantiated at time t, only if it’s ancestor is instantiated before time t. This seems to allow infinite regresses still.

            To wit, we may label the causes in a causal chain by the integers, say that n is caused by n-1, at time n-1/2, and remains instantiated from that time onwards. I don’t see how this scenario violates your premise at all. Each cause is instantiated by another cause that has already been instantiated by something else. If you think this violates your premise 1, I would love to know which of the causes in the chain violates it, because I don’t see any. Remember, they are labeled by the integers, so an answer will take the form: n caused n+1 at time t without being instantiated itself, where n is an integer and t is a real number.

            As a further point, since my intuition tells me we may be at an impasse, I should also point out that you are the one who claims to have a proof. A proof which fails to convince is no such thing. If I don’t see a logical connection between your premise 1 and 3, it is your responsibility to come up with a generally accepted principle which uncontroversially leads from 1 to 3. Preferably, this will take the form of a machine-verifiable substitution rule.

            I would also add that in coming up with such a commonly accepted principle of reasoning, you are doing yourself no favors by using nonstandard terminology like “instantiated,” since it is very unlikely that I accept any principles using such nonstandard terminology.

          • Ray

            VS

            I don’t see how
            “an element that is not instantiated cannot instantiate anything else.”
            leads to a the claim that all causal chains must have a starting point.

            There are two plausible way of parsing this premise:

            1) The unordered way: “the elements that are instantiated” refers to a single set of elements, S. In this case the claim just implies the existence of an ancestor function, so that f(B) = A means that A caused B. In this case, the premise is not violated as long as for all x in S, f(x) is also in S. This allows not only infinite regresses, but causal loops as well.

            2) The ordered way: “the elements that are instantiated” refers to a set whose membership changes with some parameter, t — which for simplicity I will refer to as time, but which could refer to some abstract concept of ontological priority or whatever you’d like.

            Then your premise would read — if x is in S(t1), then there exists some t2 less than t1 such that x is not in S(t2), but f(x) is in S(t2.) i.e. x is instantiated at time t, only if it’s ancestor is instantiated before time t. This seems to allow infinite regresses still.

            To wit, we may label the causes in a causal chain by the integers, say that n is caused by n-1, at time n-1/2, and remains instantiated from that time onwards. I don’t see how this scenario violates your premise at all. Each cause is instantiated by another cause that has already been instantiated by something else. If you think this violates your premise 1, I would love to know which of the causes in the chain violates it, because I don’t see any. Remember, they are labeled by the integers, so an answer will take the form: n caused n+1 at time t without being instantiated itself, where n is an integer and t is a real number.

            As a further point, since my intuition tells me we may be at an impasse, I should also point out that you are the one who claims to have a proof. A proof which fails to convince is no such thing. If I don’t see a logical connection between your premise 1 and 3, it is your responsibility to come up with a generally accepted principle which uncontroversially leads from 1 to 3. Preferably, this will take the form of a machine-verifiable substitution rule.

            I would also add that in coming up with such a commonly accepted principle of reasoning, you are doing yourself no favors by using nonstandard terminology like “instantiated,” since it is very unlikely that I accept any principles using such nonstandard terminology.

  • Martin

    Dorlf,

    The argument is talking about change. So in that case, something or someone changed the ambient temperature or moved the water from one area with a certain temperature to another. So for the water to be liquid and then change to solid, the air temperature must change or something must relocate the water to a different place with a different air temperature.

    • Dorfl

      Here you’re assuming that the water/ice will have ambient temperature to begin with, and so it takes a change in ambient temperature for the water/ice to start changing as well. This is a reasonable assumption, since by the time you look at it, it probably will have. But it’s not anything more than a reasonable assumption.

      To return to the nuclear decay example: say that you have got a lump of uranium floating in a vacuum. There is nothing else for it to interact with. Over millions of years, it will gradually turn into a lump of lead, instead. So potential lead turned into actual lead, without anything external causing the change.

      • eric

        Or better yet, say you have just empty space intead of that lump of uranium. Its going to produce particle pairs. Albeit with brief lifetimes, but they can have significant secondary impact on the world – this production is how black holes decay; they are also responsible for greater than 90% of the mass of the proton.

        • Dorfl

          True.

          Interestingly, if you have empty – or mostly empty – space, it’s also going to expand exponentially as long as the cosmological constant is non-zero. So Quantum Mechanics and General Relativity both predict empty space will behave very strangely.

      • Martin

        And in that case, it is just in the nature of uranium to decay spontaneously. Because of the way it is structured, it will simply decay without anything causing it to do so.

        Let’ say that’s correct.

        So then you are talking about formal and final causes, which are part of the Aristotelian worldview, and rejected by the mechanistic worldview, and thus still confirming that worldview and its associated monotheism. In other words, if that were true, it might be a counterexample to the First Way, but the counterexample just confirms premises from the Second and Fifth Way, and so still leads to the same conclusion.

        • Ray

          Martin

          Not only did you shamelessly put the words “nature” and “structure” into the mouth of your interlocutor, but even if you hadn’t, the mere use of those words does not imply they mean the same thing a Thomist would mean by those words. Merely using the word “nature” does not force one to equate it with telos, as Aristotle did, not does it force one to accept Aquinas’s premise that telos cannot occur in the absence of an intelligence. Likewise, the use of “structure” does not commit one to Aristotle’s hylomorphic dualism. Do you actually have an argument or are you just going to invent words your interlocutors might have said, but didn’t, and overinterpret them?

        • Dorfl

          Again, I’m really not sure what you mean.

          You seem to acknowledge that some things change because they’re changed by something external, and that other things change of their own accord. But you then seem to use the word “nature”, to dismiss the cases of things changing of their own accord as not quite counting, and therefore still allowing you to hold on to the idea that change must be induced by an external changer.

  • DavidM

    I don’t know if it has already been pointed out (I couldn’t be bothered to wade through all the half-baked commentary here), but Edward Feser himself has made a convincing case, based on an analysis of the article above, that “to call Hallquist an incompetent hack would be an insult to incompetent hacks.” Check it out at Feser’s blog!

    • Chris Hallquist

      I’m going to decline. I really need to make a point not to encourage his behavior.

  • baal

    Well, that was all hard to follow. I’m a simple-acist so I’ll try it my way.
    1. Physcical mechanisms happen.
    2. Phase II
    3. therefore god
    I don’t see how throwing any argument about movers, moving or causation really helps phase II. At the end, the whole prime mover argument has a problem jumping from the real world of things in motion upto the supernatural fingertips of a deity pushing metaphorical toy soldier around. I might have heard something lyrical that reminds me of this, “Jump on it, grab it like you want it”. Oh well, back to work. I’m just being vulgar.
    (initially planned segment on large and complex arguments being a sign of failed thinking omitted due to lack of time and authorial skill)

    • Mr. X

      “I don’t see how throwing any argument about movers, moving or causation really helps phase II.”

      Then maybe you ought to try reading some of the cosmological argument’s defenders to see what they say.

  • Chris Hallquist

    Note to all: I just released three comments from the spam filter that had gotten incorrectly marked as spam. I guess sufficiently arcane philosophical discussion looks a lot like the spam of 2012.

    • Annatar

      lulz

  • http://verbosestoic.wordpress.com Verbose Stoic

    Dorfl,

    We’re getting stuck in a long reply chain, so let me pull it out here:

    “In my argument above I used an implicit definition of “A is the cause of B iff B would not have happened if A had not happened.” ”

    That’s pretty much what I’d use, although I wouldn’t have the “iff” in there, because it runs afoul of cases where B is caused both by A and by C, and both A and C happen. Then if A had not happened B still would have happened, but in most cases we still want to say that A caused B (about the only time we really have trouble is if they happen at the same time).

    • Dorfl

      Fair enough. Now that we have an agreed-upon definition of cause, I’ll address your previous post

      “My definition of “dependence” isn’t dependent — pardon the unintentional pun — on time at all. Thus, you can do all the manipulations you want to time and all of the dependencies — and the relevant descriptions — will stay the same.”

      It would probably have been better if I had said “the dependencies are always double-ended”. That is, if B depends on A, then A also depends on B, meaning that if the statement, “B is caused by A” is true, then by our definition the statement “A is caused by B” also ends up being true. It can be a lot less intuitive-sounding, as in my example of “I dropped the egg because it smashed on the floor” or your “my parents engaged in sexual relations because I was born”, but it will still be true.

      “You can escape this by saying that there is no sort of “objective” cause for an event, which is what I think you’re aiming at here.”

      Correct. Or at least, which one of two causally-connected events you consider to be cause and which you consider to be effect is a matter of convention.

      “The problem is that I think it objectively true that, say, I was born at least in part because my parents engaged in sexual relations, and that it is muddled descriptions to argue that my parents engaged in sexual relations because I was born.”

      It’s certainly a muddled description, in the sense that it’s deliberately confusing and not particularly useful except as an illustration of some curious facts about physics – just like my egg-example. But it’s not an incorrect description, which is all that matters for our discussion.

      “Perhaps, as eric aims at, this is a case where our intuitions lead us astray, but I’d need a lot more than that to accept so radical a proposition.”

      It is. In most courses on thermodynamics, some thought experiment equivalent to mine is likely to be used at some point, precisely because the idea of a fall being caused by the landing is so wildly counter-intuitive – even as it’s pretty much trivially true, once you accept that the laws of physics are time-symmetric*. The concept of entropy and the second law of thermodynamics are then used to explain why it contradicts our physical intuition so much.

      I’m not sure what I can do to make you accept that though. I mean, I can’t really do much more than say “trust me, a random stranger on the internet with unverifiable credentials, that this is something which physicists accepted long ago and which is no longer seen as controversial” and possibly point you in the direction of some Wikipedia articles.

      * There are some caveats to this statement, but they don’t really matter for our discussion.

      • http://verbosestoic.wordpress.com Verbose Stoic

        “It would probably have been better if I had said “the dependencies are always double-ended”. That is, if B depends on A, then A also depends on B …”

        Well, that’s not part of my definition, and I don’t thnk it follows from dependence. For me, if B depends for its instantiation on A, that in no way implies that A depends on B for ITS instantiation … and, in fact, it couldn’t be the case because you’d have a circular dependency, with both A and B depending on each other for their instantiation but then with neither of them actually being able to be instantiated (because, basically, the other would have to be instantiated first).

        So, because i don’t think the relation is bidirectional, then I don’t think that you can do the reversal that you’re trying to do. That’s exactly what I meant by a muddled description, because it is technically true that you can say that if B hadn’t happened A wouldn’t have happened if the causal link is tight enough, but that doesn’t translate to A depends on B. So, I see that as an analysis of the form “We know that B happened, and so we know that A happened (since B would not have happened otherwise)” but not one that actual estalishes a dependency relation. To me, in the case we describe we have basically two different relations, in that B depends on A but A has a dependent event B. Cause, to me, basically describes that combination.

        “But it’s not an incorrect description, which is all that matters for our discussion.”

        The problem is that I think that in this discussion, it IS incorrect; that usage of the term is not how we are using the term dependence OR cause.

        “It is. In most courses on thermodynamics, some thought experiment equivalent to mine is likely to be used at some point, precisely because the idea of a fall being caused by the landing is so wildly counter-intuitive – even as it’s pretty much trivially true, once you accept that the laws of physics are time-symmetric*. The concept of entropy and the second law of thermodynamics are then used to explain why it contradicts our physical intuition so much.”

        The question, though, is if the definition used in physics really is the one that’s being used in this argument. I’d argue that the problem here is that physics is using a time dependent definition of cause, and so when they discover time-symmetry they then translate that notion of cause directly using that and get an absurd result for these sorts of questions, and so point out that the mistake that’s being made is thinking that the meaning of the word “cause” is time depedent. I’d need more than physics finding it convenient to decide that this really is the definition as is being used in this philosophical argument; if an argument from physics results in a philosophical and conceptual absurdity, I think it only reasonable to ask physics why it thinks it got it right and we got it wrong.

        • Dorfl

          “Well, that’s not part of my definition, and I don’t think it follows from dependence. For me, if B depends for its instantiation on A, that in no way implies that A depends on B for ITS instantiation … and, in fact, it couldn’t be the case because you’d have a circular dependency, with both A and B depending on each other for their instantiation but then with neither of them actually being able to be instantiated (because, basically, the other would have to be instantiated first).”

          As long as you’re using the definition “A is the cause of B if B would not have happened if A had not happened”, then the bidirectionality of causality follows from the definition – even if you didn’t explicitly include it – because the physical laws of the universe we inhabit happen to be such that you can’t make a non-arbitrary definition of unidirectional causality. It is simply the case that the statement “B would not have happened if A had not happened” is true if and only if “A would not have happened if B had not happened” is also true.

          “The question, though, is if the definition used in physics really is the one that’s being used in this argument.”

          Physicists, in my experience, don’t have a very rigorous definition of cause and effect, since the concept turns out to be of surprisingly little use in most of the situations they have to deal with. It doesn’t matter though, since we already have an agreed-upon definition of cause and effect that we’re using in this discussion. I’ve tried to show that given that definition, your decision to treat one thing as cause and the other as effect is a matter of convenience.

          If I succeed at that, we can then continue to discuss whether a chain of causation can continue indefinitely in both direction, and from there return to whether the universe needs to have a finite past or not.

          “I think it only reasonable to ask physics why it thinks it got it right and we got it wrong.”

          Physicists have done their best to empirically test which laws the universe ultimately obeys. So far, looking at the universe, making tests and thinking about the consequences of what we discover seems to be the best method of learning about the universe. The answers physicists have found put certain constraints on how you can define the concept of causality and still hope to reflect any aspect of reality. Since your definition falls outside of those constraints, I think physics has got it right and you’ve got it wrong.

          I think the reason you get it wrong is that you are basing your reasoning on some ideas that are so deeply intuitive that the human brain nearly treats them as axiomatic. The purpose of my thought experiments was partly to explain why we find those ideas so reasonable and intuitive, even though they are wrong.

          • http://verbosestoic.wordpress.com Verbose Stoic

            Dorfl,

            The problem is that the battle of the intuitions works both ways. I say that you think that dependence and cause works the way you think they do is because you have an axiomatic intuition that they depend on time, which I think is wrong. So when you do physics and find that time looks really strange, you confer that to cause and dependence as well, but I don’t see any real reason to do that. So I don’t accept your demonstration because I don’t think you’ve demonstrated that; we still interpret the terms differently. But we might be able to get a start on how to prove that from here:

            “The answers physicists have found put certain constraints on how you can define the concept of causality and still hope to reflect any aspect of reality. ”

            So, what are those answers? I don’t know of any answers that constrain my view away. I’ve seen the discussions of time, but that’s arguing against a view that is not mine and that I don’t agree with. So, other than the directly controversial cases, do you have an example of a causal relation that seems obviously a cause to everyone but that my time-independent view cannot handle?

          • Dorfl

            “I say that you think that dependence and cause works the way you think they do is because you have an axiomatic intuition that they depend on time, which I think is wrong.”

            I think causality requires the existence of time to be at all meaningful, but I don’t think it depends on the direction of time. The entire point of seeing causality as pointing both ways is to get rid of the dependence on time’s direction.

            “So, what are those answers? I don’t know of any answers that constrain my view away.”

            The laws of physics are such that if a simple* system is in state A at time t1 and state B at time t2, then it will be the case that:

            - The system could not have gotten to state B at t2 without passing state A at t1.

            - The system could not have been at state A at t1 without reaching state B at t2.

            This means that the events (A, t1) and (B, t2) both ‘depend’ on each other. Neither of them requires the other in a way that the other does not also require the first. This means that a definition of causality will be, if not wrong, then at least unaesthetically arbitrary if it treats either event as somehow further down a causal chain.

            * If the laws of physics are non-deterministic, then the situation changes a bit for complicated systems during some interval of time, but not in a way that matters much for our discussion. See my response to Andrew above for a slightly longer discussion.

            ” So, other than the directly controversial cases, do you have an example of a causal relation that seems obviously a cause to everyone but that my time-independent view cannot handle?”

            How about this?

            At time t1 a perfectly elastic ball bounces off of a heavy obstacle and at t2 collides with another ball of equal mass, stopping completely and sending the other ball flying off in the same direction. Did the bounce off the obstacle cause the collision, or did the collision cause the bounce?

          • http://verbosestoic.wordpress.com Verbose Stoic

            “I think causality requires the existence of time to be at all meaningful, but I don’t think it depends on the direction of time. The entire point of seeing causality as pointing both ways is to get rid of the dependence on time’s direction.”

            And I get rid of that dependence by pointing out that the only reason we thought time and the direction of it as being so critical to what it means to be a cause is because in this universe in almost all of our examples the cause precedes the effect, making us think that the cause HAD to precede the effect. But conceptually this isn’t required, and then when you have the issues with the direction of time that come from physics I think the easiest way to solve the problem of the directedness of time is to recognize that that whole project was wrongheaded to start with.

            “This means that the events (A, t1) and (B, t2) both ‘depend’ on each other. Neither of them requires the other in a way that the other does not also require the first. This means that a definition of causality will be, if not wrong, then at least unaesthetically arbitrary if it treats either event as somehow further down a causal chain.”

            But I disagree with this. If it is NOT the case that one of them depends on the other in a way that the other does not, then I submit that you don’t have a cause at all. My view here is that you get to this discussion by only considering casses where the causal link is very, very strict, and that in general if A occurs then B MUST occur. But this falls apart in cases where B may be caused by other causes, or where things can interfere between A and B so that if A happens B still doesn’t happen. The only way to make this dependency, it seems to me, is to interpret “cause” as saying “A can be said to cause be of it can be said that A happens iff B happens”. But that last part actually rarely happens in reality, and so we generally can see many cases where A happens and B does not, and where B happens and A does not, and yet we can clearly identify specific cases where A did cause B, and where B’s occurrence happened because of A’s occurrence and interaction, and yet it is not the case universally.

            “How about this?

            At time t1 a perfectly elastic ball bounces off of a heavy obstacle and at t2 collides with another ball of equal mass, stopping completely and sending the other ball flying off in the same direction. Did the bounce off the obstacle cause the collision, or did the collision cause the bounce?”

            The bounce caused the collision. It is ridiculous to suggest that the collision caused the bounce since if I either moved the other ball out of the way or caught the “causing” ball the bounce would still have occurred, and it would merely be the case that it would have another effect. See, this is one of those cases that we aren’t agreeing on; your example was suppsoed to be obviously uncontroversial [grin].

          • Dorfl

            “My view here is that you get to this discussion by only considering cases where the causal link is very, very strict, and that in general if A occurs then B MUST occur. But this falls apart in cases where B may be caused by other causes, or where things can interfere between A and B so that if A happens B still doesn’t happen.”

            Hmm…

            I think we’re partly talking past each other, since I’ve been using A and B to refer to specific events, while you seem to use them to refer to classes of very similar events.

            To use the example of a smashing egg again: For that particular egg to smash on the floor in that particular way, required me to drop it in the way I did at that particular time. Nothing else could have caused that one egg-smash. Conversely, once I dropped the egg the way I did, that particular egg-smash had to happen. That’s why I consider the causal link to be so strict. Assuming determinism it is completely strict. Granting some amount of non-determinism, it will generally be equally strict in either direction (as long as the entropy is maximal, which it is for the majority of imaginable situations).

            You on the other hand would – if I understand you correctly – look at the more general class of “egg-drops” and “egg-smashes” and note that not every egg-smash is preceded by an egg-drop, and not every egg-drop leads to an egg-smash, we cannot make a strict causal link between egg-drops and egg-smashes.

            There is nothing wrong with looking at things that way. Like I said, it’s definitely the most practical way of doing things. But I don’t think it’s really relevant to this discussion, since we’re discussing how cause and effect works at the most fundamental level. And at that level, we cannot lump two egg-smashes together as being basically the same, no matter how similar they are for practical purposes – such as cleaning the kitchen.

            “The bounce caused the collision. It is ridiculous to suggest that the collision caused the bounce since if I either moved the other ball out of the way or caught the “causing” ball the bounce would still have occurred, and it would merely be the case that it would have another effect.”

            But what I didn’t tell you was that t2 happened before t1. So when you thought you were catching the ‘”causing” ball’ you were actually throwing it at the stationary ball, knocking it into the obstacle and causing it to bounce.

            Does this clarify the problem? When determining the causal relationship, you still ended up looking at the time-ordering implied by my choice of the terms t1 and t2 to label the events, and based on that concluded that the event at t1 must be causing that at t2.

  • Josh

    Hey Hallq. I read Feser’s response to this post just now.
    (http://edwardfeser.blogspot.com/2012/11/the-incompetent-hack.html)

    Damn, what an absolute demolition of your argument. I had heard Feser was a good writer. I just didn’t think he could be so devastating.

    Hope you’re doing okay.
    Josh

  • Orangebookbag

    Laws of physics. Always there to come in at the last moment and save us from Thomas’ Five Ways.

    Have you ever studied philosophy of science?


CLOSE | X

HIDE | X