The Instigator
Thrasymachus
Pro (for)
Winning
2 Points
The Contender
modivarch
Con (against)
Losing
0 Points

The argument from evil is successful

Do you like this debate?NoYes+1
Add this debate to Google Add this debate to Delicious Add this debate to FaceBook Add this debate to Digg  
Post Voting Period
The voting period for this debate has ended.
after 1 vote the winner is...
Thrasymachus
Voting Style: Open Point System: 7 Point
Started: 8/4/2011 Category: Philosophy
Updated: 5 years ago Status: Post Voting Period
Viewed: 2,123 times Debate No: 17768
Debate Rounds (5)
Comments (17)
Votes (1)

 

Thrasymachus

Pro

I call an argument 'successful' if, after considering the argument, one should raise one's credence of the conclusion of that argument. I wish to argue that there is a successful argument from evil: that, after considering the experience of evil in our world, we should hold God's existence as less likely than we did before.

R1 accept, No new arguments in R5, references can be included in comment thread. Feel free to ask about concerns or clarification in comments, email, etc.

Enjoy life,

Thrasymachus
modivarch

Con

Excellent, thanks for the topic. Let's have some fun and attempt to improve our understanding.

I did note one comment about the lack of definition for god. I will assume that we are referring to the classical definition of the "omni-god" including, most importantly, omniscience, omnipotence and omnibenevolence. If any changes are desired feel free, but this seems right for our present purposes. Furthermore, I will attempt to refute the opposition by holding to this definition. In other words, I will not reject any of the three omni's in order to subvert the argument. Such a tactic would only serve to make this exercise fruitless.

Let's have a good discussion.
Debate Round No. 1
Thrasymachus

Pro

The argument from evil

Here is a modern-ish version of the argument from evil, although variations on this theme go back at least to (Rowe, 1979)

1) If God exists, gratuitous evil does not exist.
2) Gratuitous evil exists
C) God does not exist.

By gratuitous evil, one means that an evil without justification. One take on gratuitous evil is "an evil that can be prevented without losing some greater good nor permitting something equally bad or worse" (Rowe, 1979). There are many other, subtly different takes on gratuity, but the argument does not turn on which of these subtly different forms is, in fact, correct.

Most agree with premise 1 (the 'theological' premise). If God is morally perfect, then (at least) he would not permit evils without moral reason. As Daniel and Frances Howard-Snyder put it: "[I]f god can get what he wants without permitting some horror (or something similarly bad), why on earth would he permit it?" (Howard-Snyder and Howard-Snyder, 1999). Given most people accept P1 (or something so close to P1 as to have no effect on the argument), I will not argue for it further unless my opponent attacks it.

The factual premise (that there really are gratuitous evils) is controversial. There are obviously lots of evils, and most agree (pace theodicy) that many of evils appear gratuitous: we cannot see how these evils could be justified. Call these apparently gratuitous evils, inscrutable evils.

But inscrutable evil is not gratuitous evil: appearances can be deceiving. Here are three reasons why inscrutable evils make gratuitous evil likely.


1: Credulity

It is reasonable to believe "probably X" from "it appears that X": in the absence of defeaters, we should treat our appearances credulously. So, as Rowe urges (see Rowe, 1979, 1996), if it appears that an evil has no justifying reason, then we should believe there probably is no justifying reason. If an evil is inscrutable, it is probably gratuitous. If we do not see a justification, that is probably because there is not one there.

This 'Noseeum' argument has been extensively discussed and criticized (see Wykstra 1984, 2009, see also Bergmann, 2001, inter alia), and I do not wish to pre-empt my opponent. But I will say now (as the literature neglects it) that for us not to trust our appearances, it needs to be shown that our moral faculties are too poor to have access to whether an evil is gratuitous: it is not enough to show that our faculties are far inferior to someone else's (e.g. God's). My chess faculties are far worse than a Grandmaster's, but although I (unlike he) am unable to determine who is winning from a complicated mid-game scenario, I am still able to reliably determine whether a position is in mate, or that Black is winning if Black has all his pieces and White only a King and pawn.

I aver that our moral world is similar: some moral states of affairs are harder to apprehend than others, but our faculties are not so poor as to make it all opaque. In the more straightforward moral cases (like the 'hard cases' for evil: bambi, sue, the mutilation, etc.) determining an evil appears gratuitous is more like 'king and pawn versus all pieces' or 'check-mate' like than 'complicated midgame'. In such cases, we can trust our faculties have good enough access (even though there are others who may have even better access) to be confident in the 'noseeum' judgement. And, so long as we are confident in making that judgement in at least one case, the argument from evil can proceed.


2: Inscrutable evils are more likely to be observed if there is gratuitous evil

Call gratuitous evil G, and inscrutable evil I. It seems plain we are more likely to observe inscrutable evils in worlds that have gratuitous evils than those that do not: we are more likely to see appearances of X if X is actually there.{1} So:

P(I|G) > P(I|¬G)

We can plug this into Bayes:

P(G|I)/P(¬G|I) = P(I|G)/P(I|¬G) * P(G)/P(¬G)

As P(I|G) > P(I|¬G), our likelihood ratio is >1, and consequently our posterior odds are greater than our prior odds. This only shows that there is some confirmation - nothing about its magnitude. But for the resolution some (even minutely little) confirmation is enough. If there is confirmation for the key premise of the argument from evil, then we should adjust our credence in God's existence down in light of this argument from evil.


3: Lots of inscrutability gives us lots of confirmation{2}

How many inscrutable evils are there? A vast number: millions, billions, perhaps more. Every single one of these instances, taken alone, is evidence for gratuitous evil (by the first or second arguments). All of them taken together give extraordinary confirmation.

A worked example, after (Bass, 2011). {3}

For N instances of inscrutable evil, the Bayesian calculus becomes:

P(G|{I})/P(¬G|{I}) = P(In|G)/P(In|G)^N * P(G)/P(¬G)

Where {I} is the set of all inscrutable evils, In is the nth member of this set, and N the total number of members. Grant that we should be really sceptical about our moral faculties, so that any given instance of inscrutable evil is only marginally more likely on G than not-G: so little that our likelihood ratio for a single instance is 1.001. Also, set the prior odds as strongly favouring no gratuitous evil - 0.01, say. Finally, pretend there are only 10,000 inscrutable evils

The posterior odds are about 219, which translates into >99.5% likelihood that there is gratuitous evil given the 10,000 inscrutable evils. It is overwhelmingly likely there is gratuitous evil, even when our prior convictions were that there almost certainly was none, and when we take our faculties to be almost completely unreliable. Set N to one million (the same order of magnitude as those killed in the Holocaust), or ten million (the same order of magnitude as those killed by HIV) the posterior odds become too big for the average calculator, and these are still ludicrously conservative. A little confirmation, repeated a lot, gives a big effect.

So even if you are really sure that there is no gratuitous evil, and you think that the fact an evil is apparently gratuitous makes it only fractionally more likely it is actually gratuitous, you should nonetheless, in the light of all these instances of inscrutable evil, be very confident that there is gratuitous evil.


Conclusion

I have presented a fairly standard modern version of an argument from evil. I have taken the theological premise as uncontroversial, and given three reasons for the factual premise. If any of these reasons are cogent, then the resolution carries, as consideration of these evidences should raise ones credence in the likelihood of gratuitous evil, and so lower ones credence in Theism.

There are several fronts my opponent can attack from. He could take issue with the theological premise. He could attempt a theodicy to rebut my presumption of lots (or any) inscrutable evil - that although many evils appear inscrutable at first, further consideration shows they are justifiable. He could offer a sceptical theist challenge to undercut the evidence presented: that there is a stronger case that we have no warrant to move from inscrutable to gratuitous than my case above that we do. Or, of course, he could do something else entirely. I look forward to seeing what he has to say.

Thrasymachus


{1} This is argued at length in (Stephenson, 2009), but I take it to be fairly obvious, and so omit the more elaborate demonstration for space's sake. My opponent is welcome to challenge me here.

{2} This is strictly unnecessary for defending the resolution - any confirmation, no matter how slight, is good enough. However, I would like to show something stronger - not just that evil is disconfirmation of theism, but that evil is stonking good disconfirmation of theism.

{3} Bass uses a slightly different Bayesian gloss, but the results are similar, and the strategy of using many instances to drive an argument I owe to his work.

modivarch

Con

The argument before us has been presented above in the following manner:


1) If God exists, gratuitous evil does not exist.

2) Gratuitous evil exists

3) Therefore, God does not exist


My opponent (OP) accepts (1) and for the time being, though I have reservations, I will as well. The argument is valid which leaves us with premise (2) as the focal point. OP has already noted as much and, therefore, supports (2) with an argument along the following lines:


4) If X appears to exist then it is more reasonable to believe that X exists than not.

5) Gratuitous evils appear to exist.

6) Therefore, it is more reasonable to believe that gratuitous evils exist than not.


Now, what is correctly noted by OP is that this argument works only “in the absence of defeaters.”

What will a defeater look like? A defeater will show that one is reasonably justified in believing gratuitous evils do not exist despite the presence of inscrutable evils. Again, what is required is not to prove that gratuitous evils do not exist, but merely that it is reasonable to believe that they do not. (I also wish to point out that I do not need to prove that God exists.) In regard to my opponents opening definition of a successful argument this would mean that one is under no epistemic obligation to raise one’s credence. Whether the defeaters I offer will “do the job” will be our point of contention. In this first reply I offer one.


Building A Case:

The most prominent defense against this type of evidential argument has typically been the idea that an omni-God could have a justifiable reason for allowing evils despite our lacking the epistemic capabilities to know what that justification actually is. In other words, even though the evil is inscrutable to us and it is, according to OP’s argument, reasonable for us to believe it is gratuitous, it is, at least, equally probable that an omni-God has a justifiable reason for allowing that evil which is not presently accessible to us. In pre-empting this response OP offers a criticism:

But I will say now (as the literature neglects it) that for us not to trust our appearances, it needs to be shown that our moral faculties are too poor to have access to whether an evil is gratuitous: it is not enough to show that our faculties are far inferior to someone else's (e.g. God's).”

Most likely, the reason the literature has neglected this criticism is that in considering the cognitive abilities of humans we make certain assumptions. For example, we have limited resources in terms of our ability to know the past and future, we are limited in our ability to compute certain amounts of information, we are limited in our perceptual capabilities (i.e. their range and precision), etc. It seems plain that this is representative of the epistemic capabilities of humans, or lack thereof. Furthermore, in considering an omni-God we are also assuming a being whose epistemic situation is vastly superior and more capable to that of a human. The point, therefore, is not to simply say that God is superior, but that due to these implications it follows, quite reasonably, that an omni-God has access to, at the very least, information that we do not.

Let’s piggy-back the chess analogy and bring to light a few of our epistemic limitations in regard to moral situations. I offer two

1) Consider the “end-game” to be the “closing” of a given event. In other words, the point of time in which, after an inscrutable evil, a moral good could no longer be instantiated that would justify said event. If one assumes it is after the completion of a given inscrutable evil then he begs the question, for instance, against any type of eschatological theodicies. The point being that our epistemic limitations do not put is in position to determine whether or not an inscrutable evil is still “open” for justification.

Since we cannot show ourselves to be at “the end game” it is unreasonable to assume we have the necessary information required to judge whether or not a given evil is gratuitous and, therefore, opens the possibility that justification can be provided. Considering our epistemic limitations the assertion that we have an adequate moral capacity to recognize gratuitous evils seems to have failed. Let’s refer to this limitation as (VEG) for “vague end game.”


2) Furthermore, even given an inscrutable evil and a moral good it is highly questionable that a group of even the most intelligent of humans would have the cognitive faculties to understand the connection, if any, and its implications. For example, consider an inscrutable evil, of which we are aware, that occurred last Tuesday. Let’s assume that, in light of the defeater in question, this inscrutable evil led to a moral good which, unbeknownst to us, justified the inscrutable evil the following day. Furthermore, consider our proverbial “chessboard” which is the world and all the events that have occurred between Tuesday and Wednesday. Now, since we know an inscrutable evil and a moral good both occurred, what is the possibility that, given our epistemic limits, we understand the connection between the two without a full understanding of the “mid-game?” (Of course, there are specific reasons for which we could narrow the parameters. This does happen in certain instances and I’ll discuss them shortly.) Furthermore, what are the limitations on discovering a connection when these events are a week apart, a year, a hundred years, etc.?

The mere recognition of “the end game” will not provide us with the information needed to understand the causal chain between the inscrutable evil and the moral good. For instance, in the “Bambi” scenario it could be that a moral good occurred sometime later as a result of Bambi’s death for which we lack the cognitive abilities to see a connection. Again, in considering our epistemic limitations this strategy seems to have failed. Let’s call this epistemic limitation (NPC) for “no perceivable connection.”


These examples accurately show two difficulties surrounding our epistemic capabilities and, specifically, our capabilities in assessing various moral situations. Were one to reject either would require assessing human cognitive abilities to be above and beyond what is presently understood and accepted.


The Defeater:

It is clear that we lack the epistemic capabilities of an omni-God which will seriously limit our ability to identify connections between inscrutable evils and moral goods. Although, this defeater is strong enough to counter my OP’s argument one more addition will put it fully to rest.

It is quite obvious that we often encounter “evil” we know is justifiable due to its positive results. OP, being in the medical profession, will be quite familiar with the immense number of situations in which present suffering will lead to someone’s healthy recovery. In fact, there is an unquantifiable number of instances that have occurred in which the justification of a given evil event can be identified. Considering this, in conjunction with our epistemic limitations, the defeater for my opponent’s argument is this:


(1) We know of “lots” of instances where evil is justified.

(2) Given such epistemic limitations as (VEG) and (NPC) it is unlikely that we will find justification for each and every evil that has occurred.

(3) Given (1) and (2) it is at least as likely that a given evil has justification as that it does not.

(4) Therefore, we are under no epistemic obligation to raise our credence concerning the non-existence of God.


In short, the evidence we have from seeing various ‘evils’ justified and the lack of our epistemic capabilities (e.g. VEG and NPC) shows we have a powerful defeater for each instance of inscrutable evil and, at the very least, an equal amount of evidence for believing it is non-gratuitous than that it is. Thus, “lots” of inscrutable evils do nothing to raise our credence of gratuitous evils because we have “lots” of reasons to think they do not exist.

Debate Round No. 2
Thrasymachus

Pro

Above, I offered an argument from evil that should lead us to lower our credence in God existing. Modi opens a sceptical theist front: to use a sceptical defeater to undercut the evidential premise. Here, I shall exegete Modi's particular approach and show it to be unsuccessful. I will also remark on problems afflicting the entire sceptical theist approach, and why they are insurmountable.


Modi’s Magnificent Manoeuvre

Modi develops his sceptical attack in an interesting way (it seems a theoretical novelty, if you'll forgive another chess analogy). This is his defeater:

(m1) We know of “lots” of instances where evil is justified.

(m2) Given such epistemic limitations as (VEG) and (NPC) it is unlikely that we will find justification for each and every evil that has occurred.

(m3) Given (1) and (2) it is at least as likely that a given evil has justification as that it does not.

(m4) Therefore, we are under no epistemic obligation to raise our credence concerning the non-existence of God.

Note that his sceptical claim (m2) is very mild. In a bog-standard sceptical theistic response it would be far too weak: that our inscrutability sensor is not perfectly specific does not defeat the sum of inscrutable evil. MMM is not a bog-standard sceptical attack: Modi hopes to lower the sceptical bar by deploying information about our knowledge of justified evil, so that he very mild (and highly plausible) sceptical claim can do the job.

As it stands, (m3) simply does not follow from (m1) and (m2). Here's the reductio:

(r1) We know of "lots" of instances where a swan is black.

(r2) Given such epistemic limitations as (visual defects) and (lack of omnipresence), it is unlikely we will observe black swans for each and every occurrence of a black swan.

(r3) Given (1) and (2) it is at least as likely that a given swan is black as that it is not.

(r4) Therefore, we are under no obligation to raise our credence concerning the existence of non-black swans [when faced with apparently non-black swans]

The problems are that "lots" can still be a minority of the relevant cases (as with black swans) and that the sceptical claim is being amplified from "plausibly, we will sometimes H1 is false when H1 is true" to "we should not change our credence when we see outcomes that conflict with H1."

But there is a way to get MMM to follow with impeccable probabilistic credentials. What we want is this:

(m1') Most instances of evil appear justified.

If (m1') was true, then (m3) would follow. In a world where we have some epistemic limitations, a small 'tail' of inscrutable evils would be entirely unsurprising in a world where there was no gratuitous evil: our imperfect faculties would (correctly) see in most cases that evil was justified, but occasionally miss a justification here and there. And so (m4) would follow: our appearances are no more surprising in the case there are no gratuitous evil then the case there is some, and so these appearances give no confirmation of Atheism.

Sadly, Modi does not have a good case for (m1') and we can see it is false:

We aren't interested in cases of evil's being justified: rather, we are interested in cases of evils having God-justifying reasons. All sorts of unpleasant medical procedures (for example) are justified through casual limitation: there is no other way of getting the goods in question. God does not labour under these limits, and permission of horrors like Alzheimer's, cancer and HIV are inscrutable evils. Worse, I see no reason to believe even much more minor evils (coughs, ice cream melting) are justified God-justified: there are lots of ways God could arrange the world, and although I would be unsurprised if these minor evils are justified, I do not know there is a justification.

So most evils do not appear God-justified (indeed, there aren't even "lots" evils that are God justified): at best, instances of evil do not appear gratuitous but no justification is known; at worst, they are inscrutable. So we should reject (m1'), which is the only way I can see to reconstruct MMM into a valid argument. Therefore MMM is unsuccessful.[1]

But maybe Modi sold his sceptical concerns short, and he can actually support a stronger sceptical claim than (m2) - one strong enough to undercut the argument from evil.


Spanking strong skepticism

Just how strong does the sceptical claim need to be in 'vanilla' sceptical theism, so that inscrutable evils are no evidence whatsoever against the existence of God? It wouldn't be enough to say that are faculties are occasionally fallible, or even that our faculties are poor. Our epistemic limitations would need to be so bad appearances of gratuity and gratuity would be entirely uncorrelated, that we would identify gratuitous evil no better than chance. Only then would P(I|G)/P(I|¬G) be either unity or unknown, and so observation of inscrutable evil gives no disconfirmation whatsoever to theism (see argument 2 in my R1 above).

Consequently, Modi's cited sceptical limitations do not get us that far. It is true we lack good insight into the far future consequences of actions (VEG), or that evils and justifying goods can be linked up in ways we cannot always foresee (NPC). So there will be cases (perhaps many cases) where we observe inscrutable evils that are in fact non-gratuitous. Yet this not enough to say that our observation of inscrutable evils is no evidence for gratuitous evil.

Of course, Modi intended (VEG) and (NPC) as examples of our epistemic limits, to support his mild sceptical claim for MMM. If one wanted to support the far stronger one necessary in vanilla sceptical theism, one could offer more lots more limitations to gnaw away to nothing the evidence inscrutable evil gives (cf. Howard Snyder, 2009).

We can show that this program is failure, simply because the conclusion (our appearances of unjustified evil bear no relation to actual unjustified evil) entails a violation of all our moral intuitions.

Consider giving money to Oxfam. It appears that doing so will lead to recognisable goods (less death, less suffering), without any apparent harms. Yet, if sceptical theism is true, for all we know giving to Oxfam may deny great goods beyond our ken, and that not giving to Oxfam may realize great goods beyond our ken. Worse, our moral appearances are uncorrelated to moral actuality: however things appear, it is unknown whether giving money or not will really be the better option. There is no answer to whether we should give to Oxfam (or any other moral decision). This is moral paralysis, which seems crazy: if sceptical theism entails this, we should reject sceptical theism.

Sceptical theism needs to be selectively toxic to inscrutable evil, and not all moral appearances. This is difficult (see Sehon, 2008). Worse, the more elaborate skeptical theisms, which introduce resources to do just this become increasingly ad hoc and 'top heavy' (Rowe, 1984, see also Russell and Wykstra, 1988). Much like adding "all people who see white swans are crazy", to defend "all swans are black" these extensions of the Theistic hypothesis protect one from the evidence at the expense of making that hypothesis implausible. I aver that sceptical theism itself is implausible, and the forms elaborate enough to avoid general moral scepticism are really implausible.


Conclusion

Modi has developed a novel sceptical approach. Sadly, MMM is invalid, and can only be made valid with premises that are false. Modi has not provided enough to support a more standard sceptical attack, especially in light of criticisms I raise. The argument from evil remains cogent.

Thrasymachus



[1] If Modi could demonstrate (m1'), for example by a broadly successful theodicy, then he could use MMM to dismiss the remainder. Thus, if Modi ventures into theodicy, he can defeat the argument from evil by merely showing that the bulk of evils appear justified, rather than every evil appears justified.

modivarch

Con

Resting on a Fallacy:

In his first section (MMM) Thras’s attempts a reductio counter-example to show why the defeater I have offered fails. Unfortunately, it rests entirely upon a fairly obvious equivocation. His claim is this:

“The problem [is] that "lots" can still be a minority of the relevant cases (as with black swans)…”

However, this is not how I used the word in question (“lots”) by any means. It should have been quite apparent to Thras that in using this term I was drawing a correlation between the experiences we have of justified evil and the “lots” of inscrutable evil that he refers to in his opening argument. Furthermore, I explicitly stated, in the paragraph above the argument, that;

“…there is an unquantifiable number of instances that have occurred in which the justification of a given evil event can be identified.” (Emphasis added)

Possibly, if I would have added the emphasis earlier I wouldn’t have to bother with this, but as it stands I unfortunately do. Just to be clear the issue is that Thras’ counter-argument hinges on his “black swan” analogy. The essential point is that black swans, even though there may be “lots”, are still in the vast minority of total swans. In other words, consider a jar that holds 1k marbles. 100 of these marbles are red and 900 green. I could say there are “lots” of red marbles, but the reds would still only be 10% of the total and thus in the minority. So what Thras is implying is that my use of “lots” is akin to this, but that is clearly false and sadly, as a result, his straw-man reductio fails. I could say more on this point, but the fallacy is quite apparent and I will just leave it at that.


Moving On:

Thras further suggests that in order for my argument to go through I would have to support a modified premise:

(m1) Most instances of evil appear justified.

If most instances of evil appear justified then a few inscrutable evils, given some epistemic limitations, would be expected. Exactly what it meant by most here is a vague, as most could be anything from 50.1% of a given sample to 99.9%. However, given the explanation below this statement it seems that Thras is going for the latter claim, or, at least, something on the higher end. (I follow this interpretation due to a few statements including “a small ‘tail’ of inscrutable evils” and “occasionally miss a justification.” These statements seem to indicate that justified evils would be in the vast majority. If this is incorrect I apologize; I am merely trying to avoid committing the fallacy just discussed.)

The problem is that this claim is unjustified. Why is it that this new premise is required in order to support the defeater? Recall that Thras’ argument is only successful if, after considering it, we ought to raise our credence for the conclusion. If that is the case then there is no reason why justified evils must outweigh inscrutable evils. Here is why: recall that Thras provided a function above to support this conclusion:

“P(I|G) > P(I|¬G)

We can plug this into Bayes:

P(G|I)/P(¬G|I) = P(I|G)/P(I|¬G) * P(G)/P(¬G)

However, this function ignores any other evidence that might explain why certain evils, even “lots” of evils, are inscrutable - what I essentially provided. Let’s call the evidence of our epistemic limitations (E), the omni-God hypothesis (O) and the evils for which we find justifiction (J). (I realize that Thras’ reply requires more discussion on (E) and (J) - I will get to that shortly.) Including these evidences the function now looks more akin to this:

P(I|E & O & J & ~G) is at least as likely as P(I|G)

In short, the defeater we have at least an equal amount of evidence for inscrutable evil being justified as we do for it not being justified. Furthermore, notice that (J) and (I) = 1. For either an evil appears to be justified or it does not and if (~G) is at least as likely as (G) then both = .5 and it seems clear that any claims against this equivalence will require methods of quantification we simply do not have access to. (Another epistemic limitation of ours.) Any assertion that (~G) is more likely given (I), in light of all the evidence, will be a result of personal credence and have no bearing on the argument.

Of course, Thras claims that I need (m1) to support the defeater. However, under this new function, incorporating the total evidence, his claim is entirely without warrant. Up till this point, then, Thras has offered a rebuttal which relies on a fallacy and asserted a premise (m1) we have no reason to accept.

As I mentioned, however, Thras does object to both (E) and (J), but unfortunately both of these contentions fail as well.


Failed Objection to (E) & (J):

Thras argues that we are not interested in simply justifying reasons for evils (J), but rather God-justifying reasons. However, by identifying evils for which we find justification we have a plausible reason why, in conjunction with our epistemic limitations and the vastly superior capabilities of an omni-God, the evils for which we do not have justification, quite plausibly, have justification we are ignorant of. This justification may be simple or complex. A God-justifying reason may involve a connection between an evil and a good unperceivable to human cognition (NPC), but similar to the same justifications we see in everyday life. On the other hand, the justification could involve a much broader situation including, but not limited to, soul-making, free will, etc. Of course, the complexity would likely leaves us ignorant of the full justification.

Furthermore, the a priori nature of Thras’ response leaves us wanting. I understand that Thras is credulous towards the possible justification of any evil. However, his personal credence has no bearing on the argument in question or on our acceptance of it. For example, what justification are we given for the claim that “God does not labour under these limits?” It is quite plausible that God has certain limits under which he “labours” himself, again, in the goal of various eternal goods?[1] Could it not be that these limitations prevent him from curing everyone to our personal satisfaction? It is clear that Thras’ assertion needs better justification than his personal feelings on that matter.

In regard to (E) Thras’ mis-states my claim:

“…there will be cases (perhaps many cases) where we observe inscrutable evils that are in fact non-gratuitous. Yet this not enough to say that our observation of inscrutable evils is no evidence for gratuitous evil.”

I made no such claim that inscrutable evil is not evidence for gratuitous evil. In fact, the claim is that justified evil provides evidence, in light of inscrutable evil, against gratuitous evil. Of course, inscrutable evil is evidence for gratuitous evil, but it is not the only evidence we have. When assessing an argument one must look at the total evidence! Straw men and ignoring relevant evidences do not add up to successful arguments.

Freedom From Moral Paralysis

Unfortunately, I have left myself little space to deal with Thras’ last contention—that skeptical theism leads to moral skepticism. A quick summar: if we claim that we are unaware of certain goods that God might bring about by allowing various evils then we are left paralyzed between behaving morally and immorally. For if it is possible that God could only bring about a given good through a certain evil we do not want to act in a way so as to prevent that by stopping the given evil.

However, in light of a free-will type defense I think this argument ultimately fails. If at least one of the eternal goods that God desires is our free will then it is perfectly consistent that he will allow either the good or evil to happen which allows humans the free choice to make the correct moral decision.

This seems quite plausible, but more is likely needed and if Thras’ desires I will address it in the next round.


[1] For example, see Plantinga’s famous FWD in God, Freedom and Evil.

Debate Round No. 3
Thrasymachus

Pro

To defeat the argument from evil, my opponent has sought to deploy both epistemic limitations and knowledge of justified evil in Modi’s Magnificent Manoeuvre (MMM). Our disagreement over MMM is twofold: 1) exactly what premises would be required to make MMM a valid defeater, and 2) whether those premises are true.

Here, I’ll show that my formulation (not Modi’s) of MMM is valid, and then show the premises have no support.

Lots (if not “lots”) of confusion about “lots”

Modi argues my prior reductio of MMM rests on “a fairly obvious equivocation” about “lots”. Modi notes that the fact black swans are a minority is what gives the reductio its intuitive bite. Yet my reductio does not require minority “lots”:

R1) We know of an unquantifiable number of instances where a swan is white

R2) Given such epistemic limitations as (visual defects) and (lack of omnipresence) it is unlikely we will observe white swans for every instance of a white swan.

R3) Given (R1) and (R2) it is at least as likely any given swan is white than it is not.

R4) Therefore, we are under no obligation to raise our credence concerning the existence of non-white swans [when faced with apparently non-white swans]

So it doesn’t matter whether it is minority “lots” or majority “lots”. (R3) still doesn’t follow: granting (R1) and (R2) does not mean we should say (for example) a swan which appears black under ideal (or even middling) viewing conditions is, actually, just as likely to be white.[1] If I did misread Modi and his use of “lots” excluded minority “lots”, then my apologies. However, MMM is still subject to lethal reductio in non-minority “lots” too.

Bayes

Modi recasts his argument in Bayesian form:

P(I|E & O & J & ~G) is at least as likely as P(I|G)

Where O is omni-god, J is justified evils, and E is epistemic limitations. Modi asserts once we add O, J, and E, we lose the apparent confirmation I provides for G.

He also adds this passage:

[N]otice that (J) and (I) = 1. For either an evil appears to be justified or it does not and if (~G) is at least as likely as (G) then both = .5 and it seems clear that any claims against this equivalence will require methods of quantification we simply do not have access to.

This is false. (J) and (I) are not compliments. Although a given evil will appear justified or not, (J) is meant to refer to other evils we know to be justified. If (J) really refers to the same evil, (so I and J are compliments) then Modi’s Bayesian gloss is uninformative, as P(I|J&anything else) = 0. (It also doesn’t seem to follow for P(~G) = P(G) that P(J) = P(I), and besides, these priors have nothing to do with the likelihood ratio).

Regardless, Modi’s argument is invalid – E & J & O are not sufficient to screen off confirmation. As the reductio above shows, if it was sufficient to point to some epistemic limitations and to some occurrences of A to defeat appearances of ¬A, we would end up ignoring lots of relevant evidence.

The way MMM can work is my suggestion last round: something like (M1’), that most of our appearances are of justified evil.[2] For if we can combine that with a (fairly modest) sceptical claim, we do screen off the confirmation of I for G. We need is something like this:

Modi’s Bayesian Assault (MBA): P(I|~G&J&E) is at least as likely as P(I|G&J&E)[3]

Where J & E entail the incidence of inscrutable evils, compared to justified evils, is unsurprising given the limited specificity of our faculties to see justification.[4]

If Modi thinks he can construct another (valid) route the argument can take, he is welcome to present it. He has failed to do so yet.

“lots” of justified evils?

I’ve argued above that the sort of justified evils we are interested in are God-justified evils – evils for which we see there is a God justifying reason, and so whatever evils humans perform that are justified are the wrong reference class. Modi replies:

“However, by identifying evils for which we find justification we have a plausible reason why, in conjunction with our epistemic limitations and the vastly superior capabilities of an omni-God, the evils for which we do not have justification, quite plausibly, have justification we are ignorant of.”

If Modi means “by identifying evils for which we find God- justifying reasons”, then I agree – doing that helps support MBA. But, as noted before, pointing to instances where humans perform evils justifiably does not help this case, no more so than pointing to instances where humans perform evils unjustifiably helps mine. The reason is that justification for evils we perform almost always relies on our limited casual powers, something God doesn’t have (Modi notes that God may labour under some restrictions, but as casual impotence is not one, his response doesn’t address my criticism.)

I also said that we do not have evidence that any evil is God justified, and that some (the inscrutable ones) appear to be God unjustified. Modi hands this off as “well, that’s just your opinion, why should anyone care?” Yet it isn’t just my opinion, it is the opinion of virtually every scholar who has written on the problem of evil. If it wasn’t the case that evils appeared (at least prima facie) unjustified, then there would be no demand for theodicy (ditto sceptical theism).

It could be that there are various eternal goods that ‘tie God’s hands’ in such a way that he can’t (for example) cure everyone with cancer. But could be isn’t plausibly, and given (at least prima facie) it appears that god could cure cancer without forestalling any eternal goods, Modi needs to make this case if he wants to include cancer as one of his God justified evils. As it stands, Modi has not shown that any evils are God justified, and, given it doesn’t appear any evils are God justified, we have no support for J.

Moral paralysis

I think Modi mistakes the moral paralysis concern. It isn’t “maybe we should behave immorally because immoral behaviour might bring justifying goods”, but rather “we have no idea what is the moral or immoral course of action to take, because we cannot trust our moral appearances – for all anyone knows the apparently ‘moral’ course of action will deny great goods beyond our ken, and the apparently ‘immoral’ action will bring great goods beyond our ken.”

So I don’t think free will helps here. Even if God has a good reason to allow us to choose wrongly, that does not help us become unparalysed: we want to choose rightly, yet have no way to find out what the right action is.

Conclusion

MBA is the way MMM could be a valid defeater for the argument from evil. It relies on justified evils (J) and epistemic limitations (E). Modi has failed to support J, and thus the only chance MBA has is very severe epistemic limitations. However, such limitations entail moral paralysis, and should be rejected. As such, MBA is a failure, and the argument from evil remains undefeated.



[1] You could read (R3) as saying “if we take a swan at random, it is at least as likely as not white”. This would follow directly from (R1). But it would be straightforwardly silly to suppose (R4) would follow from that, so I presume this is not the reading my opponent has in mind.

[2] The value for ‘most’ depends on the degree of epistemic limitation. What is needed is E&J to show that our incidence of I is what we’d expect given the limited specificity of our faculties. For example: if we could be convinced that our ‘justification sensor’ was 90% specific, and we observed a ratio of 10J:1I, then that is not surprising on the hypothesis that there is no gratuitous evil. Progressively weaker sceptical claims demand progressively stronger claims about the incidence of justified to inscrutable evil.

[3] I’ve dropped O, as if we’re granting (1) of the argument from evil, P(O|G) = 0.

[4] Sceptical theism is a ‘special case’ of MBA. When E is such that our specificity is low (<50% or so), that is sufficient to screen off confirmation regardless of J.

modivarch

Con


A Failed Reductio

Having removed the equivocation Thras offers a modified reductio for my defeater. Thras wants us to believe that our sense perception has something to do with our moral perception. His argument is that if something seems to be gratuitous then it probably is—in the same way that if a swan seems to be black it probably is. Even given our various perceptual short comings we shouldn’t turn our back on our perceptions. Thus, for his analogy-dependent reductio to work we have to believe that there is a strong enough comparison between our sense perceptions and moral perceptions. However, this is something we simply should not do as the “metaphysical baggage” surrounding the justification of a given evil bares little, if any, resemblance to a swan, or any other physical object.

Quite obviously there is an immense difference between the metaphysical nature of “evil justification” and spacio-temporal nature of swans. (That is, they exist in space and time.) Furthermore, our sense perceptions have direct access to that spacio-temporal existence—assuming we are within “perceptual range.” Being out of range is no matter because we fully accept that if one was in range with normally functioning sense perceptions (i.e. they are not blind) then the person would be fully capable of experiencing the swan. Now consider moral justification. Does it have a spacio-temporal existence? No, of course it doesn’t. Although, an event may be spacio-temporal that in no way implies that the moral justification of that event is spacio-temporal. Furthermore, consider an inscrutable evil which has occurred—can we consider that to be within our moral range? Not unless we assume that our faculties are capable of accessing the “metaphysical arena” in such a way as to reasonably assess the justification of moral events.

Of course, we reasonably believe ourselves to have access to the color of a swan, but it is far from certain as to whether we have sufficient access to something so metaphysically weighty as moral justification, especially in consideration of a omni-God. The point, once again, is that God would have full access to the metaphysical arena including the full knowledge of how an evil is justified. In fact, it seems plain that only a omni-God would have this access. Thus, even were we to have access to 99% of what God does it still leaves open the possibility that God knows of a reason we do not. (Of course, we don’t have anywhere near 99% of the abilities that a God would have.) In short, for Thras’ argument to hold he has to show that swans and moral justification are analogous in regard to our access of them via our sensory or moral faculties. He simply has failed to do so and that is why the analogy doesn’t work.


Bayes

In addressing a few issues I will begin by saying that (J) and (I) were definitely not meant to be compliments (they do not refer to the same evil). Admittedly, I should have been clearer on this point and used a different letter for (I). Recall that (J) is the totality of justified evils. In the quotation in question (I) was intended to represent the totality of seemingly unjustified evils—not a single instance, but this confuses us with the previous use of (I). Let’s consider the totally of inscrutable evils to then be (U).

Having made this adjustment the simple point here is that for all occurrences of evil .5 are justified and .5 are inscrutable. I make no bold assertions concerning the amount of justified evil (medical procedures, child discipline, etc.) over inscrutable evils or vice versa. It seems clear that were we to weigh one over the other would be question-begging and a bold claim on our abilities to access all the needed evidence, assess it, and quantify its implications. Furthermore, Thras’ criticism of my credence function continues to rely on his failed reductio which was already discussed above. In short, he claims that we have adequate moral access to conclude whether or not an evil is justified. I have provided evidence to think we do not. (Also, see his fn. 2. Why should we think that our “sensor” is 90% effective? Again, compared to an omni-God this would be plainly false.)


Five more points:

1) He continues to claim that in order for my argument to work most evils need to appear to be justified. This is simply false and I dealt with it more in depth in the previous round. Also, he writes;

"...pointing to instances where humans perform evils justifiably does not help this case, no more so than pointing to instances where humans perform evils unjustifiably helps mine."

This claim is false. If we know of instances of human justification in addition to the plausibility of an omni-God then it is further plausible that God has justification for evils we do not know.

2) there is no reason to remove (O) from the argument. It is simply a statement concerning the possibility of God existing and nothing more. Thras argues in a footnote that if (G) then (~O). Notice, however, that I did not use (O) in the side of the function with (G). (O) may be false with (G), but it is perfectly plausible with (~G). There is no reason that both sides of the function need to incorporate the same evidence.

3) in regard to his section on “lots of evils” Thras still misses the point of my claim. To repeat myself, again, it is that we experience justified evils (call them “human justified”), but we also know that we have serious epistemic limitations especially in regard to the metaphysical. Considering an omni-God, and God’s epistemic superiority, it is likely that he knows/has justification for evils we are not aware. Consider the epistemic difference between an infant and a grown-up. Are we to conclude that an infant has the needed epistemic access to assess and quantify the implications of various “evils”? (Specifically, for example, punishment from a parent) Of course not and neither should we assume that if a God exists we have the needed access to what justifies his actions either.

4) Thras does ask for a plausible explanation of why God might be causally limited. Although, I didn’t give a full explanation I did offer the well-known free-will defense. If God desires that we have free-will which includes making either destructive or constructive decisions then this would plausibly prevent God from coercing the world in such a way as to limit our freedom and thus contradict his desire that we have free-will. (This is also an example of a God-justifying reason.) Considering that Thras wants me to show why FW and cancer are preferable to ~FW and no cancer. I might ask why Thras doesn’t think that FW is good enough, but I doubt this will get us anywhere. Exactly how we would go about assessing the value of FW and decide if it is worth the existence of cancer is beyond me. However, full-on comparison isn’t needed. All that is required is a plausible reason why God would allow evil and FW looks to be it.

5) As for moral paralysis Thras’ misunderstood my argument. (I’m actually confused as to what my argument had to do with any of his criticisms.) As simply as possible this is the point: (1) God desires that we have FW and act morally within that context, (2) G gives us guidelines to live by, (3) In order to maintain that we have FW G must allow some evils to occur; thus, we know what is right and wrong and we have the FW to choose either. Thus, in a given moral situation we don’t have to guess if the evil results in a better outcome because it is our free choice of “the good” that God desires.

Conclusion

So my original defeater still works despite Thras’ claims. His claim against (J) is false via the dis-analogy and, consequently, his whole reductio fails. FW is a plausible reason why God allows evil, and FW also shows why moral paralysis is avoided. In short, through the course of our debate, I have reasonably shown that we have no reason to raise our credence of the non-existence of God due to the existence of evil and that is all that is needed.


Debate Round No. 4
Thrasymachus

Pro

Thanks to Modi for the spirited exchange. I’ll finish with a (suitably partisan) summary.


Modi’s Magnificent Maneouver – Massively Mistaken!

Modi’s original defeater:

(m1) We know of “lots” of instances where evil is justified.

(m2) Given such epistemic limitations as (VEG) and (NPC) it is unlikely that we will find justification for each and every evil that has occurred.

(m3) Given (1) and (2) it is at least as likely that a given evil has justification as that it does not.

(m4) Therefore, we are under no epistemic obligation to raise our credence concerning the non-existence of God.

My objection was a reductio – MMM-esque reasoning means we ignore evidence wrongly, like seeing swans:

(r1) We know of "lots" of instances where a swan is black.

(r2) Given such epistemic limitations as (visual defects) and (lack of omnipresence), it is unlikely we will observe black swans for each and every occurrence of a black swan.

(r3) Given (1) and (2) it is at least as likely that a given swan is black as that it is not.

(r4) Therefore, we are under no obligation to raise our credence concerning the existence of non-black swans [when faced with apparently non-black swans]

Modi initially objected that this relies on a minority ‘lots’ case, and that isn’t the ‘lots’ he meant. So I showed that the reductio doesn’t hinge on a minority ‘lots’ case, by switching to white swans. Modi now says the reductio doesn’t work because visual faculties (that see stuff) aren’t the same as moral faculties (that do stuff like apprehend justification). That’s true, but the reductio does not hinge on that either. We could concoct yet another reductio about moral justification (e.g. you see me punching a child, but I point out the many instances you have seen me do good, and your limited moral faculties).

There is a reason MMM is easy to reductio. The inference rule that makes (m3) follow from (m1) and (m2) is:

Modi’s rule: When there are lots of observations of X, and we know it is unlikely our faculties will see X for each and every occurrence of X, we should believe an appearance of not-X is at least as likely to be X as it is not

Modi’s rule is mad. All our faculties - moral, perceptual, whatever – are unlikely to see X for each and every occurrence of X (for pretty much any X). But that doesn’t mean apparently not-X is still likely X! When we see a white swan, there probably is a white swan, even though we’ve seen lots of non-white swans and we know our vision will not pick up swan non-whiteness in each and every case. When we see an unjustified evil, there probably is an unjustified evil even though (rather, even if) we’ve seen lots of justified evils, and we know our moral faculties will not pick up evil justification in each and every case.

There are valid arguments that motivate scepticism in our moral faculties. I’ve even revised MMM myself to make it valid. Modi has rejected these alternatives in favour of the moribund MMM. The reductios show MMM to be fallacious. As he has not presented a valid (let alone sound) defeater, he has provided no reason to believe the argument from evil is unsuccessful.


Bayes

Modi’s new Bayesian claim is that about half observed evils are justified, and half are inscrutable (“[F]or all occurrences of evil .5 are justified and .5 are inscrutable”). He seems to think that unless we can provide some evidence that there are more than half inscrutable, the argument from evil is not successful. But look:

P(U|E & O & J & ~G) is at least as likely as P(U|G)[1]

Where does the proportion of unjustified evils come in? Nowhere! In principle, a single case of inscrutable evil can be evidence against God – that there are as many members of the set J as set U doesn’t prove the likelihood above. It is orthogonal to the argument.

What would be on point is a demonstration of how E&J (and O, if you like) screen off this apparent confirmation. But Modi’s argument is a fallacy – so whatever evidence he thinks he has his premises doesn’t matter.

Again, there are valid Bayesian defeaters (and, again, I have offered them), but, again, Modi has rejected these. Too bad for him.


(Parenthetical remarks)

1) Modi has never argued why human-justified evils are evidence of God justified evils, asserting so instead:

R3: Thras argues that we are not interested in simply justifying reasons for evils (J), but rather God-justifying reasons. However, by identifying evils for which we find justification we have a plausible reason why, in conjunction with our epistemic limitations and the vastly superior capabilities of an omni-God, the evils … have justification we are ignorant of.

R4: If we know of instances of human justification in addition to the plausibility of an omni-God then it is further plausible that God has justification for evils we do not know.

[W]e experience justified evils (call them “human justified”), but we also know that we have serious epistemic limitations especially in regard to the metaphysical. Considering an omni-God, and God’s epistemic superiority, it is likely that he knows/has justification for evils we are not aware.

Besides calling the lack of argument, I also showed why it is wrong.

R3: [evils we perform] are justified through casual limitation: there is no other way of getting the goods in question. God does not labour under these limits…

R4: [J]ustification for evils we perform almost always relies on our limited casual powers, something God doesn’t have…

Modi asserts this is false; again, no argument why.

2) Modi, in R4, notes an Alston-esque analogy to buttress his sceptical claims:

Consider the epistemic difference between an infant and a grown-up. Are we to conclude that an infant has the needed epistemic access to assess and quantify the implications of various “evils”?... [N]either should we assume that if a God exists we have the needed access to what justifies his actions either.

I pre-empted this all the way back in R2:

[F]or us not to trust our appearances, it needs to be shown that our moral faculties are too poor to have access to whether an evil is gratuitous: it is not enough to show that our faculties are far inferior to someone else's (e.g. God's).

That are moral faculties are far inferior to God, if he exists, doesn’t show we lack access to whether evils are justified or not, in the same way that our chess faculties are far inferior to God, if he exists, doesn’t show we lack access to whether White is in mate.

3) Modi asserts that Free will is a plausible reason for stuff like cancer. Firstly note that plausible accounts are not strong enough to show an evil is probably God-justified, so these accounts do not show there is lots of justified evil.

Besides, cancer isn’t a necessary constituent of free will, so we can have both free will and no cancer. So free will isn’t even a plausible reason to permit cancer.[2]

4) None of Modi’s free will story answers the moral paralysis charge[3]: it might be ‘enough’ for God that we are freely willing to do the good, but good will is not enough. If want to know what actually does good, and we can’t know because we cannot trust appearances, then we’re still paralysed. (see Sehon 2011, pp. 76-79)


Conclusion

Modi has failed to provide a valid argument. He has also failed to support any of his premises. For these reasons his defeater is a failure. The argument from evil remains undefeated, and consequently unsuccessful. Vote pro.

Enjoy life,

Thrasymachus



[1] Switched I to U.

[2] Plantinga’s FWD (cited above by Modi) does not attempt to show free will is a plausible reason, but the weaker claim that it is epistemically possible there is no possible world with creaturely free will without evil.

[3] Modi is right: moral paralysis does not interact with MMM, as MMM a) uses a weaker sceptical claim, and b) is fallacious. I offered moral paralysis to rebut valid sceptical defeaters which Modi never presented. (All these parenthetical remarks are unnecessary, as I do not need to rebut the premises of a fallacious argument).

modivarch

Con

I find it interesting that Thras says my defeater is “easy to reductio” given that his first two accounts failed. The first hinged on an equivocation, he then modified it, but the second hinged on a dis-analogy. I’ve discussed both of these flaws in detail above. Thras doesn’t make a third attempt, although he alludes to the possibility of one. However, he maintains that it is still a successful reductio based on my faulty “rule of inference.” A few comments concerning this rule:

(1) It fails to make any mention of the shortcomings of our moral faculties. Thras needs to show why (VEG) and (NPC) are false or, at least, have little bearing on our recognition of “not-X”.

(2) It also fails to mention the hypothesis of an omni-God. Although, not included in my original defeater I have discussed this as relevant information multiple times in the debate.

(3) I have never claimed anything to the effect that, “apparently not-X is still likely X!” (This quote comes from the paragraph below the inference rule.)

The comment in (3) misrepresents my argument and is surprising considering that it does not follow the inference rule he just stated. (It is much stronger than the rule) I have only argued that it is at least as likely and that is all that is needed. If, due to the limitations of our moral faculties and the possibility of an omni-God, what is possibly a not-X is at least as likely an X then Thras’ original argument fails.

Following this Thras regresses back to the swan scenario and ignores the dis-analogy between it and the use of our moral faculties. In order to argue that we can trust our moral faculties enough to claim that some evils are probably gratuitous Thras needs to show that we have the proper justification in regard to our moral faculties not our sensory faculties. Simply because that have a word in common between them is not enough to draw an analogy. As I described above these two faculties deal with completely different realms (the physical and the metaphysical) and any comparison between has serious flaws. Thus, repeating the swan example over and over again fails to disprove my defeater or provide the necessary justification for trusting our moral faculties to the extent that we can conclude an omni-God would lack justification for allowing evil.

On Bayes:

Wow, we have really gotten off track here. Who is to blame? I don’t know. I made a mistake in choosing my variables, but I thought I fixed it as I clearly stated “in the quotation in question” (I) ought to have been (U), i.e. the totality of inscrutable evils. (U) was not to replace (I) in the Bayesian formula. Nowhere did I mention this.

I find it odd that Thras claims I need a premise that says “most evils are justified” while concurrently claiming that (1) human justification is useless we need “God-justification” and (2) only one inscrutable evil is enough. Given these two claims I’m not exactly sure how even this presumptuous claim (that most evils are justified) would be enough evidence.

Again, however, the claim that most evils are justified isn’t needed. Instead of a presumptuous claim like this I make the claim that (J) and (U) are equal. In other words, half of all evils are justified and half are inscrutable. That said in conjunction with the fact that we lack various faculties that put us at a great disadvantage, e.g. (VEG) and (NPC) and we also have the omni-God hypothesis—recall that the attributes of this God are all-good, all-powerful, and all-knowing—It is possible that an omni-God would have justification beyond our ken for allowing evil analogous to the way in which we have justification for allowing various evils.

Final Remarks:

(1) Thras says that showing our moral faculties to be inferior to God’s isn’t helpful. This is just false and I have presented a very clear argument for why—if God exists, given his attributes, he (or she) could very possibly, in the same way we do, have reasons for allowing evil that we do not. This is similar to the way in which a parent has reasons for doing things that their children don’t understand. This epistemic divide is easy for us to see and, furthermore, it is easy to see how this applies to the argument. Again, if God existed there would be a huge divide between his epistemic capabilities and ours. Thras needed to address this and he simply did not. Thras says that he “pre-empted” this, but whether he did or not is a moot point because he failed to adequately show why the parent-child/human-God analogy is insufficient evidence. Especially in conjunction with (VEG) and (NPC) this shows we are at a great disadvantage for assessing whether or not an evil is gratuitous. Specifically (VEG) argues that we don’t actually know when “white is in mate.” Thras never provided a single reason why we should believe we have seen white is in mate; i.e. that an evil event has “completed” in such a way that we know all the relevant information and contingencies in such a way as to accurately assess whether or not it is justified.

(2)He claims that God does not labor under the same causal limits that we do. This is obvious. Also, he seems to allude to the idea that I think God has the same causal limitations as we do—claiming that I deny R3 and R4. This is completely uncharitable and I’ve never said anything close to this. Rather I argued that if God desired to maintain that we have free will this would clearly limit his causal efficacy lest he act in such a way to override our free will.

(3)Thras still does not understand my response to his moral paralysis claim. Simply put, we are not paralyzed if God allows evil for the benefit of free will while maintaining a moral law which we are, at least for the most part, aware of. This would imply that a child being abused allows free will although we know, very clearly, that the act in and of itself is morally wrong and, if possible, we ought to (freely) prevent it. In other words, by preventing the abuse we are “actually doing good.” Again, no moral paralysis here.

Conclusion:

I have successfully shown that my argument isn’t fallacious, as Thras claims. Failed reductio’s are useless as counter-arguments. As a result the defeater provides a solid basis for us to think that inscrutable evils could very well be justified or, at the very least, that his argument is unsuccessful. However, this is not to say that the evil is justified, for instance, if a person was to abuse a child. Clearly, for the person abusing the child, this act is not justified. The point is to say that an omni-God might allow this in lieu of an over-arching justification, i.e. free-will. Of course, it’s quite possible that God doesn’t exist. However, the argument that (1) since some evils are inscrutable to us implies (2) that an omni-God probably doesn’t have justification and, therefore, doesn’t exist is just plain irrational. It heightens our epistemic situation well beyond its means and makes presumptions about our moral knowledge it simply cannot obtain.

In short, our moral knowledge is limited especially in regard to justification. If a benevolent God exists he or she would very likely have justification for allowing certain evils. That information alone is enough to say that we don’t have a clue whether or not an evil would be unjustified for an omni-God. That being the case inscrutable evils do not provide a good reason for thinking God doesn’t exist; i.e. Thras’ argument is unsuccessful.

Anyone else confused this statement in his post above:

“The argument from evil remains undefeated, and consequently unsuccessful.”

Eh?? Clearly, he wants us to vote con.

Just a playful jab J


Debate Round No. 5
17 comments have been posted on this debate. Showing 1 through 10 records.
Posted by modivarch 5 years ago
modivarch
Thanks for the debate.
Posted by Thrasymachus 5 years ago
Thrasymachus
R5 references:

Sehon, S. (2009). The problem of evil: skeptical theism leads to moral paralysis. International Journal for Philosophy of Religion, 67(2), 67-80. doi:10.1007/s11153-009-9213-1

(My inline citation for Sehon was wrong - should have been Sehon, 2009, not Sehon, 2011. Sorry!)

Thanks for the debate, Modi. Doubtless see you around.

Enjoy life,
Thrasymachus.
Posted by modivarch 5 years ago
modivarch
R4 Reference:

Swinburne, Richard. "Natural Evil" American Philosophical Quarterly, Vol 15 No.4, 1978. 295-301.
Posted by modivarch 5 years ago
modivarch
Late on the references, sorry...

For R3:

Maitzen, Stephen. "Skeptical Theism and Moral Skepticism" International Journal for Philosophy of Relig (2009) 65:93-103
Bergmann, Michael. "Skeptical Theism and Rowe's new evidential argument from evil." Nous, 35, 278-296
Posted by Thrasymachus 5 years ago
Thrasymachus
R3 references

No references for me this round

Enjoy life,

Thrasymachus
Posted by modivarch 5 years ago
modivarch
I'll post my R3 references soon.
Posted by Thrasymachus 5 years ago
Thrasymachus
Hello there,

Sorry, that was unclear. I should have said something like "appearances of gratuitous and actual instances of gratuitous evil would need to be entirely uncorrelated etc. etc." It seems to me we'd need something that strong to say our appearances are no evidence at all.

Best,

Thrasymachus
Posted by modivarch 5 years ago
modivarch
I think I know what you are pointing to here, but I have to admit that the phrase "appearances of gratuity and gratuity..." is a bit confusing. Is this a typo?

"Our epistemic limitations would need to be so bad appearances of gratuity and gratuity would be entirely uncorrelated, that we would identify gratuitous evil no better than chance."
Posted by Thrasymachus 5 years ago
Thrasymachus
R2 references

Howard-Snyder, D. (2009). Epistemic Humility, Arguments from Evil, and Moral Skepticism. In J. L. Kvanvig (Ed.), Oxford Studies in Philosophy of Religion (Vol. 2, pp. 17-57). Oxford: Oxford University Press.
Rowe, W. L. (1984). Evil and the theistic hypothesis: A response to Wykstra. International Journal for Philosophy of Religion, 16(2), 95-100. doi:10.1007/BF00136568
Russell, B., & Wykstra, S. J. (1988). The "Inductive" Argument From Evil: A Dialogue. Philosophical Topics, 16(2), 133-160.
Sehon, S. (2009). The problem of evil: skeptical theism leads to moral paralysis. International Journal for Philosophy of Religion, 67(2), 67-80. doi:10.1007/s11153-009-9213-1

With apologies for the slightly large text: it turns out 'paste from word' is a bit capricious.
Posted by popculturepooka 5 years ago
popculturepooka
So basically "since there is evil in the world, god doesn't exist?"

That wasn't his argument. All he's arguing is that if the AoE is successful it lowers the probability of God's existence. One could consistently hold that AoE is successful and but hold that, on balance, the probability of God's existence is very high.
1 votes has been placed for this debate.
Vote Placed by CD-Host 5 years ago
CD-Host
ThrasymachusmodivarchTied
Agreed with before the debate:-Vote Checkmark-0 points
Agreed with after the debate:--Vote Checkmark0 points
Who had better conduct:--Vote Checkmark1 point
Had better spelling and grammar:--Vote Checkmark1 point
Made more convincing arguments:--Vote Checkmark3 points
Used the most reliable sources:Vote Checkmark--2 points
Total points awarded:20 
Reasons for voting decision: Terrific debate, bravo to both of you. I thought the argument were solid all around. Both sides did a nice job hitting the ball back and forth across the net so I'll tie convincing. Pro made use of sources while Con did not.