Monday, 26 August 2013

Trinity college vs Muslim Nobels: "Spurious use of data"? The definitive version...

Richard Dawkins' tweet about the number of Nobel prizes won by Trinity College vs those won by the Muslim world sparked a kerfuffle with pro/con splitting about 50/50.
Many of the articles about the controversy quoted Faisal Islam, pretty much exactly as follows:
Channel 4 News Economics Editor Faisal Islam questioned Prof Dawkins' "spurious use of data".
I wondered about this.  What did Faisal Islam mean by "spurious data"?  So I Googled that phrase:
" Faisal Islam, spurious use of data ".
And what I got were pages and pages of pretty much the same story.  Bounce around, boing-boing, echo echo of the internet.  Same story, recycled. (I do the same, I know, but then I'm not the Independent, or Telegraph, or the Guardian...)
So, off to Channel 4 site... but nothing there in any of Islam's publications. Direct to Twitter then.  This gave these several tweets from @faisalislam:
Well, see what he says, not "spurious use of data", but "spurious data points."  Now it's clear.
What F-Islam means is the classic "fallacy of range", whereby you can prove just about any agenda you want to, depending on what beginning and end points of data you choose.  The most (in)famous perhaps is in climate science, where the use of different starting and end points in annual temperatures can prove either warming or cooling of the globe. (note: whatever you do with those data, the longterm warming trends are ineluctable)
But what of this case?  Faisal Islam claims Dawkins uses spurious data points.  But Dawkins data points are from the beginning of the Nobel prizes (1895) to today. What on earth is wrong with that?  Nothing I can see; perfectly logical and fair.
Instead F-Islam picks his own data points!  He picks the last two decades.  Why that?  Why not pick the last five years or the last 50?  His own choice is purely arbitrary, it strikes me, and not at all more logical data points than Dawkins' full-range data points.  Even then, F-Islam gets the numbers wrong: it's not 8 Muslim Nobels in the last two decades, it's 7.  Then again, Dawkins also made a slight error, in that the number of Peace Nobels by Muslims for all time, is 6 not 5....
Then F-Islam says that Dawkins' should "strip out" Economics Nobels, because they "aren't quite real".  Not "quite real"?? This is from the Economics Editor of Channel 4!  Still: strip out the Economics Nobels, even though they are labelled-- by the Nobel Committee --  "Economics Sciences".   Then surely we must also "strip out" the Peace and Literary Nobels as well.  For they are surely even less "real" and after all, the point Dawkins was making was about Science.
So, here it is, the full version spreadsheet.  Trinity College vs Muslim Nobels, for all time, and for the last two decades.  And with differing combinations of non-science prizes stripped out, for comparison:
Looking at this, it's clear that you have to really wrestle with the figures to make them come out favourably for the number of Muslim Nobel prizes.  Faisal Islam does so only by making his own -- arbitrary -- choice of "data points" and then by getting rid -- again arbitrarily, and oddly for an economist -- of prizes for Economic sciences.

Spurious moral equivalence: why pick on Islam?

The Dawkins Nobel kerfuffle also inspired an article by Murtaza Hussain in the AlJazeera website. He worries about Muslims being unfairly picked out. Why not make the Dawkins claim about other groups - Hindu, Blacks, Chinese -- who similarly don't have as many Nobels as Trinity College? Dawkins covers that very point here.  He says it's because (i) Muslims often talk about their huge and growing population and (ii) we are told that we should "respect" Islamic sciences.  No other group makes those two claims together[*]. Hence the focus on Islam and Muslims.
Ignoring Dawkins' logic for his tweet, Hussain then asks us to do a "simple test"
If you're ever unsure whether a statement about Muslims is bigoted, simply substitute the name of another minority community [Hindu/Black/Chinese] into the same sentence. If it sounds uncomfortable or even heinous to you upon doing so, rest assured that the original statement is probably just as malign.
This formulation appeals to many Islam apologists. "Aha!" they say, "yes, I see what you mean!  It's just as bigoted to make statements about Muslims as it was about Jews or blacks in the past."
But this formula doesn't hold water if you do a simple test on Hussain's own "simple test".
For example, a common statement about Muslims:
"Most of the terrorist acts in the world are done by Muslims".  
This is a true statement -- though some try to deny it... by changing data points!  Still, many Islam apologists find this statement "bigoted".  Do the Hussain "simple test" and you see that the opposite claim is nonsense: "Most terrorist acts in the world are done by Hindu/Black/Chinese".  "Bigoted" or not, more importantly, it's simply FALSE.
"Most Muslims want the countries in which they live to by ruled by Sharia law"
Many polls show this statement to be correct.  But substitute "Hindu/Black/Chinese" and the statement is incorrect.
One could go on; but won't.
Hussain's false moral equivalence doesn't wash, no matter how beguiling.
What's really needed, instead of claims of Muslim victimhood, is for Muslims to address the issue. Given the paucity of Scientific achievement in the Islamic world of the last 500 years, why is that?  Could the fact that boys are only taught to memorise one book, the Koran, and girls are kept at home, have something to do with it? Well, golly, I suspect it does.  As does the Islamic concept of Bid'ah, innovation, which is "haram" (forbidden) in Islam.  That might have something to do with it. Then there's the concept of the literal omnipotence of Allah, which means that you can't -- as a mere human --  you state scientific laws, since that would constrain the omnipotence of Allah.
(For more on this, see "The Closing of the Muslim Mind, How Intellectual Suicide Created the Islamist Crisis", by Robert Reilly)
In conclusion, to quote the wonderful, late, Christopher Hitchens:
What is needed from the supporters of this very confident faith is more self-criticism and less self-pity and self-righteousness.
[*] China also has a large population and less than stellar performance in Nobel prizes.  However, it does not make an issue of making the two points that are made by many Muslims: a large and growing population and the need to "respect" their science.  Quite the opposite: China tries to limit its population (and that's a whole other issue...) and it tries to improve its science performance by "studying from the west" and sending hundreds of thousands of its young abroad to study the latest in scientific discovery. In the years I lived in China, they constantly downplayed their science prowess, said they needed to study from the best in the world and to try to catch up.  That's in stark contrast to Islam, which asks us to "respect" contributions to science, which are now centuries in the past.
We should also recall that China had three "wasted decades", as it went through the ructions of the "Great Leap Forward" and the "Cultural Revolution", which pretty much put paid to any science from the 1950s to the early 1980s.
Even so, and using Faisal Islam's arbitrary last two decades of Nobels, China has twice as many, in the sciences, than does the Islamic world....