[Osx-nutters] Denial

[Chris Gehlker]

On Aug 14, 2007, at 9:09 AM, Stefano Mori wrote:

>
> On 2007-Aug-14, at 13:26, Chris Gehlker wrote:
>
>>>> It seems to me that Stefano's reasoning can be applied to any
>>>> area of
>>>> empirical enquiry. The fact that researchers  in any area reach
>>>> consensus is prima facie evidence that they are engaged in a
>>>> conspiracy to fake or misinterpret the evidence.
>>>
>>> They reached a consensus about their predictions of the future.  
>>> Right
>>> there it stopped making sense, and they stopped being scientists.
>>> Please see the latter part of my earlier post about empirical
>>> forecasting.
>>
>> So if there  is a  consensus  that a higher percentage of smokers
>> will get lung  cancer than will non-smokers the people who reached
>> the  consensus stop being scientists?
>
> When a doctor says that the treatment is 70% effective, what that
> means is that from experience, 70% of the thousands of patients
> already treated were cured. That's empirical.
>
> But if they maintain that in future it will remain 70% effective,
> then they are not being empirical. Bacteria develop resistance, so
> the figure for the future is not known. The drug may become
> completely useless. This is already been observed on a time scale of
> 10 years.
>
> Lung cancer has many contributing factors. You can't just predict
> that 25% of today's smokers will get cancer, because if they come to
> stop smoking before middle age (perhaps a new social trend) then 90%
> of their risk is eliminated. Meanwhile environmental pollution is a
> factor, and just to put that in perspective, out of lung cancers so
> far, 20% of the women were not smokers. Where is that 20% coming
> from, and why is it so large? And most people who smoke haven't
> gotten the disease. If a new social trend caused people to start
> smoking later in life, then the number of cancers due to smoking
> might be reduced considerably. The people who would have got cancer
> from the primary factor, smoking, may come to get lung cancer from a
> secondary factor, pollution, instead. Perhaps there is a genetic
> susceptibility and the people who get the disease tend to get it one
> way or another. It only takes another factor to come to the front of
> the queue and your future percentages will change.
>
> If by "will" you mean you are making a forecast about the future,
> then you need to empirically show why your forecast will work, and
> why other factors won't affect the outcome, otherwise it ain't
> scientific.
>
> Stefano

I congratulate you. You really are willing to get close to the Philip  
Morris position to maintain consistency. I find that perversely  
admirable.

You  may  be heartened to know that the mechanism by  which cigarette  
smoke causes lung cancer is less well understood than the mechanism  
by which greenhouse gasses cause global warming. The  latter  can,   
after all, be  demonstrated with a lamp, a shoe box, a thermometer  
and a canister of C02. Also the  scientists who shilled for the  
tobacco companies kept pointing out that there could  be some   
undiscovered factor, perhaps genetic, that predisposed some people to  
both lung cancer and nicotine addiction. This hypothesis was never  
refuted because it can't be.

The point is there was never anything like certainty that cigarets   
'cause' cancer. As you point out, may people who smoke do not get  
cancer and may who get cancer don't smoke. There was just an emerging  
consensus that the evidence was strong enough and the costs off  
doing  nothing high enough that various policies to reduce smoking  
were tried.

I think the same thing is starting to happen with GW.

--
No matter how far you have gone on the wrong road, turn back.
  -Turkish proverb