Monday, 13 February 2012

The Efficacy of the Evidential Argument from Evil

After Stephen Law’s debate with William Lane Craig there was a lot of discussion on the web about the efficacy of Law’s argument, the Evidential Problem of Evil, in countering Craig’s arguments.

Many of those who think the Evidential Problem of Evil ineffective argue that it is a probabilistic argument that, at best, shows that God is unlikely. (e.g. Chab123 2011) Others (e.g. Stuart 2011) felt that insufficient attempt was made to show what was wrong with Craig’s arguments for God, especially the Kalam Cosmological Argument.

Craig’s opposing arguments for God are deductive arguments.  If the premises are true and the logic valid the conclusion is conclusively established:  it is simply impossible for the conclusion to be false. 

The best a “probabilistic” argument can do is make it very, very unlikely that the conclusion is false.  And, as Sherlock Holmes pointed out to Dr. Watson, no matter how unlikely a conclusion seems if nothing else is possible then it must be true.

This is Chab123’s complaint. Law claimed that the Evidential Argument from Evil, when plugged into the Modal Ontological Argument, demonstrated the impossibility of God:

“His confusion on this point is breath-taking. His evidential argument from evil, at its very best, shows, at most, that it is probable that God does not exist. The probability is less than 1. To defeat the ontological argument with an argument from evil, his argument would have to entail that God does not exist. The probability that God does not exist would have to be 1. It would have to prove, as he says, that the conclusion of Craig’s argument is false. But Law’s own argument, as a matter of logic alone, cannot achieve this goal. It is a probabilistic argument. As such, it leaves open the possibility that God exists, even if the probability is quite low.”

Nothing probabilistic about modus tolens

Craig’s arguments make use of a rule of logic called modus ponens.  The rule is:

  1. If x then y
  2. x
  3. Thus y

The picture many people have of a “probabilistic” argument is along the lines of:

  1. The first swan seen is white
  2. The second swan seen is white
  3. The third swan seen is white
  4. The nth swan seen is white
  5. Thus all swans are white

It can seem that the Evidential Problem of Evil follows this pattern.  Just as the first white swan is not enough to establish that all swans are white the first bit of gratuitous suffering is not enough to cast doubt on God.  The seemingly endless succession of uniformly white swans seems to make it likely that all swans are white and the endless widespread pointless suffering seems to make it likely that there is no God.

In fact not all swans are white, despite how overwhelmingly improbable that seemed before the discovery of Australia.  “y” in the example above, cannot be wrong, “all swans are white” both can be, and is

The Evidential Argument from Evil, though, is not a “probabilistic” argument.  It is just as deductive as Craig’s arguments; making use of a rule of logic called modus tolens.

Modus tolens runs:

  1. If x then y
  2. Not y
  3. Thus not x

And the Evidential Problem of Evil runs like this:

  1. If God existed then there wouldn’t be so much pointless suffering in the world
  2. There is so much suffering in the world
  3. Thus God does not exist.
This is just as “logically airtight” as any of Craig’s arguments; if the premises are accepted then the conclusion must be accepted.    We can, of course, dispute the premises (I am, though, persuaded by Law’s “evil god hypothesis” (Law 2009) that rejection of them would be irrational).  As for the comparison arguments from Craig: though in the above example of modus ponens the conclusion was absolutely certain, the premises were just assumed to be true and the logic was kept simple to ensure its validity.  Craig's premises, and the validity of his logic, are both at risk: they are not at all certain and that uncertainty carries through to the conclusion.

The idea that the Evidential Argument from Evil is a member of some poor-relation-species of argument compared to Craig’s is erroneous.

But you haven’t refuted Craig’s arguments!

The complaint that Law did not adequately argue against the Kalam Cosmological Argument is equally erroneous.

When criticising an argument we could “undermine” it: dispute the premises or find fault with the logic.  Or we could show it to be false. Law certainly did not explain why the Kalam Cosmological Argument concluded falsely but he did give a good argument that it does conclude incorrectly.

If the Kalam Cosmological Argument is a sound deductive argument then no evidence will count against its conclusion.  That evidence does count against its conclusion is good evidence that it’s not a sound deductive argument!   

Similarly, with the Modal Ontological Argument.  This, very briefly, assumes the possibility of God and concludes with His existence.  If one accepts the logic of the argument (I don’t) then, far from showing “breath-taking” confusion Law’s claim is perfectly cogent:

  1. If God is possible then God exists (Modal Ontological Argument)
  2. God does not exist (Evidential Argument from Evil)
  3. Thus God is not possible

Of course we would like to know where the argument falls down, knowing where the argument falls down may expand our knowledge and that knowledge may help us avoid error in future arguments.  But look at this, readily criticised, argument:

  1. All elephants are pink
  2. Socrates is an elephant
  3. Socrates is mortal

The “logic” is hardly worthy of the name, the premises are junk but the conclusion is true.  A demonstration of the falsity of the premises or the invalidity of the logic does not establish that the conclusion is false; a demonstration of the falsity of the conclusion, obviously, does.

The Evidential Argument from Evil is a better counter argument than picking holes in Craig's logic and questioning his premises.  Whilst direct criticisms of Craig's arguments may cast doubt on those arguments the Evidential Problem of Evil shows them to be false.


Craig, W. L. (2010, February 4). Five Arguments for God. Retrieved February 8, 2012, from The Gospel Coalition:

chab123 (2011, October 24) The Missing Ontological Argument in the Craig vs. Law Debate. Retreived February 13, 2012, from Ratio Christi-at The Ohio State University:

Stuart (2011, October 20). How Stephen Law Failed in His Debate with William Lane Craig. Retrieved February 13, 2012, from THINKINGmatters:

Law (2009, November 30) Draft of PART of chpt 3. VSI Humanism. Retrieved February 13, 2012, from Stephen Law (blog):

No comments: