Research Funding (fwd)

Alexander Berezin berezin at MCMAIL.CIS.MCMASTER.CA
Mon Oct 2 19:09:17 EST 1995



---------- Forwarded message ----------
Date: Mon, 2 Oct 1995 19:57:22 -0400 (EDT)
From: Alexander Berezin <berezin at mcmail.CIS.McMaster.CA>
To: William Tivol <tivol at wadsworth.org>
Cc: biocan at net.bio.net
Subject: Re: Research Funding



On Mon, 2 Oct 1995, William Tivol wrote:

> Dear Alex,
> 
(BEREZIN)
> > So NIL rate of
> > about 1/3 is the "natural" optimum for Darwinian evolution
> > of the suppressive system.

(TIVOL) 
> 	I'd have guessed 1/e. :-)

(BEREZIN) 
> > "Funding ideas has merit". Yes, but only in theory, not in 
> > any (conceivable) practice. As Donald Forsdyke agrues at few
> > places, to submit truly innovative idea for funding will be
> > a professional suicide. The logic of ANY collective decision
> > process is ALWAYS conformistic. This is pretty independent of 
> > the personal qualities of people involved. Correspondingly, 
> > system which does indeed foster innavation and risk-taking
> > must function INDEPENDENTLY of the quality ("expertise level",
> > etc) of people running it and the ONLY viable strategy here 
> > is some sort of bet hedging algorythm which essentialy means 
> > sliding scale without cut off threshold. 

(TIVOL) 
> 	I was recently at my undergraduate alma mater (I'm involved 
> in their alumni fund), and one point the provost made is that
> seed money--derived in
> part from the alumni fund--for innovative research is essential 
> since there are many worthwhile projects which would have no 
> chance for ordinary grant funding. This supports your argument 
> for the present USA system. Is it really always true that no 
> system can be set up to fund innovative proposals,
> or is it merely true that such a system would never be adhered
> to by the present-day grantsmanship mafia?

(BEREZIN)
You unlikely can suggest a system which will be VERY effective 
in fostering innovation. But you can EASILY design a system which
will be REASONABLY risk-taking-friendly + much more cost 
effective than the present NSF/NIH/NSERC system. So, what's
the matter ?

The problem is that the second part of your statement is 
profoundly true : ANY such system goes against grantsmanship
mafia interests and correspondingly unlikely to be adopted from 
within the scientific community, UNLESS enforced externally by 
politicians and/or money-contolling bodies. This is, of course, 
a tough job to get through. 
 
The key (and this is what grantsmanship club fiercly denies)
is that (almost all) research is NOT very expensive. Having
basic funding, even just few thousands dollars per year, 
professor with "good ideas" retains SIGNIFICANT flexibility,
can look for various options to cooperate with others, etc.
Without ANY research funding you are (almost unexceptionally)
marginalized to a pariah and your training and potential is
rended to unefficiency. Yes, there are some lucky exceptions to
it (e.g. people having some independent income, or those
who can squeeze some few thousands from their teaching
salaries, or those who are fully pen-and-paper [ not 
computer ! ] theoreticians, etc), but they are just 
these - exceptions.  

So, the most rational system from the point of view of
common sense will be to fund (almost) all people at SOME 
basic level subjected to only reasonable evidence of 
continuing research productivity (peer review papers,
citations, etc). Funds on top of this can be awarded on 
competitively selective basis. 

This system would be less stressful, more risk encouraging
and MUCH more cost effective that the present one. 
However, this is precisely against interests of grantsmanship
mafia which insists that ALL funding be on compatitive
(selective) basis. You can easily grasp that 
even $ 1,000 per year (token) grant which is awarded 
on NON-COMPETITIVE to any prof. who, say, publishes one 
(peer-reviewed) paper per year will lead to DISASTEROUS 
consequences to the power controlling establishement,
whose monopoly start to erode instantly.    

Correspondingly, common sense DOES NOT work here and 
we have (essentially, irrational) selectivity system as
practiced by, e.g., Canadian NSERC. As I have 
explained earlier, American NSF is somewhat better 
(not much) due to its multiple-grant structure; however 
the fact that it is also a "project funding" (rather
than RESEARCHERS' funding) system, remians its major 
conceptual defect (unhealable under the present
circumstances).   

 
> > > 
> > > : However, such reform will ALSO reduce the power base of the
> > > : grantsmanship elite.

(TIVOL) 
> > > 	One of its greatest benefits.

(BEREZIN) 
> > Yes, but the key point is: FOR WHOM ?

(TIVOL) 
> Reducing the power base of the grantsmanship elite is of benefit for
> the rest of us--and, BTW, for the benefit of the progress of 
> science as a whole.  I am in complete agreement with you that 
> the present system fosters a few large labs with (in the USA) many 
> grants to the detriment of researchers with better ideas.

(BEREZIN)
Another falllacy here is the paranoic obsession with so
called "research excellence" - largely a pseudo-term,
which unfortunately politically is rather sellable 
gimmik (e.g. network of "Centers of Excellence"
in Canadian universities). 

(TIVOL) 
> If you are correct that the collective decision-making
> process will always be too conservative (or even that it has a 
> high probability of getting that way) that is a convincing 
> argument against making the idea important to funding. If good 
> ideas will not be funded, it makes no sense to fund ideas.  

(BEREZIN)
You can't fund "ideas", you can only fund people who 
carry them (or in some cases specific technological 
megaprojects like Appolo - but this is out of our scope 
now).  

> I'm not quite convinced that funding researchers will not evolve
> into a "who you know" system.  There would have to be an automatic
> algorithm for evaluating track records (with some adjustments 
> for special circumstances) otherwise I think the elite will 
> still get $$$ and the rest will get little or nothing.

(BEREZIN)
Having funding caps, public openness of ALL pertinent
records and AT-ARM-LENGTH funding personel (yes, bureaucracy,
few are OK !), it is not that difficult to implement. Actually,
Rosum Roy of Penn State suggested an algorythm. Of course, 
"who knows whom" can't be completely avoided, but it can be 
done a rather minor factor, something like [ unavoidable ]
harm of shoplifting to commerce. 

> 				Yours,
> 				Bill Tivol
> 




More information about the Bioforum mailing list