Monday, May 31, 2010

A simple argument against presentism

  1. (Premise) Only an actually existing event can be seen.
  2. (Premise) A wholly past event can be seen.
  3. Therefore, a wholly past event can be actually existing.
  4. (Premise) If presentism is true, a wholly past event cannot actually exist.
  5. Therefore, presentism is not true.
A simple example for (2): If the moon were to almost instantaneously change color all over, we could see (without any instruments) the event of this change. But we'd see the event about one second after it happened (only relative to our reference frame, but in any case the presentist has no room for such a relativity).

One might want to make the stronger claim that all the events we perceive are past. I don't know if that's true. Maybe we can have direct perception of the future. Maybe a tennis player's noninferential predictions of future ball positions are like that. Moreover, even if that were true, the presentist only says that no wholly past event is actual. Many of the events we perceive are both past and present—we can only see them because they are in part past, but they are still going, and hence in the future.

Saturday, May 29, 2010

Obligatorily directional actions

According to the Christian tradition, sexual activity is impermissible unless engaged with procreation or union as a goal (the latter is variously understood, and "union" isn't always the best word). So just as deontology recognizes actions that may not be done for any end whatsoever, the mala in se, we can also introduce a category of "obligatorily directional action". A type T of obligatorily directional action comes with a set G of goals such that (a) G is a non-empty proper subset of the set of all goals permissible to human beings, (b) G is not defined solely in terms of the amount of value embodied by the goals in G, and (c) an action of type T is impermissible unless it is done with the intention (and not just foresight) of achieving at least one of the goals in G. I don't quite know how to put condition (b). I want to rule out such cases as this. Causing a minor pain can only be done for the sake of a greater good, so if we didn't require (b), we could let G be all goods greater than a minor pain, then any action type that entailed a bad of some paticular sort would count as obligatorily directional. I shall say that G is the set of T's proper goals.

The mala in se are a degenerate case where the set G is empty. The plausibility of the claim that there are mala in se makes it necessary to take seriously (apart from particular claims from the Christian tradition) the possibility that there could be obligatorily directional actions.

I can imagine two kinds of obligatorily directional actions, which I will name the suspicious and the sacred (without that implying any theological content; think of fairly secular uses like "sacred duty"). A suspicious obligatorily directional action is one that carries in itself some innate reason to think it is morally problematic, but that problem disappears when, but only when, the action is done for a goal in G. A sacred obligatorily directional action is one that it would in some way be in some sense (perhaps a fairly secular one) sacrilegious to do except at least in part in pursuit of a particular goal. One might, for instance, think that intentional killing is a suspicious action, and is only permitted when done at least in part as a just punishment or as a defense of the innocent against a wrongful aggressor—then G would include the goals of just punishment and defense of the innocent. On the other hand, one might think that prayer to God is a sacred obligatorily directional action, and that it is wrong to pray except with some sort of communion with God being a goal. Thus, it would be wrong to pray solely because someone will pay us to do so. Indeed, Catholic tradition condemns as a sin someone's attending Mass (solely?) for the sake of secular benefits.

I do not know if the sacred and the suspicious cover all the cases. For instance, I've been drawn to the view that when asserting one is obliged to have the intention of asserting nothing false. I don't think asserting is suspicious. And I don't know that I'd say it's sacred. On the other hand, maybe given that the second person of the Trinity is the Word, it is?

At times, there can be disagreement over whether an action's status as obligatorily directional is because it is suspicious or sacred. Thus, I suspect that many Catholics in the past thought that sexual activity is obligatorily directional because it is suspicious, while the more common current view is that it is because it is sacred. Actually, it could even be that an action is both suspicious and sacred. The sacred is often dangerous, after all.

Note that to every obligatorily directional action type there correspond an infinity of action types of mala in se. For if T is obligatorily directional with proper goals G, then for any g not in G, the type an action of type T done solely for g is a malum in se.

I think the concept of obligatorily directional actions needs more investigation.

Friday, May 28, 2010

Augustine quote

Just came across this in Augustine's The Teacher:

Who is so foolishly curious as to send his son to school to learn what the teacher thinks?

Yet another ontological argument

I've never heard the following version of the ontological argument put quite in this way, though there may be things in Anselm and Descartes that suggest it:

  1. (Premise) An impossible being is imperfect.
  2. A perfect being is possible. (By (1))
  3. (Premise) Necessarily, a contingently existing being is imperfect.
  4. A perfect being exists necessarily. (By 2 and 3)
  5. (Premise) What exists necessarily also actually exists.
  6. A perfect being actually exists. (By 4 and 5)
I don't know if (2) validly follows from (1). Or maybe there is a problem with (1), in that we simply cannot attribute about impossibilia?

Why believe (1)? Well, one line of thought is that impossibility is an impotence. Another is that an impossible property entails all properties, and in particular such properties as being imperfect, and no imperfect being is perfect.

Ontological arguments

Continuing my project of reading my hitherto unread reading assigned for the medieval comp, I've been reading Anselm's replies to Gaunilo. As it happens, I never read all of it. When I assigned the text for class, I assigned only an abridged version which at one point says "Anselm continues as some length, but much of what he says seems repetitive". Well, maybe it seems that way to the translator, but the full text is really good stuff. I haven't digested it all, but I think there may be more to Anselm's ontological argument than has caught my eye before. It's at least as good as the S5 ontological argument.

That said, here's another ontological argument, inspired, if memory serves, by a humorous remark my wife made to me once.

  1. (Premise) To be incapable of existing is a great impotence.
  2. (Premise) Necessarily, anything that is all powerful lacks all impotence.
  3. (Premise) A being that exists and is all powerful in one world must exist in all other worlds.
  4. (Premise) God is essentially all powerful.
  5. God lacks all impotence. (2 and 4)
  6. Possibly God exists. (1 and 5)
  7. There is a world at which God exists and is all powerful. (4 and 6)
  8. God exists in all worlds. (3, 7 and S5)
  9. God exists and is omnipotent. (4 and 8)

Step 3 gets a subsidiary argument. More than one comes to mind. But here is one:

  1. (Premise for reductio) Suppose x exists and is all powerful at w but does not exist at w*.
  2. (Premise) Necessarily, to be unable to be an efficient cause of any sort (remote or immediate, full or contributing, etc.) of a possible but non-necessary state of affairs is an impotence.
  3. (Premise) Necessarily, nothing is able to be an efficient cause of any sort of its own failure to ever exist.
  4. x's failure to ever exist is a possible but non-necessary state of affairs. (10)
  5. It is true at w that x is unable to be an efficient cause of any sort of its failing to ever exist. (12)
  6. It is true at w that x is not all powerful. (2, 11, 14) Which absurdly contradicts (10).
  7. So if x exists and is all powerful at w, it must exist at every other world w*.

I don't know how seriously this argument is to be taken. By the way, it reminds me of something I heard attributed to Scotus.

Thursday, May 27, 2010

Ockham on the paradoxes of entailment

While looking through an edition of Ockham, I came across this interesting text:

Other rules are given:
(10) From an impossibility anything follows.
(11) What is necessary follows from everything.
Therefore this follows: 'You are a donkey, therefore you are God'. This also follows: 'You are white, therefore God is triune'. But these consequences are not formal ones and they should be used much, nor, indeed are they used much. (Summa totius logicae III, III, C. XXXVI)

What is interesting to me is (a) that the consequences are not formal ones (don't follow in relevance logic?) and especially (b) that these rules shouldn't be used much. Why not?

Presumably, the only time you establish an impossibility in a sound argument is as part of a reductio, and if you do that, you don't use (10) next—you close the subproof and use reductio ad absurdum.

And using (11) leads to arguments that are not as perspicuous as they could be. For if you've established necessarily(p), it is more perspicuous to conclude p from necessarily(p) by axiom M of modal logic than to do something like:

  1. necessarily(p).
  2. 2+2=4.
  3. p. (By (12), (13) and rule (11))

However, while (10) and (11) are useless considered as rules of inference, as true propositions they can be quite useful, and do in fact occur in philosophical discussion, for instance in providing counterexamples (suppose you say that x depends on y if and only if exists(x) entails exists(y); then if numbers are necessary beings, everything depends on the number 49, which may seem to be absurd). So we need to distinguish between (10) and (11) as truths and (10) and (11) as rules of inference.

This distinction is needed anyway for modus ponens; for if modus ponens is simply the universally quantified truth:

  1. For all p and q, if (p is true and it is true that if p, then q), then q is true,
then to apply modus ponens given the truth of P and of if P then Q, you will need to do universal instantiation on (15) to get:
  1. If (P is true and it is true that if P, then Q), then Q is true.
And then using the antecedent of (16) as a premise to get to the conclusion Q, one will have to use modus ponens, which lands one in a vicious regress. (This is an argument of Sextus Empiricus against the very idea of rules of logic. But Sextus confuses truths qua truths and rules. Not that I know exactly how to draw the distinction either.)

Wednesday, May 26, 2010

Boccaccio's argument for the Catholic faith

In the second story of the first day of the Decameron, we have the story of how Giannotto tried to convince a Jewish friend named Abraham to become a Christian. Giannotto is a fairly ignorant merchant, but his arguments have sincerity. Abraham, on the other hand, is a theologically well-educated Jew. But instead of making mincemeat of his friend's arguments, out of friendship and perhaps a movement by the Holy Spirit (so the narrator suggests), he resolves he'll go to Rome to see what the alleged vicar of Christ is like, in order to decide which faith is correct. Giannotto thinks all is lost:

if he goes to the court of Rome and sees the wicked and filthy lives of the clergy, not only will he not change from a Jew to a Christian, but if if he had already become a Christian before, he would, no doubt, return to being a Jew.
Nonetheless, he sends his friend with his blessing. Abraham goes to Rome and sees all the sin among "the Pope, the cardinals, and the other prelates and courtiers". Abraham returns, and Giannotto is sure that there is no longer a chance of conversion. He asks Abraham what he thought of the Papal court. Abraham responds:
I don't like them a bit, and may God condemn them all; and I tell you think because as far as I was able to determine, I saw there no holiness, no devotion, no good work or exemplary life, or anything else among the clergy; instead, lust, avarice, gluttony, fraud, envy, pride, and the like and even worse (if worse than this is possible) were so completely in charge there that I believe that city is more of a forge for the Devil's work than for God's: in my opinion, that Shepherd of yours and, as a result, all of the others as well are trying as quickly as possible and with all the talent and skill they have to reduce the Christian religion to nothing and to drive it from the face of the earth when they really should act as its support and foundation. And since I have observed that in spite of all this, they do not succeed but, on the contrary, that your religion continuously grows and becomes brighter and more illustrious, I am justly of the opinion that it has the Holy Spirit as its foundation and support, and that it is truer and holier than any other religion.... So, let us go to church, and there, according to the custom of your holy faith, I shall be baptised.[note 1]

Now, while over the past century we've been blessed by popes of exemplary holiness (though of course there has been much wickedness elsewhere among the clergy and laity), the argument does not require present papal wickedness. What it requires is the surprising way that despite all the wickedness, the Church survives and grows. One might object: but if the Catholic faith were the true faith, wouldn't we expect that the hierarchy would be holy in the first place? While the analogy is not perfect, this is similar to asking, in the case of someone who was apparently miraculously healed, why God would have permitted the illness in the first place. The question is a good and tough one, but it does not make the healing (in the case of the cancer) or the survival and growth (in the case of the Church) less wonderful.

We might enhance the above by recalling another argument. The Catholic Church's formal teaching is coherent, despite having been developed over twenty centuries. The teachings are not only coherent at one time, but are coherent over time (and cohere with Scripture as well, but I don't want to rely on this if the argument is to be convincing to Protestants). The best explanation of this coherence is that it is the work of the Holy Spirit. Arguments along these lines have been developed by Menssen and Sullivan. Observe, too, how this consistency is not observed in most other Christian bodies—sexual ethics is a nice example, with contraception once condemned by all theologians (including Luther and Calvin) and now widely accepted by non-Catholic bodies (with the notable exception of some individual Protestants and some Orthodox bodies—though even in the latter, there is a reluctant acceptance of remarriage after divorce). But now combine this argument with Boccaccio's. The consistency over time is amazing enough—but when one notes that the consistency includes popes who were, apparently, quite wicked, but who, nonetheless, did not formally teach the Church anything contrary to the earlier faith, the argument becomes even stronger.

Monday, May 24, 2010

Plurality of divine ideas

I've been suspicious of divine ideas, but now I like what St Thomas does in S. Th. I.15. St Thomas seems to be insisting that the sentence "There are many divine ideas" gets a semantics according to which it is made true not by some plurality in things, but by God's understanding himself in this way, and in that way, and so on, without these being separate acts of understanding. It is possible, I suspect, for an experienced physicist to understand light as a stream of particles and as a wave simultaneously and without there being two separate acts of understanding here. Likewise, then, God simultaneouosly, and without a multiplicity of acts of understanding, understands himself as something that can be participated in by a donkey, and as something that can be participated in by an oak tree, and as something that can be participated in by an angel, and he does this with a single indivisible (and the indivisibility may not be present in the physicist's case) act of understanding that suffices to make true all of these particular claims about what God understands. So the apparently quantified claim "There are many divine ideas" is made true by a single indivisible entity.

I see Aquinas' project here and in his discussion of the Trinity and the Incarnation as an attempt to provide a semantics compatible with divine simplicity for hard to avoid philosophical truths (this case) or orthodox theological doctrines (the other two cases). This semantics must make the right sentences true, and it must also be plausible.

Sunday, May 23, 2010

Aquinas on the senses of Scripture

The Tradition holds that Scripture has many senses. I found really striking what St Thomas does with this: "The author of Holy Writ is God, in whose power it is to signify His meaning, not by words only (as man also can do), but also by things themselves." In other words, it seems that the Angelic Doctor thinks that the words of Scripture directly only have a literal meaning (which of course isn't the "literalistic meaning"; the literal meaning of an assertoric text is that proposition which is asserted in the text; as Aquinas says, "When Scripture speaks of God's arm, the literal sense is not that God has such a member"). The non-literal meaning is not a meaning of the words of Scripture, but it is a meaning of the realities signified with the words understood in their literal meaning. Thus, the description of the Israelites' crossing of the sea in the Book of Exodus has as its meaning that the Israelites cross the sea. This text signifies a historical event—the Israelites' crossing of the sea. And the further meanings, such as a future baptism in Christ, are had not by the text, but by the historical event itself. The events of salvation history are thus a text, and the non-literal meaning of Scripture is thus not a meaning of the text of Scripture, but a meaning of salvation history itself.

I really like this. If the non-literal meanings of the text were really meanings of the text as such, it is hard to see what would distinguish them from the literal meaning. One alternative is to say that the non-literal meanings are intended by God but not by the human author. But if so, then that makes the human author less fully the author of the text, and it makes God an embedder of secret messages in the text. On Aquinas' view, the human author gets to be fully the author of the text, even if the human author does not grasp any of the non-literal meanings. For the non-literal meanings aren't really meanings of the text. (And there is nothing unusual about the events described by a historian having a meaning going far beyond that which the historian sees in them. But these meanings are of the events, not of the history book.) Moreover, this shifts us from being unduly book-centered. Those who experienced the Sinai event were not in principle worse off than the readers reading about the event. But if the additional meanings were meanings of the text and not of the event itself, then those who experienced the Sinai event were in principle worse off than those who read about it.

All this gives a strong sense to the idea that the non-literal meanings of Scripture depend on the literal meaning. For if the Israelites did not in fact cross the sea, then there is no historical event of the crossing of the sea to bear any of the non-literal meanings. The words of Scripture don't signify a future baptism—they signify a crossing of the sea. The crossing of the sea would signify a future baptism, but since the crossing didn't take place on this hypothesis, that is irrelevant.

Moreover, if Aquinas' idea is right—and I think it is right as the notion that God writes not just in words but in historical events is deeply embedded in the Christian tradition—then the theologian who denies the various miracle stories but hopes to save a non-literal meaning is in even greater trouble. For while we assert by speech and not by silence (unless we set up a special convention—"If I say nothing, assume I agree"), we implicate both by speech and by silence. If the Israelites did not in fact cross the sea, not only is the crossing of the sea not there to bear a non-literal meaning, but the non-existence of the crossing of the sea—i.e., God's refraining from causing a crossing of the sea—carries an implicature that we should be cognizant of. But the implicature carried by refraining from an utterance s is typically (though not always, but a special case would need to be made out that the present case is such an exception) opposed to the meaning that s would have had. So the theologian who reads the miracle stories ahistorically, if Aquinas is right, may well make God out to be implicating something opposed to the non-literal meaning that the theologian hopes to find there.

Saturday, May 22, 2010

Punishment and dependence

Suppose I have extremely good statistical evidence that a random member of the student body at some ordinary college (I am not talking here of some underworld Thieves' University) has committed a crime of type C, and that the statistical evidence is probabilistically as strong as paradigmatic cases of "evidence that excludes a reasomable doubt". (Imagine, for instance, that the prevalence of criminal copyright infringement among college students was a couple of orders of magnitude higher than during the peak of the Napster era.[note 1]) Would it be permissible to convict a random member of the student solely on those grounds?

The idea would, I think, make most people uncomfortable. But why? It can't just be "because he might be innocent". For the same is true of someone convicted on paradigmatic evidence that excludes a reasonable doubt, and the chance of error in the latter case is no less than in the purely statistical case by my stipulations.

My suggestion for what is wrong here is that we want punishments to be dependent—say, counterfactually—on the crime. Even though we can be quite certain that Sally who is a student at the college has committed a crime of the relevant sort, we want it to be the case that had she not committed it, she would not be convicted.

Now, the above story might seem far-fetched. Here is a version of the story that isn't quite so far-fetched. The likelihood that a contract killer reports all his income to the tax authorities is vanishingly small. But we should not tack on a conviction for tax fraud unless we actually check the tax records.

Here is an interesting consequence of the counterfactual dependence thesis. Suppose that responsibility is compatible with determinism, and two days ago we were able to tell with certainty from Jones' brain scan and other data that a day later Jones would torture his uncle, leaving no physical marks, and then would erase all marks of the torture from the uncle's mind. Can we convict Jones of the torture? The uncle denies having been tortured, but that is exactly what we would expect. We have excellent evidence of Jones' guilt. But our evidence does not counterfactually depend on Jones' choice, and hence if we convict him, his punishment will not counterfactually depend on his choice. And hence, even though we have evidence of his guilt that excludes reasonable doubt, we are not permitted to convict Jones if I am right.

Friday, May 21, 2010

Adultery, fornication and marital relations

Start with this intuition: all adulterous actions are intrinsically wrong. Therefore, any action that is intrinsically just like an adulterous action is also wrong. Now the intrinsic character of a successful action is defined by the intentions or action plan—by what the end is and how it is intended to be achieved. Many cases of adultery do not, however, involve an intention to commit adultery. Sam knows that sex with Suzy would be adulterous, but he need not intend the sex qua adulterous. He might intend it qua pleasant or qua unitive-with-Suzy. The distinction is important. There are cases of adultery where there is an intention to commit adultery as such, as when Sam intends to make Suzy's husband a cuckold or make his own wife jealous. Such malicious cases are, ceteris paribus, morally worse than run-of-the-mill adultery done for the sake of pleasure or union.

Thus the intentions in run-of-the-mill adultery are the same as those in typical cases of fornication—to share pleasure with this person, to unite with this person, etc. If adultery is intrinsically wrong, so will these typical cases of fornication be. (And I don't think there are any atypical permissible cases of fornication, either.)

Moreover, so will cases of sexual activity within marriage when the intentions are the same kinds of intentions that typical adulterers and fornicators have. Thus, if Sam's intention is simply to share pleasure with Tamara, he is doing intrinsically the same thing as when he commits adultery with Suzy, even if Tamara happens to be his wife and Suzy doesn't. Thus, if there is to be an intrinsic difference between marital activity and adultery, the marital activity must involve intentions that adulterers cannot have, properly marital intentions such as to unite maritally with Tamara or at least to share pleasure with his own wife, Tamara. It is clear, thus, that it is possible to do something that is relevantly like adultery with one's spouse. Is this why, perhaps, when Jesus said that the man who looks lustfully at a woman has committed adultery with her in heart, he did not limit his remarks to the case of the married man or the married woman? Taking his remarks literally, to look lustfully at one's spouse is to commit adultery with her in the heart. Lust is an essentially non-marital attitude.

Now, for every possible kind of action, there are negative conditions that the intentions have to satisfy. For instance, so that a financial transaction, a hammering of a nail, a drinking of a cup of coffee, a sexual act or an act of teaching be permissible, it must not be done for a malicious ulterior end. However, sexual relations, unlike the hammering of a nail, must satisfy a positive condition on their intentions to be permissible. The intentions must be marital—of a sort that could not be satisfied outside of a marriage.

It is sometimes said that there is something wrong with a couple that stays together only because of their marriage vows rather than because they like each other. Be that as it may, if my above are right, there is something wrong with a couple that stays sexually together only because they like each other. The fact of being married needs to enter into their reasons.

Thursday, May 20, 2010

Text to speech

I've decided to step beyond the limits of the audio books available at librivox by using text to speech to generate audio books from Project Gutenberg files. I'm currently using the Ivona Amy voice, which is a lovely British voice, but I may use Brian, too (for free 30 day trials, download their Ivona Reader). My initial conversions of the text files to ogg were with Expressivo, but my plan is to do the rest of the conversions with a perl script that calls the Windows SAPI interface for the voices. You can get the script here (which I hereby release under the FreeBSD license). It can be used as follows: perl txt2ogg --voice=amy InputFile.txt (this assumes you have no other voice with "amy" in it; if you do, you may need to include more text--the voice=... option takes a perl regular expression). If you want to pre-listen to the voice, do: perl txt2ogg --voice=amy --outloud InputFile.txt. Interrupt with ctrl-c.

Counterfactuals with possible antecedent and impossible consequent

I used to think the following principle was certainly true, if → indicates subjunctive conditionals:

  1. If pq and p is possible, then q is possible.
Obviously, (1) holds on Lewis-Stalnaker semantics. More generally, it holds on any semantics whose picture of counterfactuals is that the antecedent selects one or more worlds in which it holds and the conditional is true if the consequent holds at these, or most of these, worlds.

But there seems to be a counterexample to (1). Suppose God promised to do A, and did A. Moreover, suppose that the circumstances are such that he is obligated to keep this promise (none of the defeaters for the duty of promise-keeping, if there can be any, apply; the promise has not been annulled by promisee's release; etc.). The following appears true;

  1. Had God not done A (or something better), he would have done something wrong.
(The parenthetical is if you believe that you can always fulfill a promise by doing something whose value dominates the value of what you promised.) Here, the consequent is impossible: God cannot do wrong. But the antecedent is possible: it is possible that God not have done A or anything better. After all, A may be a positive action with respect to a creature, while it is possible that God not have created anything.

The argument against (1) does not apply on views of counterfactuals on which relevant background assumptions automatically get imported into the antecedent. For part of the background here will be that God made the promise, and that the circumstances were such that he was obligated to keep it. But views like the Lewis-Stalnaker view, on which (a) there is no importation of background assumption, but (b) rule (1) holds, will have a problem here.

Wednesday, May 19, 2010

Responsibility and consciousness

Consider this claim:
Thesis: It is possible to have two worlds, w1 and w2, and a person x such that: (a) in w1, x is responsible for A, and (b) in w2, x is not responsible for A, but (c) there is no difference between what x is consciously aware of in w1 and in w2.

Here is an argument for the thesis. Note that "responsible" can be read as "morally responsible", "rationally responsible" and "epistemically responsible" as far as the argument goes. One can even replace "responsible" with "praiseworthy", "blameworthy" or "criticizable", in the moral, rational and epistemic senses. I shall assume that times can be represented with real numbers.

  1. (Premise) No human being is aware of any mental episode that lasts no more than d0=10−22 seconds.
  2. (Premise) There is a human being x and times t1 such that (a) x exists between t1d0/2 and t1 inclusive, (b) x exists at t2, (c) x is not responsible for anything at or prior to t1 and (c) x is responsible for something at t2.
  3. (Premise) Whether there is something that x is responsible for at t does not depend on what happens after t.
  4. (Definition) Let T be the set of times at which x is responsible for something.
  5. All the members of T are greater than t1. (By 2c and 4)
  6. Let t* be the infimum of T, i.e., the largest real number t* with the property that every member of T is greater than or equal to t*. (t* exists by a standard theorem about real numbers as T is bounded below by 5.)
  7. t*≥t1. (By 5 and 6)
  8. (Premise) For some x satisfying the above premises, there are worlds w1 and w2 whose laws and arrangement of matter are such as to still satisfy (1) and (3) and that are such that (a) w1 is exactly like the actual world up to t*−d0/2 and at t*−d0/2, x's mind is wiped in such a way that x is never aware of anything (except perhaps some innocent bliss, if divine goodness does not permit eternal unconsciousness or cessation of existence) and is never responsible for anything; (b) w2 is exactly like the actual world up to t*+d0/2, and at t*+d0/2, x's mind is wiped such a way that from t*+d0/2 on, x has the same mental episodes in w2 as in w1.
  9. In w1, x is never responsible for anything. (For x is not responsible for anything in the actual world up to t*−d0/2, and by 8, w1 is just like the actual world up to t*−d0/2, and we have 3.)
  10. In w2, x is responsible for something at some time after t*. (For x is responsible for something in the actual world at some time between t* and t*+d0/2 by the definition of t*, and by 3 and 8, x is responsible for something at w2 at some such time.)
  11. x's mental episodes diverge between w1 and w2 only on an interval of times of length at most d0. (By 8)
  12. x is not consciously aware of anything different at w1 than at w2. (Because 8 guarantees that 3 holds at both of these worlds)
  13. Therefore, w1 and w2 differ in whether x is responsible for something but not in what x is consciously aware of. And that's the thesis.

I think the best way out of the argument would be to deny (1). I think (1) is pretty plausible if naturalism is true--d0 is less than the time it takes light to cross the Bohr radius of the atom--but since naturalism is false, it might be that (1) is false as well. Personally, I have little desire to get out of the argument. One might also think vagueness will get one out of the argument. But I do not think something as basic as moral, rational or epistemic responsibility is something there is vagueness about.

The insatiaty of the will

I committed myself to read all the texts for our medieval comprehensive exam that I haven't already read so I can be a minimally competent grader. The reading will probably be giving rise to various posts (already hss).

Aquinas argues that our beatitude can only consist in God. The argument is interesting:

the object of the will, which is man's appetite, is the universal good, just as the object of the intellect is universal good. From this it is clear that nothing can put man's will to rest except the universal good. But the universal good is found only in God and not in any created good, since every creature has participated goodness. Hence, only God can satisfy man's will....

An initial worry is that the argument rests on an equivocation in "universal". The object of the will is the universal good in the sense that every good that a human can have is an object of the will. And God is the universal good in the sense that all goods are goods by participation in God.

Here is a more charitable take on the argument—whether it's what the angelic doctor means, I don't know. Every created good that a human can have is a good we desire. We cannot have them all, however. For instance, no matter how many friends we have, we could wish for people with a new configuration of characteristics to be friends with. Even if over the course of eternity we were to be friends with all possible kinds of friends, we couldn't simultaneously be friends with an infinite number of people. At any given time, we can only enjoy a finite number of goods. Now, in this life, we sometimes feel ourselves satisfied by a single created good—say, when we are engrossed in a wonderful conversation. But this is due to our lack of sensitivity, due to the fact that the presence of the good blocks out the fact that we lack other goods. True beatitude is not built on lack of sensitivity. Moreover, a component of a purely created happiness will be a commitment, for a length of time, to a form of activity, and such commitments can be incompatible, with there being a kind of sorrow that one is not simultaneously engaging in others.

If we are sensitive, we appreciate every created good we are capable of having, but we cannot have them all. So, if we are limited to created goods, our happiness will always either be blind or have a sorrow. Limited to created goods, our will is insatiable. But if it is possible to possess that by virtue of participation in which all the created goods are good, then by possessing that one being, one would satisfy the universality in our will. If we have that by participation in which friendship with Albert Einstein is valuable, then we do not need friendship with Einstein to satisfy the will. Thus, one kind of universality in our will—the universality of every—is satisfied by the other kind of universality.

Tuesday, May 18, 2010

Is there a cost to two-boxing in Newcomb cases?

The following may be well-known to folks in the field, or it may be well-known to them to be mistaken.

It seems to be acknowledged by both sides that it is a significant cost of being causal decision theorist that one has to two-box in the Newcomb Paradox, and hence one loses out in Newcomb cases.

I suspect the causal decision theorist should not grant that this is a cost of the theory, or at least not a significant one. First, distinguish between Newcomb cases where there is weak counterfactual dependence between what one chooses and what the predictor predicted and ones where there is no weak counterfactual dependence. (I say that B weakly counterfactually depends on A provided that B might not have happened had A not happened.)

Case A: Weak counterfactual dependence. David Lewis thought that wherever there was counterfactual dependence between non-overlapping events, there was causal dependence. I think he was wrong. E.g., that God promised A counterfactually depends on A, since the non-occurrence of A entails the non-occurrence of the divine promise, at least given prior linguistic conventions. Nonetheless, I think that counterfactual dependence, and even weak counterfactual dependence, is generally a pretty good indicator of at least explanatory dependence. Perhaps something like this principle is true: If (a) B weakly counterfactually depends on A and (b) there is no event C such that (b1) C is not weakly counterfactually dependent on A and (b2) C together with B entails A, then B is explanatorily dependent on A. (Condition (b) rules out the divine promise case: let C be the linguistic conventions that God's act of promising depends on.) Maybe some additional conditions are needed. Nonetheless, I am optimistic that in the Newcomb case, from the existence of a weak counterfactual dependence between my choice and the prediction it follows that my choice is explanatorily prior to the prediction. (Condition (b) is satisfied, at least assuming one's choice is indeterministic. I don't know what to do about deterministic choices. I don't know if the notion makes sense. Nuel Belnap once said that to make a choice, you have to have choices.) But since the causal decision theorist should be willing to generalize causal dependence to explanatory dependence, the causal theorist in this case says to one-box.

Case B: No counterfactual dependence. In this case, the following counterfactual is true for Tamara the two-boxer: She would have got less had she chosen only one box. For had she chosen only one box, the predictor would still have made the same prediction that was in fact made, and Tamara would have done poorly.

But wouldn't the two-boxer have gotten a lot more had she been a one-boxer? Yes. Here, we need to distinguish two antecedents of counterfactuals:

  1. Tamara has a disposition to one-box in Newcomb cases like this one
  2. Tamara one-boxed in this Newcomb case.
And so:
  1. Had (1) been true, Tamara would have got a lot more than she got.
  2. Had (2) been true, Tamara would have got a bit less than she got.
This means that we need to distinguish between two different questions of rationality. Should Tamara have a disposition to one-box and should Tamara one-box in this case?

Perhaps the causal theorist could say: Tamara should have a disposition to one-box and should two-box in this case. There is nothing absurd about that sort of an answer, and according to SEP, it's a standard move here. There are, after all, standard cases where it is (narrowly self-interestedly) rational to be irrational, such as where you will be killed if you are thought to be a rational witness to a crime.

But there may be a better answer. Fact (3) does not entail that Tamara should have a disposition to one-box. All that fact (3) tells us is that if the only situations Tamara faces are Newcomb ones, then she would do better to have a disposition to one-box. But there are other possible situations where someone with a disposition to two-box would do better. For instance, cases where an evil two-boxing philosopher becomes a dictator and kills everyone who doesn't have a disposition to two-box. (Or cases where people are mistaken in thinking the predictor accurate.) So in Newcomb situations, Tamara would do better were she by disposition a one-boxer. But in crazy two-boxing philosopher situations, Tamara would do better were she by disposition a two-boxer. In other words, both of the competing theories have the consequence that in some worlds it is better to self-induce a disposition to abide by the opposite theory. And hence there seems to be little or no special cost here to being a causal decision theorist.

Monday, May 17, 2010

An interesting thesis in Aquinas

In Summa Theologica I-II 1 2 repl. 2, Aquinas makes the interesting claim:

To order something toward an end belongs to one who impels himself toward that end.
Aquinas' claim here seems to be that if A has a teleological directedness at E, then if B is a cause responsible for A's directedness at E, then B is also directed at E. If this is correct, then the telè of God's creatures must all be goods that God's goodness impels him to. Our ends must be God's ends, and hence we have a metaphysical argument for the benevolence aspect of divine love.

Is Aquinas' thesis true? This sort of thing would be a counterexample: As a computer science class exercise, I am suppose to make a computer program that sorts an array of numbers. I thereby order the program toward the end of sorting the array of numbers, but I am not myself impelled to sorting the numbers—in fact, I don't care about sorting the numbers, because my grade depends on the program, not on the actual sorting. I think Aquinas has to say that the computer program is not really impelled to sorting the numbers. It is, at best, impelled to getting me a good grade. Or maybe Aquinas will simply deny genuine teleology in artifacts.

Sunday, May 16, 2010

Another fun case in favor of causal decision theory

Imagine scientists find that people having brain cancer have a subconscious desire to drink orange juice, and consequently on any given occasion of choosing what to drink, they are about 10% more likely to choose orange juice than if they don't have brain cancer. The also find that drinking orange juice has no causal effect on the progress or development of a cancer. Suppose background information such as the frequency of headaches and family history gives me a 0.001 probability of having brain cancer.

So, now I am choosing between orange and grapefruit juice. I like the taste of orange juice a little more, but sometimes I drink grapefruit juice. On a non-causal, purely Bayesian, decision theory, it seems I have reason to drink grapefruit juice. For, suppose, P(drink orange juice|cancer)=0.75 and P(drink orange juice|no cancer)=0.65. Then, plugging into Bayes' Theorem, we get P(cancer|drink orange juice)=0.00115 and P(cancer|drink grapefruit juice)=0.00087. Now the utility of the taste of orange juice is maybe 1.0 and of the taste of grapefruit juice is 0.8. But the disutility of brain cancer is at least as bad as -30,000. (Quick calculation to make sure this is in the right ballpark: I would be very willing to give up a daily pleasure equal to drinking one glass of orange juice for life to certainly prevent brain cancer, and if my favorite fruit juice has utility 1.0, and I live for 75 years, that's 27394 units of utility I would be willing to forego to avoid brain cancer.) Now, Utility(drink orange juice)=1.0+(0.00115)(-30000)=-33.5 and Utility(drink grapefruit juice)=0.8+(0.00087)(-30000)=-25.3. So I should drink grapefruit juice, even though I don't like it as much.

But it seems to be an established fact about health policy and health advice that non-causal correlations are not to be acted on. Hence, we need causal decision theory.

[Fixed a crucial typo. -ARP]

Friday, May 14, 2010

Newcomb's Paradox and Pascal's Wager

Let Egalitarian Universalism (EU) be the doctrine that God exists and gives everyone infinite happiness, and that the quantity of this happiness is the same for everyone. The traditional formulation of Pascal's Wager obviously does not work in the case of the God of EU. What is surprising, however, is that one can make Pascal's Wager work even given the God of EU if one thinks that Bayesian decision theory, and hence one-boxing, is the right way to go in the case of Newcomb's Paradox with a not quite perfect predictor.

Here is how the trick works. Suppose that the only two epistemically available options are EU and atheism, and I need to decide whether or not to believe in God. Given Bayesian decision theory, I should choose whether to believe based on the conditional expected utilities. I need to calculate:

  1. U1=rP(EU|believe) + aP(atheism|believe)
  2. U2=rP(EU|~believe) + bP(atheism|~believe)
where r is the infinite positive reward that EU guarantees everybody, and a and b are the finite goods or bads of this life available if atheism is true. If U1 is greater than U2, then I should believe.

We'll need to use our favorite form of non-standard analysis for handling infinities. Observe that

  1. P(believe|EU)>P(believe|~EU),
since a God would be moderately to want people to believe in him, and hence it is somewhat more likely that there would be theistic belief if God existed than if atheism were true (and I assumed that atheism and EU are the only options). But then by Bayes' Theorem it follows from (3) that:
  1. P(EU|believe)>P(EU|~believe).
Let c=P(EU|believe)-P(EU|~believe). By (4), c is a positive number. Then:
  1. U1U2=rc + something finite.
Since r is infinite and positive, it follows that U1U2>0, and hence U1>U2, so I should believe in EU.

The argument works on non-egalitarian universalism, too, as long as we don't think God gives an infinitely greater reward to those who don't believe in him.

However, universalism is false and one-boxing is mistaken.

Epistemic probabilities, decisions and determinism

Suppose I have a brother, and I want to ask him to lend me money. When I go to him asking him for a loan, I need to decide whether to wear a blue shirt. How should I decide? An obvious answer is:

  1. I need to evaluate the conditional epistemic probabilities P(I get loan | I wear blue) and P(I get loan | I don't wear blue), and act according to the higher of these.

The obvious answer is wrong. Here is a case. My brother is completely color blind. But my mother informs me that

  1. my father has brought me up to have a tendency to wear blue shirts if and only if he brought up my brother to have a tendency to give loans to relatives
If I know (2) with certainty (or even high credence), the epistemic probability P(get loan | wear blue) is higher than P(get loan | don't wear blue), because the information in (2) induces a correlation between loan-getting and wearing blue. However, this correlation gives me no more reason to wear a blue shirt than the woman who wishes to avoid a hereditary disease afflicting Germans has reason to move to France. (One can sharpen the case by supposing the brother isn't colorblind but has a slight distrust of people who wear blue, with the probabilistic effect of that distrust being smaller than that of the tendencies in (2); in that case, I have positive reason not to wear the blue shirt, contrary to (1).)

This has got to be in the literature. It's reminiscent of standard examples involving probabilistic theories of causation.

So what is the right answer as to how to decide? My strong intuition is that I need to as best I can estimate the integral of P*(I get loan | Q & S=x) dP(x), where P* is objective chance, S is an epistemic random variable representing the complete state of the world (including the laws) just before my choice and P is epistemic probability measure on the set of values of S compatible with my making a choice, for Q = "I choose to wear blue" and Q = "I don't choose to wear blue". Since P*(I get loan | I choose to wear blue & S=x) = P*(I get loan | I don't choose to wear blue & S=x), for every x, given the brother's colorblindness, the two integrals are equal, and so I don't have reason either way with respect to the shirt. And that's the right answer.

But notice an interesting fact. If determinism holds, then for any complete state x of the world just before my choice, either S=x entails that I will choose to wear blue or S=x entails that I won't choose to wear blue. In the former case, P*(I get loan | I don't choose to wear blue & S=x) is undefined, and in the latter case P*(I get loan | I choose to wear blue & S=x) is undefined. Thus, in the two integrals I am supposed to compare, the values of the integrands are never both defined at the same time, if determinism holds. Therefore, if the above is the right way to make decisions—and I think it is—then knowing determinism to hold would make decision theory non-viable.

Thursday, May 13, 2010

Shrinking block: An underexplored view in the philosophy of time

The Shrinking Block (SB) theory of time holds that only present and future objects and events exist—our spatiotemporal reality is a four-dimensional block that ever shrinks. I have never seen anyone advocate SB. However, interestingly, SB has some advantages over either Presentism ("only the present is real") and Growing Block (GB; "only the present and the past are real"). For instance, it is puzzling on Presentism and GB why it is that we should worry more about future physical pains than about past ones, and why we shouldn't worry about past physical pains. After all, on those two theories, future pains don't exist. SB neatly explains this: future pains are real, but past ones aren't. Moreover, SB neatly explains why it is better to have a past of vice and a future of virtue rather than the other way around, which neither Presentism nor GB on their own explain. SB fits with such maxims as "Let bygones be bygones" and encourages us not to "live in the past."

Of course, SB has disadvantages. It fits better with a reformative view of punishment than with a retributive one, and I think retribution is going to be an ingredient of any good theory of punishment (retribution is at least needed to place limits on punishment). Moreover, SB has the same difficulties with propositions about the past that GB has with propositions about the future and that Presentism has with propositions about the past and future. However, if we take the pragmatic stance that what matters most is deliberation, then it is propositions about the future that are the most important anyway.

Now, as I said, I've never seen anyone advocate SB. Why are GB and Presentism more popular? One answer is that both GB and Presentism may seem to have the advantage of accommodating the intuition which some have that the future is open, in the sense that it is neither the case that there will be a sea battle nor that there will be no sea battle. But GB and Presentism do not by themselves yield an open future. Presentists hold that bivalence about the past should hold and that there are non-trivially true propositions about what happened in the past, and if they have the right to say that, then by the same token it is compatible with GB and Presentism that bivalence about the future should hold and that there are non-trivially true propositions about what will happen. In other words, to get any help with regard to an open future, GB and Presentism have to be supplemented with either the thesis

  1. Bivalence fails for propositions about the future, and it is neither true nor false that there will be a sea battle tomorrow
or the thesis
  1. "Not (will p)" does not entail "will (not p)" even when conjoined with the thesis that time will go on, and it is false that there will be a sea battle tomorrow and it is false that there will not be a sea battle tomorrow.

But why can't SB be supplemented at least with (1)? According to SB, there are future events, yes. But SB is compatible with saying that it is not true that there is a sea battle tomorrow, and it is also compatible with saying that it is not true that will not be a sea battle tomorrow. Moreover, unless bivalence is a correct rule of logic—which Type (1) Presentists and GBers deny—why couldn't SB be comaptible with saying that both it is not true that there is a sea battle tomorrow and that it is not true that there will not be a sea battle tomorrow?

Here is a tempting line of thought. If SB holds, future events make propositions about the future true. Now, it either is or is not true that tomorrow's sea battle exists. If it exists, then it's true that tomorrow there will be a sea battle. If it doesn't, then it's not true. But there are two objections to this line of thought. First, the line of thought assumes a truthmaker/falsemaker principle about the future? But why should the SBer accept it? After all, neither the Presentist nor the GBer do! Maybe the line of thought is: "If there are some future events, then they are apt truthmakers for some statements about the future, and so the SBer should accept an appropriate truthmaker principle about the future." But this seems a shaky line of thought: it moves from there being some truthmakers to a full truthmaker principle. Second, the line of thought assumes bivalence about what does or does not exist. But if one can deny bivalence about what will or will not exist, why can't one deny bivalence about what does or does not exist?

Moreover, the SBer can accept (2). She can simply say that as a matter of fact neither tomorrow's sea battle nor the non-occurrence of tomorrow's sea battle exist. Granted, if she says this, she is committed to a changing future—maybe in an hour, once the free choice is made, tomorrow's sea battle event will pop into existence. That's weird, but perhaps not much weirder than in an hour it coming to be true that there will be a sea battle tomorrow, which all open futurists think can happen.

In fact, the above shows that if one can be an open futurist, one can be an eternalist and an open futurist.

Now, if there is very little reason to prefer GB or Presentism to SB, and SB is absurd, we should say that GB and Presentism are little better than absurd. But SB is absurd. Ergo, etc.

Wednesday, May 12, 2010

Video games as art

Roger Ebert recently argued that not only are video games now not art, they never will be, or at least won't be in the lifetime of any gamers now alive. I see three arguments implicit in Ebert:

  1. Video games are, in fact, of fairly low artistic value.
  2. Video games involve significant user interaction and hence are not sufficiently expressions of an individual or collective author's vision to count as art.
  3. Video games have rules and objectives.
Now, first we need a distinction. We use the word "art" in two ways: we use it as implying high artistic value (the evaluative sense), and we use it as describing the kind of thing we're dealing with (the generic sense). Thus, in the evaluative sense, my children's drawings, precious as they are to me, are not "art". But they certainly are in the generic sense. The question whether video games are art is, thus, ambiguous between these two senses. To be art in the generic sense, an item can have very low artistic value. Hubert Lanzinger's "The Standard Bearer" is art in the generic sense, though of absurdly low artistic value. Argument 1 addresses the question whether video games are art in the evaluative sense; arguments 2 and 3 address the question whether they are art in the generic sense.

Take the generic sense first. It seems that a sufficient condition for x to be art in the generic sense is that it be appropriate to judge x on its artistic value. This is a low bar, and intentionally so. Industrial design counts as art. One might want to raise the bar by requiring that the primary axis of evaluation be that of artistic value. But that (a) is mistaken and (b) likely not to have much effect anyway. It is mistaken, because Dante's Divine Comedy would still be art in the generic sense even if it were primarily intended for the moral improvement of the reader (maybe it even was!) and that were the primary axis of evaluation. (In fact, one might think that for any human creation, the primary axis of evaluation is the moral, since that is always what matters most, on Socratic grounds.) Moreover, one could always slice more finely and get around the "primary axis of evaluation" condition. Thus, instead of asking whether the Mac Plus's primary axis of evaluation is that of artistic value—obviously not—we would ask about the non-functional features of the shape of its case or something like that, and if we narrowed it down enough, we could get to something where the primary evaluation should be artistic (unless, as I think, the primary axis is always the moral).

Now, it appears perfectly reasonable to judge of the artistic value of games. For instance, Ishido (of which I once made a free PalmOS remake) is of higher artistic value than Pong. Nor is this just due to the availability of better technology by the time Ishido came out. After all, The Zork Series and Myst are of much higher artistic quality than Leisure Suit Larry: Box office Bust, even though the Zork Series was made in the late 70s and early 80s using a text-only interface, Myst was made in the early 90s with fairly static 2D graphics, while LSL:BOB was made in 2009 and uses fancy animated 3D graphics (I haven't played LSL:BOB, though I have played Zork, and I am basing my judgment of LSL:BOB on the reviews and this trailer). (I don't know if Myst is better than Zork—I am inclined to give a slight edge to Zork, but I think a lot of people will disagree.) If at least some video games weren't generically art, these judgments would be inappropriate.

What about Ebert's arguments 2 and 3. First, take 3. Here, I think the problem with the argument is that it is not clear why rules and objectives are incompatible with art. Each genre has its distinguishing features, and rules and objectives are distinguishing features of games. We might as well say that film isn't art because it has motion, or that some beautiful sword-like object is art if and only if it can't cut.

Argument 2 is interesting—it puts a heavy emphasis on authorship. However, the claim that I am defending is that the game, rather than the gaming, is the art. And the game, as such, has authorship, and the author (who may be a collective) guides the viewer's interaction with the games. This is true of many, and maybe all, forms of art. The great painter draws attention to certain features of the painting, guides the viewer's eye along certain paths. There are some things one only sees when one comes closer; others, when one steps further back. Without that interaction, which the great painter plans for (explicitly or implicitly), the full appreciation of the game is not possible. Or think of a formal garden or (to use one of Ebert's examples) a cathedral as a work of art—the viewer must physically walk through the work, along certain predefined paths, to fully appreciate it. Or take a bard who carefully watches his audience's reactions, and adapts her poetry, with certain outlines, to the audience. This not only does not transform her work into the generic realm of non-art; it does not even decrease the artistic value of her performance. The video game in an interesting way combines features of the garden and the bard's performance: like in the case of the garden, the artist is no longer involved during the enjoyment of the art, but had set up the experience earlier, and like in the case of the bard, there is a dynamism and change. Interactive art is no less art, in the generic sense.

The evaluative case is a matter of degree. I've seen works in respected art galleries of lower overall artistic value than Zork, Myst, Ishido and the flash games from Amanita Design (one can get Samorost 2 with the by-donation Humble Indie Bundle right now). Some of these are by respected artists. Do I think the best video games equal in quality the best works of other genres? Maybe not, but cross-genre comparisons may not be very objective. In my subjective judgment, the artistically best novels, poems, films and paintings are better art than the artistically best video games. However, note that novels, poems, films and paintings have all been produced for a significantly longer amount of time than video games, so the comparison is not fair. I am also inclined to think that the artistically best video games are pretty close in artistic quality to some of the best individual short stories (not considered as parts in a larger collection). And I am inclined to think the artistically best video games may exceed in artistic quality all but the very best of secular sculpture. And, yes, I really like short stories, though I am not so fond of secular sculpture. But as I said, cross-genre comparisons are not very objective—here, they are mainly statements of my aesthetic feelings.

In any case, once one admits that generically, video games are or can be art, any evaluative prediction that it will never, or not in our lifetime, be great art is dubious. On what grounds could one say that some genre will never be great art? Well, first, the genre could be very narrow, so narrow that we can predict what could be done within it. Maybe some particular subgenre of science fiction or detective fiction is like that, though even there I would not express confidence that great art will not come from the subgenre. Even inductive evidence is not very helpful, because we know that great artists can transform a genre. But all that said, the video game does not exhibit that narrowness of genre. Video games can have almost any subject. They can range in form from very close to a live-action film, to something close to an animated film, to something close to a novel or short story (those the last genre has largely atrophied out with the end of Infocom), and the form and degree of interaction, as well as the amount of structure, can vary a great deal. The other way that one might predict that a genre cannot be great art is if it suffers from severe technical limitations. For instance, music made only by hitting one's teeth against each other without removing them from the jaws is unlikely to rise to the form of great art. But, obviously, video games do not suffer from significant technical limitations.

Monday, May 10, 2010

Double effect reasoning without a Principle of Double Effect

Consider a fairly normal formulation of the Principle of Double Effect (PDE). An action that is foreseen to result in an evil is permissible if:

  1. the action is not intrinsically wrong
  2. at least one good is intended
  3. no evil is intended, either as a means or as an end
  4. the foreseen evils are not disproportionate to the intended good or goods.
In the previous post I argued that if this (or something like it) is true, it isn't just a peripheral principle, but all of normative ethics—a necessary and sufficient condition for permissibility. Unfortunately, I do not think this is true, not because of any difficult issues about intentions, but for the simple reason that there are many ways for an action to go wrong, and (1)-(4) do not exclude all of them. Suppose, for instance, I drop bombs on the enemy headquarters, even though I know that some civilians in the vicinity will die. It may well be that (1)-(4) are satisfied. But that is not enough for permissibility if, for instance, my commanding officer has forbidden the bombing or I am under a valid vow of non-violence. Yet that the bombing was forbidden or that I am under a vow of non-violence does not affect (1)-(3).

Granted, the proportionality question is going to be affected by the command or vow, but proportionality does not do justice to the reasons that come from commands or vows. Proportionality weighs goods, while commands and vows give rise to exclusionary reasons, which make some goods no longer count. Of course, one could redefine proportionality to take into account all the reasons available, including the exclusionary ones, but if we do that, then (4) will presumably by itself be sufficient for permissibility. Moreover, the sort of "proportionality" that will let one adequately account for reasons arising from commands and vows will likely just be another word for permissibility, and then the account is trivial. Moreover, intuitively, even when one considers the evil of disobeying one's commander's prohibition, the bombing of the enemy headquarters could be proportionate. But when one's commander prohibits it, the bombing is no longer an act of war, but a private lethal act which one has no right to perform.

Now, one could add conditions like that the action is not forbidden by vow, promise or command. But the resulting PDE would look ad hoc, and I don't think we could be sure we listed everything needed.

So what is to be done about PDE? Here is a short and dogmatic suggestion. One of the basic deontic moral intuitions is that one should produce no evil. However, as soon as we start reflecting on the world around us, we realize that many of our actions have bad consequences for some people. A letter of recommendation that I write for my student is likely to either hurt my student or hurt my student's competitor. When I cross the road, I incur the harm of an increased risk of being run over. And so on, in various day-to-day things. Moreover, there are less day-to-day cases, such as the polio vaccine manufacturer who knows that the vaccine will kill some patients, but also knows it will save more lives. The consequentialist solution is to refine "Produce no evil" into "Do nothing that produces less utility than you could produce." I think it's easy to see that this doesn't do justice to the deontic "Produce no evil" insight.

The basic insight of double effect reasoning is that "Produce no evil" should be refined into as "Intend no evil", with a supplement of "Do nothing disproportionate." Discomfort over trolley cases then shows that we are sometimes unsure whether "Intend no evil" (together with the proportionality condition) really does capture all of the force of "Produce no evil", but I think "Intend no evil" does in fact come close to capturing the force. (I prefer: "Accomplish no evil.")

What is the PDE, then? It is simply an observation of the conditions under which the refinement of "Produce no evil" is satisfied. Seen in this way, it does not provide a sufficient condition for permissibility. It does provide a necessary condition for permissibility, and satisfaction of the conditions does show that the action is permissible insofar as the deontic question of the production of evil is concerned. But there may be other deontic questions. Seen in this way, the PDE can be simplified greatly. It simply says that "Produce no evil" is not violated when one intends no evil and does not act disproportionately.

Sunday, May 9, 2010

Is the Principle of Double Effect the sum total of normative ethics?

The Principle of Double Effect (PDE) is often stated in something like this form: An action that is foreseen to have an evil effect (maybe better: an effect that is a basic evil—I shall not worry about this) E is permissible if:

  1. the action is not intrinsically wrong
  2. a good is intended
  3. E is not intended, either as a means or as an end
  4. E is not disproportionate to the intended good.
This sort of formulation is obviously incorrect. After all, conditions (1)-(4) which are supposed to be sufficient for permissibility are compatible with the action also being intended to have another evil effect E*. That particular problem is easily fixed. We just say that an action that is foreseen to have at least one evil effect is permissible if:
  1. the action is not intrinsically wrong
  2. at least one good is intended
  3. no evil is intended, either as a means or as an end
  4. the foreseen evils are not disproportionate to the intended good or goods.
This is better.

Now, observe two interesting facts. First, if an action that is foreseen to have at least one evil effect and that satisfies (5)-(8) is permissible, then a fortiori an action that is not foreseen to have any evil effects and that satisfies (5)-(8) should also be permissible. (It would be really, really weird if an action would become permissible as soon as one noticed some tiny evil side-effect.) So, in fact, the PDE gives a set of sufficient conditions for permissibility. Second, it is clear that (5) and (8) are necessary conditions for permissibility. Moreover, if (7) weren't a necessary condition for permissibility, there would be little need for a PDE. Finally, it is plausible that an action that aims at no good is a perversion of will, and hence on Natural Law grounds (6) will be necessary. Moreover, it may even be the case that every action has to intend at least one good in order to be an action, and hence (6) may be trivial.

If so, then if (5)-(8) are the right conditions to put in the PDE, they are a complete account of permissibility. This makes the PDE not just something peripheral to a deontic ethics, an epicycle for handling some wartime and medical cases, but in fact it makes the PDE be all of normative ethics.

However, as tomorrow's post will show, a PDE like (5)-(8) is not the right way to think of the insights embodied in double effect reasoning (to use Cavanaugh's phrase).

Friday, May 7, 2010

Translating A-sentences to B-sentences

I won't spell out a full method here, but I will give an example.

  1. It was sunny five minutes ago.
  2. It is sunny five minutes prior to t0 (where I let t0 be the time of this utterance).
This translation method makes use of a communicative tool that is always available to us—to stipulate new terms or symbols. Note that the parenthetical phrase is not a part of the assertion and while it is grammatically declarative, this is a declarative of stipulation rather than a declarative of assertion. In fact, the parenthetical functions exactly like prefixing with the grammatically imperative sentence: "Let t0 be the time of this two-sentence utterance."

This avoids the standard problems with the token-reflexive method of translating A-sentences to B-sentences, such as the problem that we don't want the B-sentence to entail the existence of an utterance when the A-sentence does not. Indeed, (2) does not entail the existence of an utterance or of myself. In embeddings, the parenthetical remark functions in the wide scope. For instance:

  1. It might have been that there are no speakers now.
  2. It might have been that there are no speakers at t0 (where I let t0 be the time of this utterance).
(I don't translate "It might have been" because it is just a colloquial way of writing the box operator rather than something for a theory of time to take account of.) There is no absurd claim that there might have been no speakers at the time of this utterance, but only that there might have been no speakers at t0.

In English, and I assume in many other natural languages, we have two ways of picking out what was said: sometimes we pick out the sentence and sometimes the proposition. Thus:

  1. A: It is sunny now. B: I hope you will always remember that.
  2. A: It is sunny now. B (after a pause): That's no longer true.
In (5), B is picking out A's proposition, and hoping A will remember it. (Maybe it's a romantic moment.) In (6), maybe it has started raining, and B is picking out A's sentence, and saying it's no longer true. Now, here are my translations of (5) and (6):
  1. A: It is sunny at t0 (where I let t0 be the time of this utterance). B: At t1, I hope that after t1 you always remember that (where I let t1 be the time of this utterance).
  2. A: It is sunny at t0 (where I let t0 be the time of this utterance). B (after a pause): At t1 a restatement of your whole sentence, including the stipulative portion, is no longer true (where I let t1 be the time of this utterance).
The stipulative parenthesis is wide-scoped for purposes of embedding, including in "is true".

So, we can translate A-sentences into B-sentences. Since the B-theory is simpler than the A-theory, and fits better with science and theology, we should all become B-theorists at t0 (where I let t0 be the time of this utterance).

Thursday, May 6, 2010

My amateur astronomy blog

Somebody asked me for photos of my travel scope, so I started off an amateur astronomy blog with them. I expected I'll be mostly posting observing lists there and my DIY projects.

Explaining the contingent with the necessary

Van Inwagen's argument against the Principle of Sufficient Reason requires the principle that:

  1. No necessary truth can explain a contingent one.
Note, though, that in ordinary language we feel quite free to violate (1) by saying things like:
  1. George couldn't divide the fourteen peanuts among the three party-goers because 14 is not divisible by 3.
  2. Arrow's Theorem explains why the United States' two-party system works better than Canada's electoral system.
  3. It takes a long time to find an optimal path between cities because the Traveling Salesman Problem is NP-hard.
  4. Patricia when told that she could have as much land as she could enclose with a four-mile-long string was wise to put the string in the shape of a circle because of the isoperimetric inequality.
  5. The distribution of error in the experiment was Gaussian because of the Central Limit Theorem.
Now, in all of these cases Van Inwagen might argue that there are omitted contingent aspects of the explanation. It is true that the explanation offered is in some way not complete. But it is, nonetheless, an explanation, and that seems to be enough to show that (1) as it stands is false.

Wednesday, May 5, 2010

The metaextrapolation fallacy

Let's suppose on Friday I am running a public star night. I am looking forward to this and checking the NOAA cloud cover forecast for Waco for that Friday night, starting the preceding Saturday. chart On Saturday, the forecast for that Friday is 75% cloud cover, let's suppose (I made up the data). On Sunday, the forecast is 70%. On Monday, the forecast for Friday is 62%. On Tuesday, Friday's forecasted cloud cover is 60%. Today, Wednesday, the forecast for Friday says 51%. What should I estimate Friday's cloud cover at? Well, a reasonable line of thought is that I can extrapolate from the data, and suppose that on Friday the forecast for Friday will say something like 40%, and since Friday's forecast for Friday will be the most accurate, my current estimate for Friday's cloud cover should be around 40%. Or I can give the appearance of being scientific and run a linear regression, and get the same answer from a pretty good regression with R2=0.95.

But this could be called the metaextrapolation fallacy. The weather data has already been pored over by experts—or by algorithms designed by experts. On Wednesday, these experts have access to the raw data behind the Sunday, Monday, Tuesday and Wednesday forecasts for Friday, and based on all that raw data, they extrapolated to 51% cloud cover. If they've done their job well—and of course that is always a question—my second-guessing extrapolation from their extrapolations should be trumped by their last extrapolation. In other words, I should suppose the cloud cover will be 51%, and that the trend I observed is just due to chance.

If, however, I find that in successive weeks there are similar trends in forecasts, I will then have reason to think that I've identified something in the data that they haven't. For if their estimates are the best possible, we would expect that the distribution of forecasts for a particular date, as one gets closer to the date, typically has no statistically significant trends. If there tend to be trends, then we might start forming hypotheses, like that longer term forecasts overestimate cloud cover. But to do that, we need more data than just these five points. (This post builds on, and qualifies, some of the remarks here.)

Tuesday, May 4, 2010

Good and evil: An asymmetry

According to St. Augustine, an evil is a privation of a due good. This makes for an asymmetry between goods and evils: there are due goods and non-due goods. For instance, the ability to run a mile in less than 3m 45s is a good, but it need not be a due good. Those who have that ability are well off for it, but, plausibly, while the vast majority who lack it are less well off for the lack, the lack of the ability is not a harm or bad to them. (Maybe this example doesn't work, because maybe before the Fall we could all run really, really fast or something like that. The example is only illustrative.) So, on Augustine's view, while there are due goods and undue goods, there is no similar distinction between the evils (or bads—to my knowledge Augustine does not distinguish the two), as all evils are opposed to due goods.

This asymmetry thesis for evils is plausible independently of the Augustinian view of evil as a privation. Among the goods, there is a distinction between due goods and what one might call supererogatory goods. There is no such distinction among evils. We might say that the distinction between bads and evils is such a distinction, but I am sceptical of that distinction when people make it. The good/evil distinction seems to basically line up with the distinction between non-moral evils and moral evils or maybe the distinction between minor evils and major evils. Besides, the distinction seems to me to be an accident of English. (For instance, the Latin malum covers both, as does the Hebrew ra` and the Polish zlo.) Moreover, even if there is a distinction, it does not cut the relevant way. A runner who can't run a mile in 3m 45m is not a bad runner.

Here is one minor reason the distinction matters. Tooley when trying to calculate the probability that a given evil has a justification assumes that an action is a priori as likely to have a wrongmaking property as a rightmaking one. However, if there are three basic axiological properties: supererogatory good, due good and evil, then plausibly there are three kinds of "-making" properties of reasons corresponding to these, two of them positive and one negative, so a priori an action is twice as likely to have one of the positive ones than the negative one in the Carnapian probability scheme he adopts. Basically the point is that "rightmaking" really should be "permissiblemaking" in Tooley's context, and there are two significantly distinct ways a property could be permissiblemaking: it could be dutymaking or merepermissiblemaking.

Monday, May 3, 2010

Horrendous evil and moral development

Suppose I end up at a concentration camp for a significant amount of time. Here are some possible moral development outcomes:

  1. becoming really bad (e.g., informer, kapo)
  2. becoming heroically good (e.g., Viktor Frankl, Maximilian Kolbe)
  3. becoming bitter
  4. remaining non-bitter
Obviously there can be overlap—you can become really bad and stay non-bitter. And they're not exhaustive because you might come in bitter. But nevermind all that. The above will be a classification of the most common cases of the most notable aspects of resultant moral development. (Thus, for the person who remains non-bitter but becomes really bad, we classify them just as becoming really bad, because that's more notable than remaining non-bitter.)

Now, I think we have no reason to think that A's outnumber B's. (It would be great to have empirical data.) Moreover, the average A is less morally bad than the average B is morally good. The reason is that there is an asymmetry here: the pressures that A's and B's are under in the camp decrease the culpability of typical A's but increase the praiseworthiness of typical B's. (There will be partial exceptions, like maybe the person who becomes needlessly cruel or the person who becomes virtuous because he's St Maximilian's cellmate. But even these exceptions are not going to be complete exceptions.) Furthermore—and again data would help—I suspect that some of the A's repent of their badness afterwards (some never do, and for some there is no afterwards), while few of the B's repent of their goodness afterwards. So, if all we know about x is that he is going to be an A or a B, the expected value of x's moral development hange will be positive.

What about the C's and D's? This is, I think, the really important case, as they'll probably be a larger group than the A's and B's. Now, it is a much greater virtue to remain non-bitter through a concentration camp than it is a vice to become bitter through the concentration camp. Part of the reason is the culpability point from the previous paragraph. Making one bitter is the "natural" tendency of horrors, and it is not a great vice to fall into that. So, unless there are way more C's than D's, we have positive expected value of moral development.

Now, add the following thesis: In terms of value, a moderate amount of positive moral development trumps a very large amount of suffering. (Socrates would say—and I think he'd be right—that any amount of positive moral development trumps any amount of suffering.) If this thesis is right, even when we add the amount of suffering, we may still have positive expected value for a random individual who suffers horrors like those of a concentration camp.

The real problem of evil, I think, is not about expected values, utilities and the like, however. The real problem is deontological. Does God have the right to allow someone to suffer so much given the expected value of moral development? I think the answer is positive. Suppose I knew that by preventing a great suffering to myself I would be losing an opportunity for significant positive moral development. Would prudence permit me to refrain from preventing the suffering? I think it would. Nor would such a refraining from prevention be morally wrong. But God is closer to me than I myself am, in some relevant sense. If I would not be imprudent or immoral to permit a suffering to myself, it would likewise not be wrong of God to permit it to me.

Sunday, May 2, 2010

The price of goods

If someone tells us that there is some great good to be achieved, we expect that either she's telling us something false or there is a string attached—a price to be paid. We are incredulous when it is suggested that there is a great good at no price. (Back when PalmOS was a going thing, I noticed more than once that raising the price of one of my apps would increase the volume of sales.)

Moreover, when a great good can be achieved at a price and at no price, we tend to think that it is more valuable when achieved at a price.

All this makes it implausible to suppose that one could have great goods, at least of the sort humans are capable of (and if it is suggested that God could have created other beings, capable of other goods, the response is: "For all we know, he did, too"), without heavy prices.

In other words, we should not expect a world with great goods and beings like us to lack evil. Moreover, we should not expect it to lack great evil.

Objection: The pessimistic intuition here is based on an inductive argument from the goods and evils of this world, while the space of possibilities is much broader.

Response: Could be. But either we do or we do not take the kinds of goods and evils we can observe to be representative of the goods and evils possible to beings like us. If we do take them as representative, the above remarks apply. And if we do not take them as representative, then a sceptical theist response to the problem of evil is very reasonable.

Saturday, May 1, 2010

A class of evils

As part of my interest in theodicy, I am interested in classifying evils. Here is a class I haven't seen discussed in the literature: major evils that are an overwhelming aggregation of small evils. What made me think of this was an email from a student who had suffered from a myriad of individually small things over the week, but where the cumulation is really large. In those cases, the whole appears bigger than the sum of the parts.