Friday, February 5, 2016

Parfit on brute facts


Derek Parfit’s article “The Puzzle of Reality: Why Does the Universe Exist?” has been reprinted several times since it first appeared in the Times Literary Supplement in 1992, and for good reason.  It’s an admirably clear and comprehensive survey of the various answers that have been given to that question, and of the problems facing some of them.  (Unsurprisingly, I think Parfit’s treatment of theism, though not unfair, is nevertheless superficial.  But to be fair to Parfit, the article is only meant to be a survey.)

Parfit appears to sympathize with the “Brute Fact View” according to which the universe simply exists without explanation, and that’s that.  The claim here is not that there is an explanation but that we don’t and even can’t know what it is.  It is rather that there is no explanation at all, no intelligibility, rhyme or reason to why this universe exists rather than another or rather than nothing at all.  This is, of course, implicitly to deny the Principle of Sufficient Reason (PSR), according to which everything does have an explanation, whether or not we can always discover what that explanation is.  I’ve defended PSR and criticized the Brute Fact View in several places, such as in Scholastic Metaphysics.  (Also in several earlier blog posts, to which you’ll find links below.) 

Parfit describes and defends the Brute Fact View in the following passage:

[On] the Brute Fact View… we should not expect reality to have very special features, such as being maximal, or best, or having very simple laws, or including God. In much the largest range of the global possibilities, there would exist an arbitrary set of messily complicated worlds. That is what, with a random selection, we should expect. It is unclear whether ours is one such world.

The Brute Fact View may seem hard to understand. It may seem baffling how reality could be even randomly selected. What kind of process could select whether time had no beginning, or whether anything ever exists? But this is not a real problem. It is logically necessary that one global possibility obtains. There is no conceivable alternative. Since it is necessary that one possibility obtains, it is necessary that it be settled which obtains. Even without any kind of process, logic ensures that a selection is made. There is no need for hidden machinery.

If reality were randomly selected, it would not be mysterious how the selection is made. It would be in one sense inexplicable why the Universe is as it is. But this would be no more puzzling than the random movement of a particle. If a particle can simply happen to move as it does, it could simply happen that reality is as it is. Randomness may even be less puzzling at the level of the whole Universe, since we know that facts at this level could not have been caused.

End quote.  Parfit’s argument here seems to me highly implausible and problematic.  For one thing, he seems to allow at least for the sake of argument that there might be a kind of “process” which “selects” whether anything exists etc. but in a “random” way that is not ultimately explicable.  This is a very odd suggestion for a couple of reasons.  First, why bother with it?  If you’re going to commit yourself anyway to the idea that the universe is just an unintelligible Brute Fact, why not simply say that the universe just exists and that’s all that can be said and leave it at that?  Why posit, between the universe on the one hand and sheer Bruteness on the other, some intermediate “process” of “selection” which in some sense accounts for the existence of the universe but itself operates in an unintelligible way?  What’s the point of positing such a “process” in the first place if one doesn’t think that it or anything else can do any real explanatory work where the sheer existence of the universe is concerned?

Second, why call something a “process” which functions to “select” the universe if one thinks it is not something whose operation is ultimately intelligible?  Other things we call “processes” are not like that, including processes that involve an element of chance.  For example, the way a population is molded by natural selection is a kind of process, and chance plays a role, but that does not make any of its results unintelligible.  Given such-and-such a variation within a certain population (larger beaks in certain birds within a group of birds, say) under such-and-such environmental circumstances (hard seeds being the main local food source), it is perfectly intelligible why there would be a change in the population (the larger sort of beak would be much more common in later generations of birds). 

(As Aquinas argues, chance always presupposes the convergence of lines of causation which are not the result of chance.  To take a stock example, when a farmer finds buried loot while he is out plowing his field, that is a chance occurrence.  But that a robber decided to bury his loot there and that the farmer decided to plow the field that day were not chance occurrences.  All chance occurrences are like that in that they resolve themselves, at some level, into a convergence of non-chance occurrences.)

So, what we ordinarily describe as “processes” of “selection” are intelligible even when they involve an element of chance.  So why call what Parfit is describing -- something which is chance all the way down, as it were, and the operation of which is not intelligible -- a “selection process,” or indeed a “process” of any kind? 

So, the stuff about “how reality… [is] randomly selected is one problematic aspect of Parfit’s view.  Then there is the suggestion in the second half of the second paragraph quoted, to the effect that logic itself essentially solves any apparent problem with the Brute Fact View.  Again, Parfit says: “Since it is necessary that one possibility obtains, it is necessary that it be settled which obtains. Even without any kind of process, logic ensures that a selection is made. There is no need for hidden machinery.”

To see what is wrong with this, suppose police come across a dead body and start batting around possible explanations -- murder, suicide, accident, heart attack, etc.  Suppose one of the policemen who has heretofore been silent interrupts and says: “I don’t know why you guys are wasting time considering these different explanations.  I say it’s just an unintelligible, inexplicable brute fact that this corpse turned up here and now.  Case closed, we can go home now.  Don’t raise your eyebrows!  After all, it’s necessary that some possibility had to obtain here and now, so it’s necessary that it be settled which one obtains.  Even without a murder, or suicide, accident, heart attack, etc., logic ensures that a selection is made. There is no need for ‘hidden machinery’ such as murder, accident, etc.”

No one would accept this for a moment, of course.  That “logic ensures” that some possibility or other will obtain simply does not make it the least bit plausible to say that we needn’t bother asking how exactly this particular possibility -- a corpse, and this corpse, here and now -- got to be the one which obtains.  Now, Parfit gives us no reason at all to believe that this sort of move is any more plausible when we are asking “Why does the universe exist?” than it is when we are asking “How did this corpse get here?”  So, his attempt to appeal to logic in order to make the Brute Fact View believable fails.

(Indeed, it is strange that Parfit would take this suggested defense of the Brute Fact View seriously given what else he says in the article.  In particular, at the beginning of the article he is critical of attempts to dismiss the need to explain the initial conditions of the universe that allowed for stars, planets, and life, on the grounds that there had to be some initial conditions or other.  The fact that there had to be some initial conditions or other doesn’t remove the need for an explanation, Parfit argues, because the specific initial conditions that happened to have obtained are so improbable.  Now, if saying “There had to be some initial conditions or other” is by Parfit’s own admission not a plausible way to dismiss the request for an explanation of why we have a universe capable of supporting life, etc., then why is saying “Logic ensures that some selection has to be made” a plausible way to dismiss the request for an explanation of why anything exists at all rather than nothing?)

A third issue raised by Parfit’s remarks is the stuff about the random behavior of particles, and what Parfit has in mind here are, of course, claims to the effect that quantum mechanics has shown that events can occur without a cause.  I’ve discussed this issue at length elsewhere (e.g. in this post) and won’t repeat here everything I’ve said before.  Suffice it for present purposes to note that when Parfit says that “if a particle can simply happen to move as it does, it could simply happen that reality is as it is,” he is overlooking a crucial disanalogy between quantum theory and the Brute Fact View, and one that should be obvious.  No one claims that the motion of the particles in question is simply unintelligible.  They don’t say “they just move and that’s that and nothing more can be said.”  Rather, they say that (what they call) the random motion of particles is something which it makes sense to think of as occurring given quantum mechanics.   The theory provides an explanatory context that makes the behavior of the particles intelligible even if their motion is said to be in some sense “uncaused.”  (Hence the motion isn’t “random” full stop, without qualification.  If you’re giving a theoretical description of some “random” phenomenon which gives it a kind of intelligibility, then you are ipso facto using “random” in a qualified sense.)

By contrast, the Brute Fact View denies precisely that there is any larger explanatory context within which the “random” “selection” of the existence of the universe can be made intelligible.  It says that the universe just exists and that’s that and nothing more can be said.  There is no larger background theory in the context of which such a “random” occurrence makes sense.  So there just isn’t any parallel here with quantum mechanics.  Hence, whatever it is Parfit and others think quantum mechanics has established, it simply lends no plausibility to the Brute Fact View.

Finally, there is Parfit’s remark that “randomness may even be less puzzling at the level of the whole Universe, since we know that facts at this level could not have been caused.”  Those who haven’t read Parfit’s entire article might wonder whether he is blatantly begging the question when he says that “facts at this level could not have been caused.”  For isn’t the claim that such facts are caused precisely what theism says?  But Parfit is not ruling out theism a priori here.  Rather, his remark here must be understood in light of what he says at the very beginning of the article, where he says:

[T]hings might have been, in countless ways, different. So why is the Universe as it is?

These facts cannot be causally explained. No law of nature could explain why there are any laws of nature, or why these laws are as they are. And, if God created the world, there cannot be a causal explanation of why God exists.

So, Parfit is not ruling out arbitrarily the claim that God is the cause of the universe.  Rather, he is saying that even if God is the cause, God’s own existence would not have a causal explanation and thus would have to be explained in some other way.  (The traditional answer is that it is God’s nature as that which is purely actual, subsistent existence itself, absolutely simple or non-composite, etc. that explains his existence in a non-causal way.) 

So, Parfit’s point is that causal explanations, specifically, cannot be the ultimate sort of explanation, so that if there is to be an explanation of an ultimate sort it will have to be an explanation in something other than causal terms.  And he is right about that.  So, when he says at the end of his essay that “facts at [the level of the whole Universe] could not have been caused,” he is just alluding to the point made at the beginning of the essay that ultimate explanations cannot be of a causal nature specifically.

So far so good, then.  The problem is with what Parfit seems, at the end of the essay, to think follows from this point.  Again, he says that “randomness may even be less puzzling at the level of the whole Universe, since we know that facts at this level could not have been caused.”  That is to say, from the (true) premise that ultimate explanations cannot be of the causal type, Parfit appears to derive the conclusion that it is plausible that the fundamental facts about the universe might be “random.”  Well, that conclusion simply doesn’t follow from the premise.  In particular, from the premise that “X does not have a causal explanation” it simply doesn’t follow that “X is random,” or even that “X is plausibly random.”  That would follow only if the only plausible alternative to causal explanation is an appeal to randomness.  And that isn’t so.  Something that lacks a causal explanation could have an explanation instead in terms of its own nature, say, or by virtue of being a necessary truth.  The fact that 2 + 2 = 4 does not have a causal explanation but it is hardly “random” that 2 + 2 = 4.  When Thomists argue that God’s existence follows from his being pure actuality, subsistent being itself, absolutely simple or non-composite, etc., they are not saying that his existence is “random.”  On the contrary, they are saying that his existence follows necessarily from his nature so understood.  And so on.

Of course, an atheist would criticize the concepts of pure actuality, subsistent being itself, etc.; someone who denies the objectivity of mathematical truth might challenge the claim that it is in any interesting sense a necessary truth that 2 + 2 = 4; and so forth.  But none of that is to the present point.  The point is rather that the claim that ultimate explanations are not causal explanations simply does not by itself lend any plausibility at all to the Brute Fact View, contrary to what Parfit implies. 

Anyway, even apart from the problems with Parfit’s account of it, we can know the Brute Fact View is false, because we can know that PSR is true.  Again, see Scholastic Metaphysics, and the first several of the posts listed below.

Related posts:







130 comments:

  1. Professor Feser,
    Does he ever define "random"? A great many people seem to use the word in a purely negative sense (happening without any sort of nature whatsoever) as if the existence of the mathematical concept with the same name somehow justifies treating their own sense of the word as not only meaningful, but referring to something we're all familiar with. But when "random" is used to mean "happening without any sort of nature or cause", this seems to me to bear an affinity to the New Atheist technique of making things more palatable by giving them names. Has Parfit, or has anyone for that matter, ever defined "random" in a way that is not simply a disguised denial of the validity of reason?

    ReplyDelete
  2. Most non-statistical types do not actually understand what is meant by random.

    ReplyDelete
  3. TOF,
    Do you mean that they're referencing the mathematical concept of random, but badly? I've always been under the impression that they were using a different concept which was inspired by the mathematical concept. Roughly that they took the layman's explanation, "we don't know what the outcome of an individual event will be," then said, "wait a minute... what if no one could!" and sort of went from there to "random" meaning "no one can, even in principle, know what the outcome will be, or afterwards, why it was". Or alternatively that they started with pseudo-random number generators in computers, then generalized to, "a thing which actually, completely has the properties we want a pseudo-random number generator to have". I might be giving people too much credit, though...

    ReplyDelete
  4. In every day life, in the natural sciences, it is certainly true that "The vast majority of things have an explanation; and when no explanation is known, then an unknown explanation is in the vast majority of cases much more likely than an absent explanation". That could be seen as a kind of weak PSR, a kind that works for everyday life and most uses in the natural sciences, yet still permits the atheist to say "The universe as a whole has no explanation and doesn't need one" (thus defeating the cosmological argument) - rather than an absolute rule, it accepts violations in exceptional cases, and arguably the existence of the universe as a whole is a vastly more exceptional case than your everyday murder or particle physics experiment. I'm a theist, and I believe God caused everything beyond himself, so that belief implies a much stronger PSR (probably about as strong as Professor Feser's). But, I don't believe I have any reason to believe such a stronger PSR apart from theism. The cosmological argument tries to argue "PSR implies God"; I reject cosmological arguments because I believe instead that "God implies PSR". Professor Feser claims we have reason to believe in strong PSR apart from the truth of theism; I'm not convinced on that point.

    ReplyDelete
  5. Finally giving up on the hope of 'Scholastic Metaphysics' every coming out in Kindle, I have ordered the paper book ($46 delivered to New Zealand - sob!).

    The reason is that I think I don't really understand what is meant by a 'brute fact' view. I recall - very vividly and very painfully - a time, in 1984, when, suddenly, out of nowhere, I 'realised' (as I would have put it) that there is no God. The world just was.

    But what seemed to me to come from this was not even something like Parfit's 'selection' business - which seems to me precisely to say there is an explanation, though we cannot know it. What seemed to me to come from this was ... simply silence. If what is, is - and that's that - then ... Well, I don't know, it just seemed to me that was the end of my saying anything - about reality or ... well, or saying anything at all.

    I recovered (as I would call it) from my attack of atheism. I decided that if there were no God, then my imaginary world in which God exists was superior to the real world - and that I refused to act in terms of such a world.

    Not very logical - but within a few weeks I was a believer again :-)

    jj

    ReplyDelete
  6. (As Aquinas argues, chance always presupposes the convergence of lines of causation which are not the result of chance. To take a stock example, when a farmer finds buried loot while he is out plowing his field, that is a chance occurrence. But that a robber decided to bury his loot there and that the farmer decided to plow the field that day were not chance occurrences. All chance occurrences are like that in that they resolve themselves, at some level, into a convergence of non-chance occurrences.)

    Chance here is meant to be different from indeterminism, correct? It is pointed out on p. 122 of Scholastic Metaphysics that the scholastic principle of causality does not require causes to determine their effect, but only that effects, when they occur, have causes.

    So perhaps the fact about, say, radioactive decay is that particles are radiated stochastically. Their radiation is still caused by the atoms from which they irradiate but need not be a convergence of underlying non-chance occurrences. Thus their radiation is caused and indeterministic, but not a chance occurrence.

    That does mean that chance has a somewhat special sense here. Is it obvious that Parfit and Aquinas are using it in the same sense?

    ReplyDelete
  7. @Chris: something like that.

    Variation in a process is of two basic sorts: assignable variation and random variation. Assignable variation is that which can be assigned to a particular singular cause. For example, if a pair of regular dice is recorded to show "13" we know there must be a particular cause, and diligent investigation will uncover it.

    OTOH, if the dice show "12," there is no particular reason. This "random" variation is due to a host of small causes acting concurrently, no one of which is dominant. Consequently, if one of these causes is controlled, the result will still appear with some frequency due to the combined effects of all the other causes. The random variation is due to the entire cause system, but it is not uncaused.

    If the effects of these common causes are additive (and a linear sum is often a good approximation) the net effect will be approximated by the so-called Normal Distribution. (If they are multiplicative, the result is a Lognormal Distribution; if polynomial, an Extreme Value Distribution; and so on.) But the variation is never uncaused.

    What many people mean when they say "uncaused" is really "unpredictable." This is another matter entirely. Newton's theory cannot tell us which apple will fall from which tree and when; but that the apple's fall is caused by gravity remains true -- whatever the heck we mean by "gravity." (There's that conceptual confusion again...)

    ReplyDelete
  8. The universe as a whole has no explanation and doesn't need one

    What is the explanation for the MoonEd? This is the merelogical sum of the Moon and Ed Feser. Of course, the Moon has an explanation and Ed has an explanation; but the set U={Moon,Ed} has no explanation. This is because the MoonEd is not a thing.

    The universe is likewise not a thing. It is simply the set U={X|X has physical existence}. The explanation for the universe is "Things Exists." But the universe is not some vast box that exists independently of the Things in it.

    ReplyDelete
  9. @TheOFloinn: But how then does one define a "thing"? What makes some mereological sums count as "things", and other mereological sums not count as "things"?

    Consider a tree. At any moment, the tree contains some number of atoms. (There can be some ambiguities about atoms on their way in or out of the tree, e.g. as the tree absorbs a CO2 molecule from the atmosphere, at which point does its atoms become part of the tree? However, let us put them aside for the moment, and suppose that, at any given moment, every atom can be definitely assigned as either belonging or not belonging to that particular tree.) Now, the mereological sum of all those atoms is that particular tree at that particular moment. If I understand you right, you are saying that mereological sum is a valid "thing", but the mereological sum of all atoms (and subatomic particles outside any atom and fields and so forth) in the physical universe is not a valid "thing". What criteria determines which mereological sums are valid "things" and which are not?

    One distinction between "MoonEd" and a tree or the Moon or Professor Feser or the entire physical universe, is that the former is an essential arbitrary mereological sum, while the others are far less arbitrary. A tree, the Moon, Professor Feser - these mereological sums all possess a certain unity and coherence which "MoonEd" lacks - but surely, in terms of unity and coherence, the physical universe as a whole is closer to tree/Moon/Feser than to "MoonEd".

    ReplyDelete
  10. @TOF: I prefer the definition of random which is a measurable function from a probability space to the real numbers (mostly because it defies misuse as a metaphor to the real world).

    When applied to genetic mutations, I've always liked, "uncorrelated with anything we have (so far) measured".

    But I do think that, as a matter of psychology, many people consider the abstraction of a random process in mathematics, then use that with little modification as a sort of embodied abstraction in the real world, as if a thing so under-specified could actually exist.

    But I do think that they mean something more than unpredictable—I should probably say something less—because for example the choices made by a free will are not predictable, but neither are they random in this popular sense. To be random, a thing must lack not only predictability, but also a nature. Probably also any relationship to its own history, if that's not redundant to it lacking a nature. It may even be required to have an even distribution to its outcomes (like a computer's pseudo-random number generator), though again this may be redundant to it not having any sort of nature. Of course such a specification precludes the thing from actually existing, but these days people are used to believing up to six impossible things before breakfast. :)

    ReplyDelete
  11. Secular philosophy, whether popular or professional, invites us towards enthusiastic participation in a mortal lifetime displayed within a scheme of dreadfully indifferent powers and laws and forms. The universe of science is a mighty pool of senseless rocks, winding like clockwork, shedding a soup of poor beings cited with the demand for yes and yes, until, in some terminal collision, they are made soundless in the molecular roar.
    Yet, if human beings or their politics act in a manner like the universe they suppose, there are cries against barbarism, totalitarianism, and unlove. How can we justify and demand superior behavior on the part of humankind if we conceive of humankind merely in terms of the mortal ego and the secular cosmos?

    If humankind does not achieve a higher Sacred and Transcendental view of self and world and Reality, then there is no real alternative to the MAD gleefulness of the proponents of conventional exoteric religion and the sorry revolt of angry secular mortals.

    With rare exception all of the proponents of conventional religion are just as much convicted of the same entirely mortalistic model of existence. Except that they attempt to add an idealistic gloss to the same mortalistic vision - aint "creation" wonderful! - "God" made all of this - "God" has a plan for humankind - you go to "heaven" when you die.

    Sacred philosophy is, when free of its historical conventions, a rigorous intuitive and intelligent discipline that examines and transcends the conventional presumptions of ordinary, or popular, and professional secular and religious knowing and experiencing. If we are limited to the model of existence, knowledge, and experience proposed by both secularism and conventional exoteric religion, then we are merely projected into the mortal domain of limits and possibilities bereft of Ultimate Paradox, Mystery, and Truth.

    ReplyDelete
  12. The word random has so much baggage, and I hear atheists use it all the time to show that the universe is purposeless. Evolution is often described as a largely random process, and it's the baggage that comes with the word random that leads many to the interpretation that evolution is purposeless. I'm sure Ed would bring up the difference between qualitative and quantitative analysis here. Anyways, instead of using the word random I prefer using words like "unpredictable" and with regards to evolution I prefer "stochastic." Stochastic has no such baggage, and sounds cooler.

    The immune system is an example of a stochastic process at work. Your immune system only has a limited number of antibodies and every time antigens enter your body your immune system starts generating a large number of random variations that are then selected upon once it finds antibodies that bind to the antigens. That's how your immune system, with such a limited number of antibodies, can recognize a very large number of different antigens. Would we say that this "random" process has no purpose?

    "Random" methods, like Monte Carlo methods, allow you to optimize something by randomly flipping a certain thing and then selecting on it. These methods are very effective when it comes to solving very complicated problems and they're used in physics, finance, engineering, computer science, etc. Ard Louis (who is a renowned professor of theoretical physics at Oxford University as well as a devout Anglican) puts it well here: "Imagine I give you a fully formed train made of Lego blocks, you might think it's cool, but if I take a box and fill it with Lego blocks and shake it and then out comes a fully formed train, even if that train has a few scratches on it and one or two pieces missing it would still be much cooler."

    I think it's time for atheists to stop using the word random as if it somehow cuts against theism, especially considering most of the time they don't even know what it means. I have noticed that many atheists these days have given up on the brute fact hypothesis and have adopted the multiverse hypothesis. Indeed, the confidence that multiverse supporters have in the theory being true is so high to the point where you get the sense that they've already been to these other universes. I personally think they've seen too many episodes of Sliders.

    ReplyDelete
  13. Never mind the philosophy. The important question is where are you getting those comic book panels?

    The Chicken

    ReplyDelete
  14. Further to JTJ's post above about "Scholastic Metaphysics," Father Schall puts the book on the latest of his periodically issued reading lists: http://www.crisismagazine.com/2016/on-catholic-intelligence

    ReplyDelete
  15. how then does one define a "thing"?

    Things are objectively delineated, identifiable and countable. They are discovered in nature; they are not simply defined. Paraphrasing Alexander Pruss:

    The MoonEd is not objectively a thing. It's not a thing because it is two things. Now, you can say that in one way the MoonEd is one thing and in another way it is two things. But this in identifying the things. Because the things are "objectively delineated and objectively countable," we would need an objective fact of the matter about whether the MoonEd is one thing or two things. It can't be both. So on the scientific grounds of what lends itself better to explanatory purposes is it objectively better to talk of the MoonEd as two things rather than as one.

    So there is certainly one thing we can say about the things. "No thing can be a mereological sum of other things." A heap of sand, then, is not a thing, for it is nothing but the mereological sum of the grains of sand. It is a heap, not a thing.

    But is not Ed simply a heap of atoms? No, because the atoms are constituted by the formal cause of Ed Himself. That is, Ed has a substantive being that a heap of sand does not. The Atoms of Ed operate as part of a whole while the grains of sand operate independently under the influence of external forces. Likewise, a carbon "atom" is a thing, because its parts are subordinate to a whole. An electron in the outer shell of the atom, for example, does not act as a free electron, but as part of a thing. The grains of sand in the heap may also be things, if we can regard the crystaline structure as giving unity to them.

    The Greek “ousia” and the Latin “substantia” can be used for “thing” if you like. "We discover the things and do not create them in the way our minds create the [MoonEd]."

    BTW, it's not always clear. Artifacts may be things even though their parts do not have a natural tendency to work together and subordinate themselves to a whole. And some natural things may be ambiguous, esp. plants. When my neighbor cut down a tree bordering his fence, the roots, which ran underground onto my own property, immediately put up shoots that, unmolested, would have grown up into their own trees. So was there one thing that perished or were all of them really a single thing? There is an Aspen grove in (iirc) Colorado named Pondo in which every tree is a clone of the same root system. Is Pondo one thing or many? Plants are weird and live in a way very different from animals.

    Hope this helps.

    ReplyDelete
  16. Hi Ed,

    I agree with your critique of Parfit's position, which strikes me as philosophically incoherent - especially with regard to Parfit's remarks on randomness.

    Having said that, I think your own formulation of PSR is too strong: every actualization of a potential requires an explanation. I would argue that since every contingent choice (including Divine choices) involves an act of will directed at some goal that need not have been chosen, it must involve the actualization of some potential. It is irrelevant whether the choice is made inside time (as with creatures) or atemporally (as with God), and it is likewise irrelevant whether the object chosen enhances the status of the agent (by conferring some additional perfection). The point is that by choosing something that you need not choose - as God does when He freely decides to create and maintain the world - you actualize your will in a way that it would not be actualized, had you not made that choice.

    Although the world is not a brute fact, God's choice to make it is "brute" in the sense of being inexplicable, although it of course presupposes the fact of God's existence. Choices have necessary conditions, and that's all we can say. The actual choices themselves don't seem to require an explanation, however.

    Even in the natural world, many actualizations of underlying potentials are doubtless accidental. It therefore seems safest to restrict PSR to the existence and the essential properties of things, since these cannot be described as accidental.

    Re PSR and God: it strikes me that an atheist (or even a theist) might try to argue that the notion of a self-explanatory being is philosophically absurd. For suppose there were such a Being. You have argued that it would have to be Pure Act: its whole Being is identical to the action that defines its essence. Now ask yourself: what kind of action could possibly be self-explanatory?

    The two actions which are supposed to characterize God's essence are knowing (or thinking) and loving. What does it mean to say that some act of thinking explains itself, or that some act of love explains itself? Suppose there is some such thought or act of love: what is its object? Itself? So the thought of X explains the thought of X, or the love of X explains the love of X? Sounds circular to me.

    It has been many years since I read it, but I seem to recall that in his book, Beyond the New Theism, Germain Grisez opted for a weaker version of Divine Necessity than the one you propose. He was persuaded by Hume's argument that any being (including God) can be conceived as non-existent. What Grisez argued for was a Being (D) which is such that, if it exists, it requires no further condition for its obtaining than the fact of its own existence. Grisez's God is thus independent, but not self-explanatory. What is your reaction to this position?

    TOF:

    You wrote above that the universe is not an object. I respectfully disagree. It is not just the set of all contingent beings. It is a unified system of things subject to the same laws and resulting from the same set of initial conditions. In fact, one cosmologist whom I spoke to many years ago defined the science of cosmology as the attempt to treat the universe as if it were a single object.

    ReplyDelete
  17. Vincent, "as if it were" is a counterfactual.

    Also, if the universe were a single substance, you would not be one.

    ReplyDelete
  18. Vincent:

    "You wrote above that the universe is not an object."

    Actually he wrote that the universe is not a thing (meaning, I take it, a substance). That is, or may be, very different; depending on the precise sense of "object," lots of non-things can be objects. I can make the universe the "object" of my thought, for example, without thereby implying that it's a single thing/substance.

    ReplyDelete
  19. Incidentally, you seem to agree, since you write It is a unified system of things without (so far as I can tell) intending to suggest that a "system of things" is a "thing" in its own right.

    ReplyDelete
  20. @Simon Kissane: "But how then does one define a 'thing'? What makes some mereological sums count as 'things', and other mereological sums not count as 'things'?"

    Luckily these questions, like so many other questions, are immensely clarified by Heidegger.

    ;-)

    ReplyDelete
  21. I seriously think, professor Feser, that while you are definitely correct that modern, contemporary philosophy (as taught in the government controlled institutions) needs to reacquire the distinction between potency and act; nevertheless, I think - vis-à-vis, for example, trope theory - the recovery of substance might be a better place to start (and I mean substance especially as opposed to accident - which is, of course, related to potency and act).

    I mean, anyone who has read the arguments Aristotle gives around the PNC in his Metaphysics knows at least that he strongly believes it is simply impossible to coherently deny the difference between substance or accident (the consequence, according to him, being that anything means anything and, necessarily of course, there is no meaning at all). That is of course impossible; otherwise, atomism, materialism, etc., would mean nothing and that God exists and that ad infinitum absurd and contradictory implications follow.

    ReplyDelete
  22. "That is of course impossible; otherwise, atomism, materialism, etc., would mean nothing and that God exists and that ad infinitum absurd and contradictory implications follow.

    I don't think it would be necessarily impossible, but no one could argue for it, or believe it. One could not even hold to radical skepticism. All beliefs would completely lack coherency.

    ReplyDelete
  23. Well,I have a question and I sincerely expect a proper answer.What if at the fundamental level of physical reality there are some forms of existence,that are in a mixture of actuality and various potentialities that are grounded in its actuality?Why thomists assume that its actuality from which all the potentialities arose is contingent?What is needed is an Actualizer that actualises its potentialities.So why must its actuality be necessarily contingent?

    ReplyDelete
  24. @Mikael:

    Well, there will likely be better answers in the morning, but as I'm up, let me start.

    "Well,I have a question and I sincerely expect a proper answer."

    Yes, indeed. I will attempt the very thing. :)

    "What if at the fundamental level of physical reality there are some forms of existence,that are in a mixture of actuality and various potentialities that are grounded in its actuality?"

    If there is a "mixture of actuality and various potentialities" that is *grounded* in an actuality, then that mixture isn't really fundamental--the actuality is. Furthermore, Thomists have good arguments that the physical itself cannot be fundamental, as the physical is both act and potency.

    "Why thomists assume that its actuality from which all the potentialities arose is contingent?"

    I hope I'm reading you right here. Thomists do not assume that the act from which all potentialities arise (so to speak) is contingent. They argue that it is necessary. (And of course, not really an "it," but that is another matter.)

    "What is needed is an Actualizer that actualises its potentialities.So why must its actuality be necessarily contingent?"

    The act, it is argued, is not contingent--it is necessary.

    ReplyDelete
  25. @TheOFloinn: If one accepts an Aristotelian-Thomist ontology as one's starting point, I take it that what you have said is correct. However, if one chooses to adopt a competing ontology instead, the same conclusions don't follow. Many (even most) atheists nowadays adopt a reductionist materialist ontology in which there is no substantial distinction between a person and the mereological sum of the atoms which constitute their body at any given moment, and given those premises your conclusions fail to follow. (I myself adopt instead something like an idealist ontology, in which the fundamental constituents of reality are minds and their contents; I don't think some of your specific conclusions hold using my premises either.)

    ReplyDelete
  26. Thanks laubadetriste.I am new to metaphysics and I'm a bit confused.

    ReplyDelete
  27. I also have another question:isn't local motion the only real change that takes place in our Universe?Because when we talk about water freezing or boiling,we are after all talking about molecules moving from place to place.

    ReplyDelete
  28. Dear Dr. Feser,

    I am a regular reader of your blog. I write from Portugal, and I came across the work, recently, of David Benatar. If time and circumstances permit, could you dedicate a post commenting on his work? I think Benatar´s main thesis runs against one of the core Thomistic thesis, if I understand it correctly, that "to be (to exist) is good".

    Thank you for your time and work.
    Best regards,

    João Gabriel
    Porto, Portugal

    ReplyDelete
  29. @Mikael: "Thanks laubadetriste.I am new to metaphysics and I'm a bit confused."

    No worries. "All that I have written seems like straw to me."

    "I also have another question:isn't local motion the only real change that takes place in our Universe?Because when we talk about water freezing or boiling,we are after all talking about molecules moving from place to place."

    Well, Thomists would deny that local motion is the only real change that takes place. So too would the preponderance of non-Thomists. Dwelling on water leaves out a large number of other, different sorts of things. Similarly, there was a philosopher named Austin who had some funny passages about those who talked only about "medium-sized dry goods," and left out things like rainbows.

    ReplyDelete
  30. Mikael,

    Even if the only root cause of changes is the change in motion, that would not imply that change in motion is the only sort of change. Other changes (from peaceful to angry, from ignorant to knowing, from opining that Trump is an idiot to opining that Trump is a moron), these may be caused by atomic motions but are not, themselves, the very same things as those atomic motions.

    Simon Kissane: certain conclusions do depend on the A-T sense of the act-potency doctrine, and the substance-accident doctrine. But that does not mean that Prof Feser merely posits these doctrines as if they were absolute starting points. He argues to them. His arguments (and others' for the same doctrines) oppose and - at least so they think - defeat other positions, such as those of reductionist materialism, and pure idealism. The Prof had a 10-part series of posts against reductionist materialist, for example. So, you seem to be asking to have all those arguments made over again here. But if Ed and we were to do that, it would leave no room for THIS argument on THIS topic.

    In much the largest range of the global possibilities, there would exist an arbitrary set of messily complicated worlds.

    Parfit certainly abuses the word "possibilities" here, and probably does the same for "global" and "world".

    There is no conceivable alternative. Since it is necessary that one possibility obtains, it is necessary that it be settled which obtains. Even without any kind of process, logic ensures that a selection is made. There is no need for hidden machinery.

    Parfit is also making a category mistake. Or two. His notion of "necessary" is pretty badly off, and he makes the mistake of using "logic" as if it eliminated the need for either a cause or an explanation.

    ReplyDelete
  31. @Tony: I'm familiar with some of Prof Feser's arguments against materialism in favour of A-T. Actually, I'm an idealist, and I haven't been able to find much in the way of arguments by him against idealism (not saying he hasn't said anything on this topic, just personally I haven't been able to find much). A-T positions make about as much sense to an idealist like myself as they do a materialist, but anti-materialist arguments aren't always directly relevant to me (since I agree that materialism is false, but the falsehood of materialism doesn't by itself establish the truth of any particular alternative to it). So, if you have any pointers to work (by him or others from his school of thought) attacking idealism specifically, I'd be interested to read that.

    ReplyDelete
  32. It's ok Tony,I'll search for those posts.I know it can be very annoying to repeat the same things whenever someone that has no clue about something arrives here.I'll search for them.

    ReplyDelete
  33. I promise this is my last question:).What do you mean when you say that God is perfect?I mean...look...I understand that God must be superior to all things that He creates,cause He won't have the limitations of the created things,but what does it mean to be perfect?Can we say that a Perfect Being can do whatever is metaphysically possible?And how do we know what is metaphysically possible?

    ReplyDelete
  34. Vincent Torley said, 'Although the world is not a brute fact, God's choice to make it is "brute" in the sense of being inexplicable, although it of course presupposes the fact of God's existence.'

    God's choice is not brute because he created out of his love, out of his desire to share his eternal love and existence -- This is because he made conscious beings (at least us) and other things to support those beings, blah blah blah. Of course this all assumes you agree with the classical conclusion that God is good.

    ReplyDelete
  35. What do you mean when you say that God is perfect?

    To my understanding, the easiest thing to say is that there cannot be any lack in God, cause when we speak of a lack in an essence, there is something in the essence that is not actualized. But something that is Pure Act cannot have anything that is not actualized.

    Speaking of possibility, I guess A-T starts off by investigating the essence, and finds that what is true for the essence, is true of it in every possible world.

    ReplyDelete
  36. @Simon

    I seem to recall Feser critiquing idealism in both Last Superstition and in the Philosophy of Mind, e.g. in the discussion of Berkeley and his successors.

    Maybe someone else can elaborate on that.

    ReplyDelete
  37. Many (even most) atheists nowadays adopt a reductionist materialist ontology in which there is no substantial distinction between a person and the mereological sum of the atoms which constitute their body at any given moment

    That only proves sadly that "many (even most)" atheists are fools. A person cannot be the sum of the atoms constituting his body at any given moment simply because those atoms are constantly changing, while the person remains. Your body today does not consist of the same atoms as it did last year, yet you remain you.
    +++++
    What do you mean when you say that God is perfect?

    By analogy with anything else said to be perfect: "thoroughly made". That is, lacking in nothing that is of its nature. Think what you would mean by a "perfect circle" or a "perfect right triangle" or some thing of the sort.
    ++++++

    ReplyDelete
  38. The important question is where are you getting those comic book panels?

    This is clearly the important line of inquiry. If I were to irresponsibly speculate I would have to go with some sort of supercomputer that can tap into to The Vestibule where The Fulcrum serves.

    ReplyDelete
  39. Step2 writes:

    This is clearly the important line of inquiry.

    I agree. Those who like to accuse Step2 of trolling, please take note.

    ReplyDelete
  40. @Simon Kissane: "Many (even most) atheists nowadays adopt a reductionist materialist ontology in which there is no substantial distinction between a person and the mereological sum of the atoms which constitute their body at any given moment..."

    This surprises me to hear, because in my own experience among atheists, which is not negligible, they *adopt an ontology* at best in the sense that Monsieur Jourdain spoke prose.

    That they might deny a *substantial* such distinction would of course be significant in why they might deny a real such distinction. Perhaps that is what you were saying.

    "A-T positions make about as much sense to an idealist like myself as they do a materialist..."

    This is a puzzling, for of course a position oughtn't fail to make sense merely because one disagrees with it. Speaking for myself and no other, it was while being a materialist that I found idealist and A-T positions made *more* sense than what I then thought.

    "So, if you have any pointers to work (by him or others from his school of thought) attacking idealism specifically, I'd be interested to read that."

    I myself have no notable A-T works attacking idealism specifically. Scott may have. Daniel Joachim is right that Dr. Feser touches on idealism in TLS and in PoM, but in both cases only very tangentially. If I may ask, what sort of idealist are you?

    (I suppose you must know of G. E. Moore's "Refutation of Idealism.")

    ReplyDelete
  41. "A-T positions make about as much sense to an idealist like myself as they do a materialist..."

    As Scott pointed out elsewhere, the difference between Plato's idealism and Aristotle's realism is not materialism (or akin to it).

    "It is more correct, therefore, to call his [Aristotle's] teaching an immanent idealism as contrasted with the transcendental idealism of Plato."

    ReplyDelete
  42. A-T positions make about as much sense to an idealist like myself as they do a materialist,

    Speaking from my own relatively modest experience, by and large materialists don't so much consider the A-T position and then reject it, as they reject it without having studied its arguments or understood them. That covers about 97% of modern scientists so far as it seems to me (and only to the extent they adhere to ANY position, which does not extend to everyday life, in which their actual behavior does not adhere to materialism). They simply accept without demur the materialism fed to them by others. Materialist philosophers may have studied A-T's realist positions to some extent, but judging from how often their arguments get the concepts wrong they usually don't actually understand them well enough to say honestly "we understood that and reject it on the basis of reasoning." Now, I do not claim this applies to ALL materialist philosophers, there are probably a few out there who really do get the arguments of A-T realism, but still reject it. I would guess that, for those, they do not DISDAIN the A-T position the way Coyne and Dawkins do.

    ReplyDelete
  43. I'm an Eastern Orthodox Christian and I must recognise that I'm very sympathetic with AT-metaphysics.But I've read some things on this blog(older posts)that looked really weird and for me they seem contradictory.According to AT,it is metaphysically impossible for God to perform evil acts.Thomas Aquinas says that God does what seems to be evil acts(for example allowing people to have cancer) only in order to obtain something better(for example sufference makes people more humble and they may return to worshipping God).Fine.According to both Roman-Catholic Church and Eastern Orthodox Church after people die they become an incomplete substance but this won't last forever cause God will ressurect us.Fine again.But according to Aquinas when animals die,they die forever and their essence perishes forever.So I have 2 questions:1.If God cannot do evil things,why according to Catholic Church does He not ressurect animals?Or plants?(Or am I mistaken?I've read the dispute between Hart and Feser and Feser said that animals won't be resurrected.).2.Isn't the act of completely destroying the essence of a thing an evil act(according to Catholic Church)?And if not,why are humas priviledged?What if God decides to annihilate all creation?(I don't want to be understood wrongly,I'm not trying to create a polemic here,but I want to see thomists' approach.I believe that all creation will be restored).

    ReplyDelete
  44. I believe that all creation will be restored

    So do people who take a Thomistic position on this point. But the restoration will not be a copying, because that would just involve all the imperfections and failings and mortality again. Further, resurrection means that this very thing is resurrected, not a copy or something like it, and that means that there has to be something that things continue to have until the resurrection. So it will be the incorruptible, or what can be made incorruptible, that will be restored; but then the only question is just whether something is corruptible or not. We see this with your first question:

    1.If God cannot do evil things,why according to Catholic Church does He not ressurect animals?Or plants?

    The question that would need to be answered before this is why it would be an evil act not to resurrect animals or plants. The underlying core of the argument to which you are referring is that animals and plants are naturally corruptible; so it would just be natural for them to live their finite space and that's it. With human beings, we are partly corruptible, partly incorruptible (as far as intellect goes); so the question is what will happen to the incorruptible part. This question doesn't even arise if there is no incorruptible part: things just end, and that's all.

    On the other hand, if we had good reason to think there is such an incorruptible part in animals and plants, they would be handled in exactly the same way human beings are handled. So it is in fact just a question of what human beings share or don't share with animals and plants, and not a question about evil or God at all. That can be a perfectly legitimate position -- but only if it is based on an understanding of what animals and plants are in themselves, not on trying to get the result that we happen to fancy at present.

    ReplyDelete
  45. Professor Feser writes,

    "The point is rather that the claim that ultimate explanations are not causal explanations..."

    I could see this use of causal explanation causing a lot of confusion. I understand you are using it in the sense that an infinite regress of causes in certain explanatory series is impossible; however, I think generally people do not see, say, positing God as Creator as positing a non-causal explanation of/for existence (though of course He is not Himself thought to be caused by anything or anyone). God Himself is not, however, in turn Himself a Brute Fact.

    It just sounds highly unintuitive to me that the First Cause or the Uncaused Cause it not positing a 'causal explanation of the universe'; to be sure, however, this is definitely not positing a causal explanation for existence as such (existence couldn't in principle have caused itself to exist). Now I only say this for practical reasons in that I expect you might have a lot of rather shocked or surprised expressions if you were to say that, e.g., classical theists don't argue for or posit a causal explanation for the universe: Brute Fact View appears to me rather to be the view that there is no causal explanation of/for the universe.

    That being said the article is excellent and gives a lot to think about.

    ReplyDelete
  46. @ TOF,

    "A person cannot be the sum of the atoms constituting his body at any given moment simply because those atoms are constantly changing, while the person remains."

    Heh. Interesting that this problem popped up here after I made my comment about the distinction between substance and accident! It may have otherwise seemed like a totally irrelevant digression or even thread hijacking.

    I'm always amazed and impressed how especially metaphysical reasoning always seems to lead back into itself.

    But furthermore, you are correct that atomism is problematic for an account of the human person, especially if you were to pretend to make it relevant for where that account is arguably most practically felt (I'd say that's in the law). But it's not only that. The atomist still seems to me to end up entertaining some sort of spin-off of trope theory, but a version of it that to him sounds much more reasonable and scientific, in that the tropes are real, individual physical entities. It's just that he tries to make either the atoms or their components explanatorily basic elements to account for 'the what' (the ultimate subject) that 'things are said of'. But still, I think, he is missing the thrust of the problem and the reason why someone who entertains tope theory doesn't appeal, as it were, to atoms; because even atoms have totally accidental features; e.g., atom X's being in this or that location or being part of (or not or no longer part of) this or that compound or molecule, etc. Of course, a virtually infinite number of even more obviously accidental relations can also be posited to any given atom (it is so-far-from this or that other atom or this or that city-centre or whatever).

    ReplyDelete
  47. "animals won't be resurrected"

    Dear Mikael: It's not so much that they WON'T be; it is that given their nature, they don't seem to require a resurrection. This is not to say that God's grace could not do such a thing--His grace perfects and builds upon nature but does not destroy it, to coin a phrase. :)

    ReplyDelete
  48. Off topic: As we enter this Lenten season, may God bless you all.

    ReplyDelete
  49. Off topic: As we enter this Lenten season, may God bless you all.

    And you likewise, Scott. Thanks for all the contributions you provide here; it is a pleasure for us all. And nice picture, by the way.


    ReplyDelete
  50. Thank you so much for your responses.In what concerns me,I think that animals too have some sort of immaterial power.They can act freely(of course not like us,they are more inclined to follow their instincts than us but certainly I don't think they are robots) for instance.But when dealing with skeptics that are inclined to think after some discussions that the God of classical theism exists,can we convince them on PURELY philosophical(and not religious grounds) that if God of classical theism exists,we are immortal and that God cannot change His mind and reverse the immortality of the soul that He created.On purely philosophical grounds,with no appeal to revelation.

    ReplyDelete
  51. Mikael,

    I don't know of anyone who argues that the soul is immortal on the basis of classical theism; whether or not the soul is immortal is a question of human nature.

    ReplyDelete
  52. Mikael, I am not sure you are clear what you are implying. God can do things that are metaphysically possible, and He cannot do things that are metaphysically impossible. God cannot change the number 2 into the number 3 for a time, for example. Nor can He decide to hate the good, or love evil as such. The "immortal soul" is immortal by reason of the very nature it has. God can (theoretically) refuse to keep such a thing in existence (because all created things depend for their continued existence on His continued willing that they exist), but the immortal soul cannot be made to "pass away" the way material beings pass away, for the same reason 2 cannot be made into 3: it's nature doesn't admit of that sort of change. That's a metaphysical impossibility If you were to then ask "but what if God were to 'make it's nature' into something else", that too would require a metaphysical impossibility, that of changing a thing's nature while leaving the thing's identity intact. (It would thus mean that the thing being made to pass away is not an immortal soul, but some OTHER sort of thing. It would mean that it is not really "2" that is being made "3", but something else.)

    .But according to Aquinas when animals die,they die forever and their essence perishes forever.So I have 2 questions:1.If God cannot do evil things,why according to Catholic Church does He not ressurect animals?Or plants?

    To add to what Brandon has already said, there is a way in which the death of a plant or animal is an "evil in a sense" but is "not an evil in the ultimate sense". God arranged a universe of bodily beings, in which the coming to be of some things implies the passing away of prior things: the coming to be of a lion implies the passing away of all the herbivores of which its flesh is made in its mother's womb. Even more elementally, (under the A-T doctrine of substantial being) the coming to be of a tree implies the cessation of the independent substantial being of the independent carbon atoms, carbon dioxide molecules, etc, that make up its stuff. The order implicit in such a universe implies that all of these changes are ordered to a good that encompasses the whole: on the level of the universe, the passing away of an antelope to feed the lion is a GOOD thing. Since the good of the higher order (the whole ordered economy of creation) is more to be willed than the good of a lower order, to speak of the death of the antelope as "evil" it must be captured within constraints: it is a qualified sort of evil.

    The punishment of evildoers in hell must be viewed the same way: they suffer an evil, that's what punishment IS, what it consists in. But that they suffer evil is a good of the economy of creation, since "the order of justice is the order of the universe" as St. Thomas says. It is would be ultimately a disorder in the universe if evildoers were not punished.

    ReplyDelete
  53. I can't believe you believe this religious mumbo-jumbo. The fact that stuff like evildoers getting punished for eternity is being discussed on a philosophy blog is hilarious. Next thing you'll be talking about angels.

    ReplyDelete
  54. @Anonymous,

    if evil is the privation of good and God is All good would'nt it be logical for evil doers to stay for eternity away from God. I think it would be "torture" if God force them to spend eternity with Him.

    ReplyDelete
  55. It's remarkable how often the phrase is hilarious is used in this sort of context to mean "seems superficially silly to someone who hasn't troubled to understand it." And we do talk about angels sometimes, so yeah, if they come up, it won't surprise me either.

    ReplyDelete
  56. Anonymous said...

    @Anonymous,

    if evil is the privation of good and God is All good would'nt it be logical for evil doers to stay for eternity away from God. I think it would be "torture" if God force them to spend eternity with Him.

    February 11, 2016 at 6:56 AM"



    Evil-doers are assumed to have a positive nature which has been corrupted in various ways, most especially through their choice of using privation as a tool directed against others, in order to satisfy what would be, if properly ordered, positive impulses.

    Thus, the answer to your question would depend on whether the deprivations the evil-doers commit are committed for the deprivation per se; or, in order to positively satisfy - no matter how perverted the method chosen is - what was before the warping occurred a legitimate organic urge.

    No doubt if the evil-doer got literally nothing out of it, they would not do it. Since few persist in doing something that results in nothing at all in return, be it no more than a pleasing chemically induced physical shiver at the thought of spitting in God's face.

    You don't even have to believe in God in order to appreciate how this conceptual distinction plays out psychologically (since that is what you are talking about) in the case of people who imagine that they have died and gone to hell only to experience nothing around them but the rage and malice of their own kind, and have then somehow been reprieved.

    It may all be in their minds of course, but it illustrates what the concepts look like as experienced by the mind.

    If Hell were real, you probably would not enjoy it.

    ReplyDelete
  57. The fact that stuff like evildoers getting punished for eternity is being discussed on a philosophy blog is hilarious.

    Not very informed about the history of philosophy, are we, Callicles?

    ReplyDelete
  58. The fact that stuff like evildoers getting punished for eternity is being discussed on a philosophy blog is hilarious.

    What would be hilarious would be you claiming that your right hand, after you have cut it off, would grow back.

    ReplyDelete
  59. @Anonymous February 11, 2016 at 4:38 AM: "I can't believe you believe this religious mumbo-jumbo. The fact that stuff like evildoers getting punished for eternity is being discussed on a philosophy blog is hilarious. Next thing you'll be talking about angels."

    Fools rush in where angels fear to tread. And rush out too, apparently, without making any arguments.

    @Scott: "It's remarkable how often the phrase is hilarious is used in this sort of context to mean 'seems superficially silly to someone who hasn't troubled to understand it.' And we do talk about angels sometimes, so yeah, if they come up, it won't surprise me either."

    Surprise! And therefore, in a spirit of épater le bourgeois, a few more choice words:

    "If an angel were ever to tell us anything of his philosophy I believe many propositions would sound like 2 times 2 equals 13."--Lichtenberg, *Aphorisms* Notebook B, B44

    "O white-robed Angel, guide my timorous hand to write as on a lofty rock with iron pen 1 the words of truth, that all who pass may read."--Blake, *Poetical Sketches,* "Samson"

    ReplyDelete
  60. [2nd] Anonymous said...

    @Anonymous,

    if evil is the privation of good and God is All good would'nt it be logical for evil doers to stay for eternity away from God. I think it would be "torture" if God force them to spend eternity with Him.

    February 11, 2016 at 6:56 AM"


    I should have with regard to Hell, said "one" (presumptively an "evil-doer") would probably not enjoy it.

    Your own, 2nd anonymous, point, i.e., that one who embraced evil or deprivation and depredation of fellows as a way of existence, would find it logically in-congruent to dwell in a domain which represented the antithesis of his own values, does stand to some reason.

    ReplyDelete
  61. @DNW,
    "Your own, 2nd anonymous, point, i.e., that one who embraced evil or deprivation and depredation of fellows as a way of existence, would find it logically in-congruent to dwell in a domain which represented the antithesis of his own values, does stand to some reason."

    Yeah that is what I meant and that God in His loving nature would not force His Will on anyone, even though He would not want His children to stay away from Him for all eternity.

    ReplyDelete
  62. atheist anon: "I can't believe you believe..."
    A stunning 'argument'.

    ReplyDelete
  63. Timocrates, you write, "It just sounds highly unintuitive to me that the First Cause or the Uncaused Cause it not positing a 'causal explanation of the universe'; to be sure, however, this is definitely not positing a causal explanation for existence as such (existence couldn't in principle have caused itself to exist). Now I only say this for practical reasons in that I expect you might have a lot of rather shocked or surprised expressions if you were to say that, e.g., classical theists don't argue for or posit a causal explanation for the universe: Brute Fact View appears to me rather to be the view that there is no causal explanation of/for the universe."

    Who's doing the positing, the First Cause or someone else who believes that there is or may be a First Cause?

    ReplyDelete
  64. In my example somebody else is, William. Perhaps I should have written, "It just sounds highly unintuitive to me that positing a First or Uncaused Cause is not positing a 'causal explanation'..."

    ReplyDelete
  65. Or William, are you asking more in the line of the "God did it" response? That is, are you asking that -granting a First Cause is necessary- does it not follow that ultimately someone's positing, say, an Uncaused Cause is necessarily itself an effect of that same Cause and, so, in a sense, would just really be the Uncaused Cause positing something (and not some individual creature - at least not really)?

    ReplyDelete
  66. Surely hilarious Anonymous was being satirical? His comment struck me as too deadpannish not to be.

    ReplyDelete
  67. I suspect the anon with the "hilarious" comment is the same uber-troll who slithers his way around com-boxes and other sites making it clear his understanding of metaphysics is...um...limited. He did, however, reference Quine once while trying to disrupt the reviews of DBH The Experience of God over at Amazon. This, of course does not mean he read Quine to any great extent, let alone understood how that would negate arguments made by DBH or others.

    ReplyDelete
  68. Coming back to the central question as to whether the cosmos can be taken at face value as simply a "brute fact," it reminded me that I wrote an article long ago entitled, "How Creation Implies God." This article addresses directly the "brute fact" claim and may be of interest to some. It originally appeared in Faith & Reason.

    Here is the link to it on my own website: http://drbonnette.com/articles/19/evolution-micro-macro-articles-home-creation-implies/

    ReplyDelete
  69. To say reality is selected, randomly or not, is to assume a meta-reality or environment within which reality has its being. God, in other words.

    ReplyDelete
  70. I’ve been tracing some of the links back to earlier blog posts and comment threads and I’m confused about something (well, many things most likely). I don’t understand the difference between the Principle of Sufficient Reason and Determinism. The earlier posts make it clear that scholastics (or at least Prof Feser) accept the former but not the later.

    If a contingent event occurs that involves something coming into existence or of having some attribute, then by PSR there must be a sufficient reason or necessary objective explanation for that event. I take it that by “sufficient reason” one means that given the existence of the reason the event must needs have occurred. But this is just determinism isn’t it? If an explanation Y for event X is such that Y could obtain without X occurring, then how could Y be said to be a sufficient reason for X? Alternately if X could occur without Y then Y isn’t the reason for X.

    ReplyDelete
  71. mackthemike:

    "If a contingent event occurs that involves something coming into existence or of having some attribute, then by PSR there must be a sufficient reason or necessary objective explanation for that event."

    Part of the problem is in the bolded bit: the PSR does require that the explanation be sufficient, but not that it be necessary. For all the PSR has to say, the event could have occurred in more than one way.

    "I take it that by 'sufficient reason' one means that given the existence of the reason the event must needs have occurred."

    And the other part is a confusion about the nature of the supposed necessity. A sufficient explanation needn't involve causal necessitation. If e.g. my wife has to choose between a Coke and a Pepsi and she chooses the Coke, there's still a clear sense in which she could have chosen the Pepsi; she just didn't, and never would, because she likes Coke and doesn't like Pepsi. That latter fact explains her choice, but in doing so it doesn't imply that the result was causally necessitated, i.e., "determined" in the sense required by determinism.

    ReplyDelete
  72. (Clarification/elaboration: That last paragraph also applies to a "necessary explanation," which I take to be the only possible explanation for a given event. That's why I say there's a confusion about the nature of the supposed necessity.)

    ReplyDelete
  73. Thanks Scott.

    The text you bolded above came originally from Prof Feser’s 10/10/2014 post, “Della Rocca on PSR”, which says in relevant part “The principle of sufficient reason (PSR), in a typical Neo-Scholastic formulation, states that there is a sufficient reason or adequate necessary objective explanation for the being of whatever is and for all attributes of any being” [internal quotation marks omitted]. I don’t think the part about an adequate necessary objective explanation is needed for my question, so I propose to drop it for the purposes of my question.

    The heart of my question is what “sufficient” means in the PSR. The most obvious and straightforward reading is that the sufficient reason must be sufficient to bridge the gap between the explanation and what is being explained. That is, one must be able to move reliably from the truth of the facts in the proposed explanation to the truth of that which is being explained. To deny this would be to assert that the explanandum might still not obtain, the truth of the explanans notwithstanding.

    I take your point that as a general matter the explanation need not be a causal one, but when the PSR is applied to spatial-temporal events surely that is causal. Isn’t this just what determinism says?

    ReplyDelete
  74. Also Scott, I think fact that your wife could have ordered Pepsi, but never would have done so is compatible with Determinism if one has the correct understanding of the uses of the terms “could” and “would.” The statement can be re-cast as “My wife could have chosen Pepsi if she had wanted to, but never would have given her dislike of it.”

    ReplyDelete
  75. Mack The Mike:

    The heart of the matter is that causation needn't, in my view at least, involve necessitation. So I would say no, determinism doesn't just say that spatio-temporal events are "caused"; it says something about the mode or manner in which they're caused. And although I used to agree with Brand Blanshard that causes necessarily bring about their effects, I now agree with Stephen Mumford that to explain causation by necessity is to "explain" the clearer by the more obscure.

    Moreover, it's one thing to say that on certain understandings of "could" and "would," the fact that my wife could but would not order Pepsi is merely compatible with determinism; even in that case, that fact wouldn't commit us to determinism. But it's another matter to show both that those understandings are the correct ones and that they entail determinism rather than merely fail to be inconsistent with it.

    As for the correctness of those understandings, I don't know which ones you have in mind and so can't comment directly. But I'll reiterate that I see nothing unclear in the idea that my wife could have ordered a Pepsi, that it was therefore (causally) possible that she not order a Coke, that her ordering the Coke therefore didn't occur by (causal) necessity, and that it therefore did not occur "deterministically" in the sense required for determinism.

    On the contrary, the suggestion that it did occur "necessarily" seems to me to make a pretty hash of any suggestion that she "could have done otherwise" in just about any relevant sense. At the very least I'd want to know what sort of "could-ness" was involved in any claim that she "could" have ordered Pepsi if she was causally necessitated to order Coke.

    ReplyDelete
  76. Ah, OK. I see I’ve made an error. One thing I noted when reading comments on the “Causality and radioactive decay” post (12/12/2014) was that the commenters were sometimes talking past each other because they were using the word “cause” to refer to different things. That’s why I avoided using the term at all in my initial comment above. But the term snuck back in under the cover of ‘Determinism’. I should not have let it.

    The problem is that the concept of causality being used in descriptions of Determinism is the Humean one of ‘constantly-conjoined’ not the Aristotelian one. Determinism does not entail the belief that events are necessarily caused by their antecedent states of affairs in the Aristotelian sense of causality --merely that any relevantly similar state of affairs will always be conjoined with a relevantly similar subsequent state of affairs.

    So the question of whether or not causation entails necessitation isn’t of direct interest. The question is whether sufficiency entails necessitation. I don’t see how a putative explanation for an event can be sufficient if the explanans don’t necessitate the explanandum.

    ReplyDelete
  77. I want to be clear that I’m not making any kind of argument for Determinism, but rather asking the question of whether Determinism and PSR must stand or fall together. You may be right that the fact that your wife could have chosen Pepsi notwithstanding the fact that she actually chose Coke disproves Determinism. But if so, then there is not a sufficient reason for her having chosen Coke is there?

    ReplyDelete
  78. I think there are three basic ways to view potentiality. Let’s call them nominalism, extreme realism, and moderate realism after the three positions on the status of universals.

    The nominalist view is that a potential is merely a hypothetical and can only be understood in terms of the mental models formed by observers with limited information. According to this account, a being with perfect knowledge wouldn’t think in terms of potentials at all, for such a being would already know all the actuals.

    An example will help. In a nine player game Texas Hold’em two players are heads up and all-in after the flop. Player one has Q-Q and player two has 9-8 off-suit. The flop came Q-10-7 rainbow. The players’ hands are both face up since there’s no more betting to be done. Player two says “Well, I could still win if a Jack comes”, since that would give him a Queen-high straight. By using the word ‘could’, player two is employing a hypothetical. He is saying that there is a potential configuration of cards in the deck which would result in his winning the hand. Given what he knows, player two is correct. But as it turns out, all four Jacks are already in the muck. If this is a televised game, the announcers already know this, since they have seen all the other players’ hole cards. So from the announcers’ point of view player two is drawing dead. There is no potential configuration of cards under which player two wins.

    Under the nominalist understanding of potentiality all uses of the words ‘could’, ‘can’, ‘potentially’, etc. are like this. They are all to be understood solely in terms of a particular point of view – a set of knowns and unknowns. Under this interpretation a Determinist may well agree that your wife “could” have chosen Pepsi notwithstanding the fact that she actually chose Coke. All the Determinist would mean here is that she could have chosen Pepsi given our limited knowledge of her state of mind beforehand.

    ReplyDelete
  79. The extreme realist position on potentiality is that all potential states of affairs are actual states of affairs in an alternate reality. These alternate realities are real, not hypothetical. The Many Worlds interpretation of Quantum Mechanics is an example of this kind of extreme realism on the status of potentiality.

    Under this view your wife necessarily chose Coke insofar as she necessarily made every possible choice. She chose Coke, she chose Pepsi, she chose rum, goats blood, etc., etc. She just did those things in different universes, or different branches of the Multi-verse. This interpretation of potentiality is also compatible with Determinism.

    ReplyDelete
  80. The moderate realist position on the status of potentials is that potentials are in some sense real, but not in their own actual universes. This version may not be compatible with Determinism or PSR. This is what I’m trying to work out.

    ReplyDelete
  81. uhg. a 6 would also give player two a straight. oh well. Say the sixes are *also* in the muck.

    ReplyDelete
  82. Hi Mack,

    I wrote this before your latest posts. Hopefully it's still close to the mark:

    I want to be clear that I’m not making any kind of argument for Determinism, but rather asking the question of whether Determinism and PSR must stand or fall together. You may be right that the fact that your wife could have chosen Pepsi notwithstanding the fact that she actually chose Coke disproves Determinism. But if so, then there is not a sufficient reason for her having chosen Coke is there?

    It may help to draw a distinction between necessitarianism and determinism. If determinism is correct, every cause necessitates its effect but the actual world isn't the only possible world[1]. In contrast, if necessitarianism is correct, every cause necessitates its effect and the actual world is the only possible world. I suspect that what's really at issue here is whether the PSR entails necessitarianism.

    Call true propositions facts. Call the conjunction of all contingently true propositions the Big Conjunctive Contingent Fact (BCCF).

    Suppose the PSR is true.

    Necessarily, every contingent fact has an explanation (PSR). There is a contingent fact that includes all other contingent facts (BCCF). Hence, there is an explanation of this fact. If the BCCF explains itself, there are contingent self-explainers[2]. There are no contingent self-explainers. Hence, the BCCF does not explain itself.

    Since the BCCF includes all contingent facts, cannot explain itself, and has an explanation, the BCCF is explained by a necessary (non-contingent) fact. So far, this isn't that different from Pruss's PSR cosmological argument.

    Let p be some explanandum and q, an explanans. If p is the BCCF, q is a necessary fact. If q explains p, q entails p. Hence, by modal modus ponens, the BCCF is a necessary fact.

    But the BCCF is by definition a contingent fact. So, the BCCF is both a necessary fact and a contingent (non-necessary) fact. So, by reductio, the PSR is false.

    It sounds like you're asking if the only way we can avoid the above (or something like it) is by arguing that there are only necessary facts[3], thereby rejecting the second premise. Am I going in the right direction?


    [1]No ontological commitment to such worlds is intended.
    [2]There are also Hume-Edwards-Campbell type responses to this premise. To save space and for the sake of illustration, I'm going to assume they're incorrect unless someone mentions them.
    [3]If there are only necessary facts, there is only one possible world and necessitarianism is correct.

    ReplyDelete
  83. John, Yes. Exactly. Well put. That's what I'm thinking. Is there a hole in it?

    ReplyDelete
  84. Mack the Mike:

    I'm afraid I'm a bit under the weather after a busy afternoon, so I'll have to continue this at a later time. However, I can address this part briefly:

    I want to be clear that I’m not making any kind of argument for Determinism, but rather asking the question of whether Determinism and PSR must stand or fall together.

    Yes. And I'm saying they're not, because there's at least one account of causation (nondeterministic or non-necessitarian) that preserves the PSR and doesn't entail determinism (or necessitarianism).

    You may be right that the fact that your wife could have chosen Pepsi notwithstanding the fact that she actually chose Coke disproves Determinism.

    That's not quite my argument but I think that's true as well. At any rate, that argument also suffices to show that the PSR doesn't entail determinism/necessitarianism.

    But if so, then there is not a sufficient reason for her having chosen Coke is there?

    Sure there is: she likes Coke and she doesn't like Pepsi. For that reason, she'll pick Coke over Pepsi every time, and would do so even if we could somehow "roll back the tape" and "replay" the same choice over and over. How could that explanation fail to be sufficient? What else would you need to know to explain her choice?

    ReplyDelete
  85. "They're not" should be "they needn't."

    ReplyDelete
  86. Hi Mack,

    Consider the sixth premise:

    If q explains p, q entails p.

    Either the explanans needs to have its explanandum in it, or it doesn't. Suppose the explanans needs to have its explanandum in it. Then the premise is true. q and p entail p.

    Suppose the explanans doesn't need to have its explanandum in it. The proposition “Fred died because he was shot”, where “Fred died” is the explanandum and “[Fred] was shot” is the explanans, seems to be a counterexample. “Fred was shot” doesn't entail “Fred died”. (People survive getting shot all the time.)

    Nor is it a matter of detail. Suppose the explanans is “Fred was shot and the bullet hit Fred in the head and [...]” right down to the minute details of the resulting injuries. It's still possible that Fred was hit by a truck at the same time, and that the injuries from the truck killed him first. So, even with such a detailed explanans, the explanans still doesn't entail “Fred died”.

    Maybe if we had to include every true fact about the world in the explanans, then it would be a matter of detail. But if that's what's needed for a sufficient explanation, we never sufficiently explain anything.

    Since the BCCF includes all contingent facts, cannot explain itself, and has an explanation, the BCCF is explained by a necessary (non-contingent) fact.

    Another, more radical reply is to argue that the explanans in this case isn't a proposition. It's a necessary being (or, if you want to avoid calling God a being for pedagogical reasons, Being Itself). Moreover, it's a necessary being able to make reasoned free choices[1].


    [1]Vincent may take issue with the “reasoned free choices” part of this reply. Since I don't want to give short shrift to his argument, I won't reply to it here.

    ReplyDelete
  87. I made a mistake in this paragraph:

    Nor is it a matter of detail. Suppose the explanans is “Fred was shot and the bullet hit Fred in the head and [...]” right down to the minute details of the resulting injuries. It's still possible that Fred was hit by a truck at the same time, and that the injuries from the truck killed him first. So, even with such a detailed explanans, the explanans still doesn't entail “Fred died”.

    All that's required for the explanans conjunction to entail Fred's death is that in every possible world in which the conjunction is true, Fred dies. It doesn't follow from the fact that Fred died for some other reason first that Fred doesn't still die in the world in which the explanans conjunction is true. So, that part doesn't work, though it's still not clear that “Fred died” follows from the detailed explanans.

    (What I was thinking was that the detailed explanans doesn't entail that “Fred died because he was shot”.)

    ReplyDelete
  88. Thanks for the thought provoking response John.

    Russell in his chapter on determinism in "Religion and Science" defines determinism as the hypothesis that:

    There are discoverable causal laws such that, given sufficient (but not superhuman) powers of calculation, a man who knows all that is happening with a certain sphere at a certain time can predict all that will happen at the centre of the sphere during the time that it takes light to travel from the circumference of the sphere to the centre.

    My only quibble with Russell here is that for the purposes of metaphysical speculation the laws need not be, as a practical matter, discoverable nor the calculations humanly possible -- merely that they be well defined.

    Such a fine level of detail would include the existence and momentum vector of the truck that kills Fred in your example. Now of course a less complete understanding of the circumstances than that described by Russell might still suffice to necessitate Fred's death. But if not even a Russellian level of local knowledge suffices to answer the question of whether Fred lives or dies with certainty, in what other sense can that knowledge be said to be sufficient?

    ReplyDelete
  89. Scott, you write:

    Sure there is: she likes Coke and she doesn't like Pepsi. For that reason, she'll pick Coke over Pepsi every time, and would do so even if we could somehow "roll back the tape" and "replay" the same choice over and over. How could that explanation fail to be sufficient? What else would you need to know to explain her choice?

    If every "replay" always results in the choice of Coke, then the antecedent state of affairs and the subsequent choice of Coke are, as it were, "constantly conjoined" and completely predictable. That means that the conditions of Determinism are satisfied.

    ReplyDelete
  90. Predictability isn't causal necessitation; the former is an epistemological notion, the latter is metaphysical. And we'd be begging the question if we simply assumed that the first entailed the second. The entire point of my example is that it seems to be a case in which we have complete predictability but not necessitation (and therefore not "determinism").

    What would "necessitation" add to the (so it seems to me) entirely intelligible explanation that she consistently chooses one over the other because she likes it better? And in what sense would she be "choosing" if her "choice" were necessitated by antecedent factors and there was no sense in which she "could" have chosen otherwise?

    ReplyDelete
  91. Scott,

    As I mentioned in an earlier comment, I think often scholastics and non-scholastics talk past each other because their respective uses of term surrounding causality are so different. I think it is telling that Russell's definition of determinism doesn't employ the concepts of causation or necessity at all. I do think many definitions of determinism do make use of these terms, but they do so using the modern, non-scholastic meanings.

    Most people today, at least those who haven't had any training in Aristotle, think of causation in a way that can be cashed out entirely in terms of spatio-temporal events and mental events.

    I agree that predictability is primarily a matter of epistemology, but it also has a secondary metaphysical meaning. We can learn about metaphysics through epistemology. This is Kant's method: to ask what must be true of metaphysics given his epistemology. To maintain the distinction between metaphysics and epistemology is is why I said I'd rather drop the parts of Russell's definition regarding laws being "discoverable" and the calculations humanly possible. The distinction between metaphysics and epistemology is also why I thought that John's point that if we had to include too much detail in an explanans we'd never explain anything is off-point. PSR asserts that the sufficient reason exists not that we can find it.

    In any case I use 'predictability' in a metaphysical sense as a relationship that obtains between temporally ordered states of affairs such that an intelligence could predict the later from the earlier given sufficient calculating power regardless of whether or not anyone can get such knowledge.

    I propose that the way to avoid talking past each other about causation is to just avoid the terminology. The question then becomes whether the PSR entails determinism as Russell defined it, as the predictability of events at the center of the sphere given knowledge of the whole sphere. This approach will avoid going off on too many tangents.

    My current understanding of your position, Scott, is that PSR does entail complete predictability but that that predictability doesn't entail causal necessitation as scholastics understand it. I'm fine with this.

    ReplyDelete
  92. Hello again, Mack:

    But if not even a Russellian level of local knowledge suffices to answer the question of whether Fred lives or dies with certainty, in what other sense can that knowledge be said to be sufficient?

    Good question. It's easiest to quote Pruss in The Principle of Sufficient Reason:

    Explanation is always a relation between two facts, that is, two true propositions. Thus, necessarily, if p explains q, then the explanans p and the explanandum q both hold. As the preceding examples show, the notion of explanation does not, however, require that the explanation be final or ultimate in the sense that no mystery remains. An ultimate explanation is one in which the explanans itself does not call out for further explanation because it is either self-explanatory or necessary or both.

    However, we will usually require explanations to be
    full. This notion requires some explication. An explanation is full provided that it does not allow a puzzling aspect of the explanandum to disappear: anything puzzling in the explanandum is either also found in the explanans or else [is] explained by the explanans. It would not do to explain why John is sad and excited by saying that he was made sad by the death of his dog, Fido. That would miss out on a part of the explanandum, namely, why he is also excited. One way to give a full, though not ultimate, explanation is to say that John is made sad by the death of his dog, Fido, and excited by a job offer he has received.

    ReplyDelete
  93. Well, all that's needed to disprove an entailment relation is that it's possible the truth of the conditional's antecedent—the explanans—doesn't imply its consequent—the explanandum[1]:

    Such a fine level of detail would include the existence and momentum vector of the truck that kills Fred in your example. Now of course a less complete understanding of the circumstances than that described by Russell might still suffice to necessitate Fred's death.

    But now you're off on a regress. Fred could have also been hit by a brick at the same time, and the injuries from the brick could have killed him. The regress terminates once every true fact about the world is in the explanans.

    The core point is that it's possible that “Fred died” and that “Fred was shot and [...]” without it being the case that “Fred died because he was shot and [...]”. So, “Fred was shot and [...]” doesn't entail that “Fred died because he was shot and [...]”.


    [1]To put it again in the slightly misleading, but illustrative, language of possible worlds: “All that's needed to disprove an entailment relation is that in at least one possible world the truth of the conditional's antecedent doesn't imply the truth of its consequent.”

    ReplyDelete
  94. Mack The Mike:

    "My current understanding of your position, Scott, is that PSR does entail complete predictability but that that predictability doesn't entail causal necessitation as scholastics understand it."

    Well, I wouldn't say PSR entails complete predictability; I don't think an event that was (say) only statistically predictable would automatically violate it. But the latter half, yep.

    ReplyDelete
  95. The question then becomes whether the PSR entails determinism as Russell defined it

    Entailment is a very strong relation, and if PSR entails something, that thing would have to follow from PSR on its own. But Russell's definition of determinism requires an account of what causal laws are, it requires us to posit that causal laws in that particular sense are real, and it requires an account of what it means to predict something. None of these follow from PSR on its own; PSR doesn't require that we have any notion of 'causal laws' at all (we would have to add something to PSR that specifically needs such things, rather than some alternative way of thinking about causation, to be explained), and it doesn't deal with prediction but with explanation, and accounts of one don't necessarily double as accounts of the other.

    In any case I use 'predictability' in a metaphysical sense as a relationship that obtains between temporally ordered states of affairs such that an intelligence could predict the later from the earlier given sufficient calculating power regardless of whether or not anyone can get such knowledge.

    How is 'predict' to be understood here? We know that actual predictions can be lucky -- e.g., we can predict rightly on the basis of sheer guess, like winning the lottery, or on the basis of patterns and streaks that were temporary due to unusual interference, or on the basis of false assumptions that nevertheless just happen to be close enough to truth in just the right way to fit. And even if we understand it in a sense requiring that it be more than luck, we can predict things with certainty knowing that, nonetheless, there are things that could make our prediction wrong. For instance, I can predict with certainty that I will not win the lottery because I did not buy a lottery ticket, while still recognizing that it's remotely possible that someone out of the blue has gifted me a winning lottery ticket that I haven't yet received. On the one hand, if this is ruled out, then John West's point seems something you are committed to: predicting correctly in the relevant sense requires knowing everything. Applying the PSR, on the other hand, quite clearly does not require knowing everything, just the explanandum and what you need as explanans for that specific explanandm, which means that the PSR does not entail this kind of predictability. On the other hand, if that kind of prediction is allowed (i.e., saying with certainty beforehand even if other things are also possible), then, given that extremely low probability events sometimes do occur, in those particular events it was impossible to predict that the extremely low probability event would have happened, because although it was recognized as possible it was extremely low probability. But PSR will function just as normally. Thus it seems in that direction PSR wouldn't entail predictability of the sort you have in mind, either.

    ReplyDelete
  96. I just noticed this (in your comment to Scott):

    The distinction between metaphysics and epistemology is also why I thought that John's point that if we had to include too much detail in an explanans we'd never explain anything is off-point. PSR asserts that the sufficient reason exists not that we can find it.

    For context, I wrote:

    Maybe if we had to include every true fact about the world in the explanans, then it would be a matter of detail. But if that's what's needed for a sufficient explanation, we never sufficiently explain anything.

    My point here was that it's an unreasonably high degree of detail to require for a sufficient explanation.

    ReplyDelete
  97. Let me offer my own definition of determinism using Russell as a jumping off point. Determinism is the doctrine that there exists a well defined function (f) such that given a complete description (d1) of the contents of a sphere at time (t1), the function evaluates to a complete description (d2) of the contents of a smaller sphere concentric with the original sphere at some later time (t2) provided that the radius of d2 is no larger than the radius of d1-c(t2-t1), where c is the speed of light. Or in other words d2=f(d1,t2-t1)

    So I've removed any consideration of the nature of causality. If by this definition determinism is false then either Relativity or PSR must be false as well, or so I'm claiming. Actually if we want to deny Relativity we can just make both sphere's infinitely large.

    ReplyDelete
  98. John,

    But now you're off on a regress. Fred could have also been hit by a brick at the same time, and the injuries from the brick could have killed him. The regress terminates once every true fact about the world is in the explanans.

    The regress terminates once every true fact about the world inside the sphere is in the explanans.

    ReplyDelete
  99. I'm no philosopher, but it also seems to me that if you really did need to include every true fact about the world in the explanans, then it begins to seem that almost amounts to determinism.

    jj

    ReplyDelete
  100. Thanks for the reply, Mack.

    If it helps, there's an analogy between my argument and Russell's problem for truthmakers (which he gives the unfortunate name “verifiers”) and general facts:

    The regress terminates once every true fact about the world inside the sphere is in the explanans.

    All we need to break an entailment relation is a possibility.

    Is there a possible world in which every fact in the explanans is true and yet something additional happened that instead explains the explanandum? The only way I see to rule the possibility out is for you to include, at least, every true negative proposition in your explanans, and there are a possibly infinite number of true negative propositions. I think that any explanans that includes a possibly infinite number of true negative propositions is far too detailed.

    As I wrote in my previous comment, if that's what's needed for a sufficient explanation, we never sufficiently explain anything.

    ReplyDelete
  101. Mack The Mike:

    "So I've removed any consideration of the nature of causality."

    …and indeed any consideration of causality at all, along (therefore) with any consideration of determinism, which even in Russell's formulation certainly does involve causality.

    ReplyDelete
  102. John

    Maybe if we had to include every true fact about the world in the explanans, then it would be a matter of detail. But if that's what's needed for a sufficient explanation, we never sufficiently explain anything.

    ...

    My point here was that it's an unreasonably high degree of detail to require for a sufficient explanation.


    Well that's the question. What counts as sufficient for the purposes of PSR? Surely if determinism is true, the PSR is true since the ability to exactly predict something sufficiently explains it, I would think. But what about the other way? Can we have a case where even given the "unreasonably" high degree of detail we are considering in the definition of determinism we still can't explain an event and PSR is still true?

    The point was never that you *have* to include every detail in the sphere in the explanans for the explanation to count as sufficient. The point was that you *get* to include every detail in the sphere in the explanans before concluding that because you have failed to find an explanation, determinism or PSR are false.

    Suppose we are trying to explain some event and we gather some cursory evidence at the scene but fail to find an explanation. Do we conclude that PSR or determinism are false? Of course not. We keep looking. Now suppose we keep gathering data and now have a complete description of every subatomic particle within one light-second of the event one second before the event and still have no explanation. Well now we have to either reject Relativity or determinism.

    ReplyDelete
  103. Scott,

    …and indeed any consideration of causality at all, along (therefore) with any consideration of determinism, which even in Russell's formulation certainly does involve causality.

    Quite true. I was in error earlier. I apologize. I misread Russell. I still think Russell is using 'causal' in a completely different way than scholastics use it. I think he's using it in a way that can be cashed out entirely in terms of predictions. That's why I've introduced a new definition above.

    Now it may be the case that my original question has already been answered. My original question was how can scholastics think that PSR could be true and determinism false? It seems clear now that one answer is that scholastics take determinism to be making claims about causality as scholastics understand it.

    I don't think determinism is making those kinds of claims at all. I think all determinism is claiming is that a particular kind of mathematical relationship obtains between antecedent and and subsequent states of affairs in the world that could, in principle, be used to make true predictions with certainty. Determinists use terms like causal in their descriptions to mean just those sorts of mathematical relationships.

    So while I think my initial question has been answered there is still this related question: Can PSR be true absent the kind of mathematical relationship I describe?

    ReplyDelete
  104. The regress terminates once every true fact about the world inside the sphere is in the explanans.

    The description of the sphere itself requires truths that are not reducible to facts within the sphere. For instance, the boundary of the sphere is a relation between what is within the sphere and what is not (or, if we assume there is nothing outside the boundary of the 'sphere', the truths that characterize this state), and this entire set-up requires that the boundary of the sphere be taken into account. It's a basic error to think that the truths describing your set-up are somehow all somehow "inside" the sphere and bounded by it. The mathematics itself is not "inside" the sphere; there are infinitely many mathematical truths that that are relevant to everything in the sphere. There are vastly many, perhaps infinitely many, metaphysical truths about alternative ways in which each of those mathematical truths could be physically exemplified, which are potentially relevant to understanding what is going on inside the sphere. Only talking about the inside of the sphere doesn't actually restrict anything when we move to talking about truths and how they are related, because truths are very obviously not related by distances in light-spheres.

    But this is all beside the point, because you didn't answer my question at all, which was about what you meant by prediction. (I'll copy you by putting the blatantly obvious in bold.)

    Determinism is the doctrine that there exists a well defined function (f) such that given a complete description (d1) of the contents of a sphere at time (t1), the function evaluates to a complete description (d2) of the contents of a smaller sphere concentric with the original sphere at some later time (t2) provided that the radius of d2 is no larger than the radius of d1-c(t2-t1), where c is the speed of light. Or in other words d2=f(d1,t2-t1)

    I'm not sure what you mean by a 'complete description'. If you mean the description is exhaustive, i.e., can be represented as the infinite set including every truth that can possibly under any circumstances be stated about the events in the spheres in question, then it's unclear why you think the falsehood of this position, holding Relativity to be true, requires that PSR be false; PSR doesn't make any appeal to complete descriptions in this sense, nor does it seem to require descriptions in such a sense exist, or that there exists any such well-defined function as you suggest here. There seems to be gap between the one and the other. On the other hand, if you mean something else, it would need to be more explicit.

    It's worth noting, incidentally, that your definition really can't just be between two descriptions and times; it has to cover all descriptions and times related in the relevant way. When it's put in causal terms because causality removes happenstance relations; if you take it out and still only talk about two states, it could be sheer accident that that particular function in question relates them, in something analogous to the way that it is utterly useless that there is a mathematical function that gets you from pounds of ice-cream consumed in the United States in years 2015 and 2016 to the number of pockets picked in Rome in 2017 and 2018. There's some expression that's bound to get you from one to the other; it has nothing to do with either predicting or explaining.

    ReplyDelete
  105. John,

    All we need to break an entailment relation is a possibility.

    Is there a possible world in which every fact in the explanans is true and yet something additional happened that instead explains the explanandum?


    That doesn't break entailment. Suppose I claim that A→C and B→C and you deny A→C, but agree that B→C. Now say that we agree that both A and C are true. If you go on to prove that B is true and I accept it, you haven't disproved that A→C.

    ReplyDelete
  106. I don't think determinism is making those kinds of claims at all. I think all determinism is claiming is that a particular kind of mathematical relationship obtains between antecedent and and subsequent states of affairs in the world that could, in principle, be used to make true predictions with certainty. Determinists use terms like causal in their descriptions to mean just those sorts of mathematical relationships.

    Yes, but this is not in fact the historical sense of the word 'determinism'; it's less than a century and a half old, was borrowed by physicists on analogy to a particular (and not universal) philosophical usage of the term that later vanished, and is still almost only ever used this way by physicists and people (like philosophers of science) whose primary association with the word comes from physics rather than other sources. Even today, people calling themselves determinists often don't mean this by the term. It's only come back into philosophical discussions of determinism at all, in fact, by way of arguments about how physics relates to discussions of determinism. It's not surprising that if you go around using the word in this sense that people are going to have some difficulty figuring out what you mean by it.

    ReplyDelete
  107. But this is all beside the point, because you didn't answer my question at all, which was about what you meant by prediction. (I'll copy you by putting the blatantly obvious in bold.)

    Apologies for this, incidentally, which (1) sounds more sarcastic than I intended and (2) is itself beside the point because you weren't actually responding to me but to John. It's still true, though, that we don't really know what you mean by prediction.

    ReplyDelete
  108. Mack,

    I'm talking about the entailment from the correction of February 14, 2016 at 9:55 PM:

    (What I was thinking was that the detailed explanans doesn't entail that “Fred died because he was shot”.)

    If I can refute that entailment, I can refute the principle that "If q explains p, q entails (q explains p)". I intended to go from there to argue that the falsity of "If q explains p, q entails (q explains p)" causes problems for the plausibility of February 14's sixth premise that "If q explains p, q entails p".

    So, I'm still dealing with the sixth premise of the argument I made on February 14, 2016 at 1:44 PM, when I asked:

    It sounds like you're asking if the only way we can avoid the above (or something like it) is by arguing that there are only necessary facts[3], thereby rejecting the second premise. Am I going in the right direction?

    And you replied:

    John, Yes. Exactly. Well put. That's what I'm thinking. Is there a hole in it?

    If we've moved to a different question (and it looks like we have), let me know. I don't want to flog a dead horse.

    That doesn't break entailment. Suppose I claim that A→C and B→C and you deny A→C, but agree that B→C. Now say that we agree that both A and C are true. If you go on to prove that B is true and I accept it, you haven't disproved that A→C.

    It does. The proposition that "'Fred was shot and [...]' entails ('Fred was shot and [...]' explains 'Fred died')" requires that "Fred was shot and [...]" implies that ("Fred was shot and [...]" explains "Fred died") in every possible world.

    ReplyDelete
  109. Mack The Mike:

    "If you go on to prove that B is true and I accept it, you haven't disproved that A→C."

    He doesn't need to. As he said, all it takes to break entailment is a possibility.

    ReplyDelete
  110. Brandon,

    But this is all beside the point, because you didn't answer my question at all, which was about what you meant by prediction.

    I agree that nothing in the first paragraph of your comment is on point, so I won't comment on it. I also agree that I didn't answer your question about prediction in my response to John West. I'm not sure why you would have expected it to.

    Regarding 'prediction', I mean it in the ordinary sense of the term. I think your questions about it are answered by my re-formulation of Russell's definition of determinism. I mean prediction as the output of that function f.

    ReplyDelete
  111. John,

    Please continue if you like. I'm interested in your line of thinking.

    I know that a certain amount of shorthand is needed, but just to be clear. I'm reading propositions like "q explains p" as "q sufficiently explains p". I understand that there are all sorts of partial explanations for things.

    You may be further along than you think. I don't think I believe that "If q explains p, q entails (q explains p)" I just believe that "If q explains p, q entails p". Although I think I might be able to retreat to "If q explains p, q predicts p" without too much damage.

    ReplyDelete
  112. Regarding 'prediction', I mean it in the ordinary sense of the term....I mean prediction as the output of that function f.

    These are inconsistent claims, though; the ordinary meaning of 'prediction' does not involve having a mathematical function of the sort you have in mind. For one thing, you can get this mathematical function long after t2, and it will still output the relevant description for t2 from the t1 description, without being a prediction at all.

    ReplyDelete
  113. Brandon,


    Yes, but this is not in fact the historical sense of the word 'determinism'; it's less than a century and a half old, was borrowed by physicists on analogy to a particular (and not universal) philosophical usage of the term that later vanished, and is still almost only ever used this way by physicists and people (like philosophers of science) whose primary association with the word comes from physics rather than other sources. Even today, people calling themselves determinists often don't mean this by the term. It's only come back into philosophical discussions of determinism at all, in fact, by way of arguments about how physics relates to discussions of determinism. It's not surprising that if you go around using the word in this sense that people are going to have some difficulty figuring out what you mean by it.


    Brandon, I'm certainly no expert on the history of the term, but I think my use of 'determinism' flows from the modern, non-scholastic, understanding of causality as being a constant conjuction of temporally ordered states which is by far how most scientists use it these days. Even this comment section shows this. The early dispute over whether causes can be simultaneous with effects shows it. Outside a scholastic blog comment section I think you'd be hard pressed to find any other usage. Russell was a major public intellectual. My undergrad QM textbook uses determinism in this sense. I don't pretend I'm using 'determinism' in the only sense it is ever used or even the historically original, but I think it's the predominate one now.

    ReplyDelete
  114. My copy of the Oxford Dictionary of Philosophy's entry on Determinism says, in part:

    The usual explanation for this is that for every event, there is some antecedent state, related in such a way that it would break a law of nature for this antecedent state to exist and yet the event not happen

    ReplyDelete
  115. Brandon,

    These are inconsistent claims, though; the ordinary meaning of 'prediction' does not involve having a mathematical function of the sort you have in mind.

    Nonsense. If you want to know where to look for Mars in the night sky of July 7, 2853, you run a set of equations. This sort of thing is very, very common.

    ReplyDelete
  116. I think my use of 'determinism' flows from the modern, non-scholastic, understanding of causality as being a constant conjuction of temporally ordered states which is by far how most scientists use it these days.

    It does not, because up to about the twentieth century everyone who accepted the Humean position on causation opposed positions called 'determinism'. This is heritage going back to Hume, who also uses his account explicitly to oppose it (although it was called the doctrine of Necessity then). Likewise, prior to very late in the nineteenth century everyone who advocates a position called 'determinism' was an opponent of Humean conceptions of causation, the entire point of which was to avoid assuming that one thing necessitates or determines another. The new usage -- and it is new, only become into broader usage in the past forty years or so -- is derived from the adapted usage of physicists, as I already explicitly noted. Thus it obviously comes up whenever physics is involved, as here, or when people give physics-based objections to simultaneous causation. Russell, as I recall, was getting his (nonstandard generally at the time) usage from physicists (among whom it had already begun to be the standard meaning); he was very deliberately doing that. It did not become widely used, even among philosophers, although it much later had a resurgence, again due to the influence of physics. Most people who have accepted the constant-conjunction view in the real world have not been determinists; they think (for the same reason Hume did) that anything might happen at t2, for all one can be sure -- causation is just what we expect to happen on the basis of what usually happens, and for all we can be totally sure the rules might change (although obviously by definition we don't expect that). The only people who use something like the constant-conjunction meaning and hold something stronger than this are physicists, or people getting it from physicists, because their expectations are based on rigorous experiments resulting in highly confirmed mathematical descriptions, which most people are not dealing with when they are talking about determinism (which is usually in the context of discussing free will). Most people don't think of causes in terms of mathematical functions, you know.

    ReplyDelete
  117. My copy of the Oxford Dictionary of Philosophy's entry on Determinism says

    And what is its copyright date?

    ReplyDelete
  118. Scott,

    "If you go on to prove that B is true and I accept it, you haven't disproved that A→C."

    He doesn't need to. As he said, all it takes to break entailment is a possibility.


    '→' means 'entails'. So, yes he does. Breaking the entailment means disproving the entailment, A→C.

    ReplyDelete
  119. Brandon,

    My ODP is copyright 2005, my QM textbook is 1990 and while my copy of "Religion and Science" is copyright 1970 with a new forward, the original was copyright 1935, So I have a 70 year span of consistent usage.

    ReplyDelete
  120. Nonsense. If you want to know where to look for Mars in the night sky of July 7, 2853, you run a set of equations. This sort of thing is very, very common.

    This is an extremely obvious non sequitur; from the fact that you can often predict with equations it does not follow that the ordinary meaning of 'prediction' includes having a mathematical function of the sort you have in mind.

    ReplyDelete
  121. So I have a 70 year span of consistent usage.

    No, you have three data points, one of which is from physics (and thus irrelevant to the question at hand, which is the meaning outside of physics and fields directly influenced by it), one of which (Russell) is a deliberate and at the time not common use of the term as it had begun to be used by physicists, and one of which is relatively recent and from an article that almost certainly was written by a philosopher of physics.

    ReplyDelete
  122. I do have one book on free will on my shelf, "How Free are You?" by Ted Honderich. Honderich distinguishes on page 2 among three uses. The first is the usage I have been using from Russell and Physics, the second is a view about people that has the upshot that people aren't free or responsible, and the third, which he uses throughout the remainder of the book is the notion that people are subject to causal laws. There is no discussion in the text of material, formal, or final causality and his focus in his discussions of causal laws is on events not substances, so I took him to mean 'causal law' in the modern sense (I think I read it about 15 years ago. the copyright is 1993).

    So I really don't think it's the case that outside of physics and in discussions of free will that the Aristotelian/Scholastic concepts of causation are the norm. Here too, the discussion uses the modern post-Hume view of causation.

    ReplyDelete
  123. So I really don't think it's the case that outside of physics and in discussions of free will that the Aristotelian/Scholastic concepts of causation are the norm.

    This is again something that doesn't follow from anything I said. We aren't talking about causation, we are talking about how people outside of fields directly influenced by physics use the term 'determinism'.

    Honderich is, fair enough, someone who does not work in philosophy of physics; but since he does work on issues involve neuroscience and the mind it's very likely that he gets it from someone who does. It's also, again, well within the time period I mentioned above for its spread in philosophy.

    ReplyDelete
  124. The upshot is that you keep talking about the matter as if 'causation' were the only problematic term, in the sense of a term that has shifted its meanings in a very significant way. My point is that 'determinism' is very clearly also a case in which this has happened, and is quite as confusing to people as 'causation'.

    ReplyDelete
  125. Brandon, you write:

    No, you have three data points, one of which is from physics (and thus irrelevant to the question at hand, which is the meaning outside of physics and fields directly influenced by it)

    A few points.

    1. The four (I added one since you posted) data points I have are four more than I've seen from you.

    2. This is a discussion about a question I had regarding Prof. Feser's ideas about PSR as linked to in the main blog post. I specifically referred to Prof. Feser's 12/12/2014 post “Causality and radioactive decay” to which he links from the main blog post because he presumably thought it relevant. Radioactive decay is of interest precisely because it is unpredictable and thus presents a challenge to determinism as the term is used by physicists. Your notion that concepts from physics are irrelevant to radio active decay is, let me be charitable here, odd.

    ReplyDelete
  126. Brandon,

    The upshot is that you keep talking about the matter as if 'causation' were the only problematic term, in the sense of a term that has shifted its meanings in a very significant way. My point is that 'determinism' is very clearly also a case in which this has happened, and is quite as confusing to people as 'causation'.

    In order to clear up my use of the term, which has been ascendant for 80 years, I have defined it explicitly. I don't know what more I can do.

    ReplyDelete
  127. The four (I added one since you posted) data points I have are four more than I've seen from you.

    As the only data point you've actually given that is relevant is Honderich (assuming, that is, that he is not directly influenced by the physics usage), this is merely rhetorical posturing.

    But if you are really interested in the underlying evidence, I recommend you start with McTaggart, who is highly influential on discussions of determinism among philosophers in the first half of the twentieth century, and move forward from there; then look specifically at people who do work in free will -- not people like Honderich who discuss free will as part of their discussion of the philosophy of neuroscience and the mind, but people who are actually talking about free will. Or you can look up Huw Price's occasional complaints that philosophers talking about determinism in matters of physics keep dragging in an incorrect idea that it involves causal necessitation. Or you can get out of your textbooks and go find ordinary people talking about determinism, ask them what they mean by it, and see how it fits with your definition. Or, you know, you could actually use all the difficulty you've had here as part of your evidence. I know it's a limited sample; but you have not actually in any way established that the confusion in the thread above about how causal 'determinism' is, is due to scholastic accounts of causation rather than common meanings of the term that don't necessarily have much to do with such accounts.

    Your notion that concepts from physics are irrelevant to radio active decay is, let me be charitable here, odd.

    This is again something that does not follow from anything I said.

    ReplyDelete
  128. In order to clear up my use of the term, which has been ascendant for 80 years, I have defined it explicitly. I don't know what more I can do.

    (1) It has not been ascendant "for 80 years". It has been the standard usage in physics for that period of time. We are, again, talking about the usage outside of physics and fields directly influenced by it.

    (2) I'm not expecting you to do anything. Your handling of it by defining it was quite admirable. I was pointing out that it was a reason for confusion that you were overlooking.

    ReplyDelete