Saturday 28 December 2013

Null hypothesis

The null hypothesis (warning, the linked cartoon is satirical) is the default hypothesis of science.  Fundamentally, before considering any other hypothesis you should first prove false some formulation of the null hypothesis.

In earlier articles, like On Evidence, I’ve talked about how Christian apologists attack atheists for their presumptions.  As an example, the Shaved Chimp points out (negatively) the fact that “the starting point for the atheists is that there is no purpose out there”.

Well, yes.  That is pretty much what we all should start out with, that’s approaching a good null hypothesis.

Similarly, if you want to consider a god hypothesis you should first find some evidence to negate the relevant null hypothesis – that there is no god.

If you want to consider a claim like “God saved my cat”, you should first disprove the null hypothesis - that there was no supernatural agency involved in the on-going existence of my cat.  Once you've done that, you can try to zero in on the nature of the supernatural agency involved, given that it might have been the witches down the road rather than some sort of sky-fairy.

But until the null hypothesis is negated, there is simply no excuse for considering anything else.

A lot of atheists are, to coin a phrase, null hypotheticians.  They don’t progress past the impossibility to show that anything supernatural is required to explain the universe.  Such atheists don’t need to actively reject god because there’s no need to even consider the god hypothesis in the first place.  To establish such a need, they’d first have to reject the null hypothesis – and no-one has been able to that.

I doubt that the average theist has even tried.

Friday 13 December 2013

Planting a Special Plea for Warrant

Putting out a Warrant for Plantinga


If William Lane Craig were the Crown Prince of Christian Apologists then Alvin Plantinga would be one of his key advisors.  A couple of times now I have found that arguments coming out of WLC’s mouth had actually originated in the brain of Plantinga.  Given the damage that could be wrought by an apologist with some intelligence and formal training, I thought that Plantinga might be worth a closer look.


This led directly to me listening to a couple of longer presentations by Plantinga, for example an Unbelievable interview hosted by Justin Brierly in which Stephen Law also took part and a “debate” between Plantinga and Daniel Dennett.   This article draws quite heavily on the latter.


The engagement between Plantinga and Dennett was titled “Debate: Science and Religion: Are they compatible?” but it was not really a debate.  It was structured like this:


·         Plantinga (43 minutes) – defence of the proposal that science and religion are compatible, consisting in the main of a basic sketch of his Evolutionary Argument Against Naturalism together with a reference to the Irreducible Complexity argument of Michael Behe.


·         Dennett (35 minutes) – attack on the proposal that science and religion are compatible, basically by indicating that religion can only be considered compatible with science if one accepts gratuitous claims and that any compatibility that religion has with science exists only to the extent that the claims of religion are not testable


·         Plantinga (10 minutes) – response to Dennett


·         Question and Answer session (17 minutes)


Plantinga spent a lot of time going over his probability theory, a version of which I have already addressed elsewhere, but I want to focus on two things he said.  The first is this little chunk of Christian goodness:


…if theistic belief is true then it probably doesn’t require propositional evidence for its rational acceptability.  As I argued in this book, Warranted Christian Belief, if theistic belief is true, then very likely it has both rationality and warrant in the basic way; that is not on the basis of propositional evidence.  If theistic belief is true then very likely there is a cognitive structure something like John Calvin’s sensus divinitatus, an original warranted source of theistic belief.  This way, belief in God, like belief in other minds, has its own source of rationality and warrant and doesn’t rely on arguments from other sources for these estimable qualities.  The demise of the theological argument, if indeed evolution has compromised it, is perhaps little more of a threat to rational belief in God than the demise of the argument from analogy for other minds is to rational belief in other minds.


To put this in context, you might want to know what Plantinga means by “warrant”:


According to Plantinga, warrant involves four factors:


·         the mechanisms producing the belief must be functioning properly;


·         they must be operating in a cognitive environment for which they are suited;


·         they must be operating according to a design plan or aspects of a design plan aimed at arriving at the truth; and


·         the objective probability must be high that operating in this manner will reach the truth.




Plantinga has argued for a new definition of knowledge, being warranted true belief as opposed to justified true belief (although it could be argued that Plantinga is merely arguing that “warrant” is a better form of justification to those in traditional theories of justification).  I touched on “justified true belief” when discussing my world view (when I did so, I was describing an empirical variant of coherentism, not that I would necessarily consider myself an “empirical coherentist”).  Because Plantinga wants to replace standard forms of justification with warrant, we might be justified in looking more closely at what both mean.


KNOWLEDGE
Warranted True Belief
Justified True Belief
That which is known is believed
That which is known is believed
That which is known is true
That which is known is true
The belief in that which is known is warranted
The belief in that which is known is justified


Comparing warrant and justification:


Warrant
Justification
Belief is generated correctly
The believer is not deluded
Belief is generated in a suitable environment
The believer is using suitably collated data
Belief generation must be operating according to a “design plan” aimed at arriving at the truth
The believer is using an appropriate method for processing available data to reach a conclusion
Objective probability that the truth will be arrived at via the belief generation mechanism is high
If the believer is not deluded, is using suitably collated data and is using an appropriate method to process available data, then the probability of the believer arriving at the truth will be high


 Plantinga doesn’t really seem to be adding anything to justification, unless his use of the term “design plan” is meant to be literal (which he denies) or if his fourth point is intended to be more than a summation of the previous three.  He does, however, seem to be taking two things away.


It’s difficult to take Plantinga at his word when he writes:


it is perhaps possible that evolution (undirected by God or anyone else) has somehow furnished us with our design plans


and also refers to “belief-forming and belief-maintaining apparatus of powers … working the way it ought to work”.  He could mean this usage to be figurative, in the sense that a materialist atheist with cancer might say “my body isn’t dealing with this problem the way it ought to” meaning “certain cells in my body are reproducing without limitation, leading to the growth of tumors which are not recognised as alien and thus not removed by my body’s defences and this state of affairs will lead to my dying earlier than otherwise would be the case – an outcome which I consider to be suboptimal”.  However, I don’t think so.  I think Plantinga means for the four factors underlying warrant to be understood (by the right audience) approximately as follows:


·         Warranted true belief is generated using a design plan which is intended to arrive at the truth, if


·         the design plan was well crafted so as to lead to a high probability that truth will be arrived at, when


·         the belief is generated according to that design plan, and


·         the belief is generated in an environment which is appropriate for the design plan


If I am correct, Plantinga’s warrant can be summed up as “Warranted true belief is a belief that God expressly designed humans to have”.  Now, iff knowledge (which is defined as warranted true belief) is defined as God given, then the existence of God is a fundamental fact, a basic truth, because without God there is no knowledge.


Note that I used “iff” deliberately, rather than “if”.  The term “iff” means “if and only if”.


This is what we atheists call a theological argument.  Plantinga, on the other hand, calls this a “philosophical” argument.


Try reading through the extract from Plantinga’s presentation again.  I’ll help by converting the phrase “if theistic belief is true” to “if God exists” (see Plantinga’s definition of naturalism as discussed in Planting a Tiger for a justification).  The word “it” becomes “belief in God” in two instances.


…if God exists then belief in God probably doesn’t require propositional evidence for its rational acceptability.  As I argued in this book, Warranted Christian Belief, if God exists, then very likely belief in God has both rationality and warrant the basic way; that is not on the basis of propositional evidence.  If God exists then very likely there is a cognitive structure something like John Calvin’s sensus divinitatus, an original warranted source of belief in God.  This way, belief in God, like belief in other minds, has its own source of rationality and warrant and doesn’t rely on arguments from other sources for these estimable qualities.  The demise of the ontological argument, if indeed evolution has compromised it, is perhaps little more of a threat to rational belief in God than the demise of the argument from analogy for other minds is to rational belief in other minds.


Now some readers might divine from this that Plantinga is saying that you don’t need evidence to believe in God and that believing in God, even in the absence of evidence, would be “very likely” rational.  Some might just read it as saying that logical arguments (resulting in propositional evidence) don’t affect the existence or non-existence of God.  The latter is of course true.


I might have a very good argument that Kim Kardashian does not exist, for example, no-one I know has ever met her.  The brilliance of this argument, however, has no effect whatsoever on the existence or otherwise of Kim (or indeed any other possible form of Kardashian).


However, I suggest that Plantinga really means the former – the very weak claim that iff God exists, then believing in such a God would be rational even in the absence of evidence.  But he goes further.  Time for the second chunk of Christian goodness from that “debate”:


And even if contrary to fact, there were scientific evidence for unguided evolution and hence for atheism, that would by no means settle the issue.  Suppose there is scientific evidence against theism, it doesn’t follow that theism is false or that theists have a defeater for their beliefs, or that theistic belief is irrational, or in some other way problematic.  Perhaps there is also scientific evidence or otherwise for theism.  But second and more important, as I mentioned, if theism is true, it is likely that it has its own intrinsic and basic source of warrant …


Yes, ladies and gentlemen, Plantinga said, in a “debate” in which he was arguing that religion and science are compatible, that if there is scientific evidence against theism (noting that evidence against a hypothesis is the only scientific evidence worth having, since hypotheses only survive in their original form as long as there is no evidence against them), then the theist can just ignore that inconvenient scientific evidence and instead rely on the possibility that there might possibly be some evidence for theism or, alternatively, rely on a completely hypothetical sense (sensus divinitatis), for which there is absolutely no scientific evidence at all, as your basic source of “warrant”.  Remembering of course that a warranted true belief looks like this:


“I believe that God made me in such a way that I will believe that God exists, therefore God exists”.


This is, in context, an admission that religion is fully compatible with pseudoscience, but not actual science.


-----------------------------------


Just in case anyone is in any doubt, in science if you have evidence against your hypothesis (and by that I mean proper evidence, not false evidence), then your hypothesis is false.  End of story.


You can have evidence which does not support your hypothesis which might not necessarily be a hypothesis killer, but this is usually due to poor framing of the hypothesis or difficulty with obtaining experimental data.  Years of not finding evidence for the Higgs Boson did not disprove the hypothesis that it exists because no evidence against existence of the boson was obtained either.


If you only accept evidence which supports your hypothesis and you ignore any other sort of evidence, then you have pseudoscience.  That approach puts you in the same category as the 9/11 Truthers, Alien Abduction enthusiasts, Loch Ness Monster spotters, astrologers, mediums, water dowsers, shamans, homeopathy practitioners, homotoxicologists, quantum healers, moon landing hoaxers, Holocaust deniers, millenarianists, flat earthers, Nibiru cataclysmists, rumpologists and theists.


Many of the claims of a pseudoscience are not strictly testable but sometimes theists, like the Intelligent Design crowd, make outlandish claims and these can be shown to be false:


God must exist because the eye cannot have evolved naturally – shown to be false


God must exist because irreducible complexity is impossible via evolution – shown to be false


God must exist because evolution is false since it’s ridiculous to imagine that humans evolved from monkeys – shown to be false (the subordinate claim about human descent from monkeys being ridiculous is agreed but it’s irrelevant to the evolution argument since monkeys and humans have a common ancestor – the link isn’t actually proof, per se, but an explanation)


What science can do, and does, is chip away at all the “God must exist because ...” arguments.  Eventually, the theist is left with no undefeated claim for the necessity of God.

What the average theist can then do, and obviously does, is retreat into a form of special pleading (“my hypothesis is special and has God in it, therefore it doesn’t need evidence”).