Asking for _proof_ of global warming, or
_proof_ of the opposite, is alas not the
whole story.
There are lots of problems (including at
present the global warming question) where
the methods of proof we use in the physics
lab are not applicable. There are a number
of plausible outcomes, and no real proof that
any of them are impossible.
Situations like this are not the least bit
uncommon. Suppose you are a manager and you
want to hire a researcher or two. Can you
_prove_ that they are going to work out?
Invent something wonderful? Win a prize?
Suppose you are playing poker. You hold
three of a kind. Can you _prove_ the other
player doesn't have four of a kind?
Suppose you are a toy company. You need to
decide whether to produce a new toy. Can
you _prove_ it will be worth the cost of
development? It's going to cost you a fortune
just to produce a mockup suitable for testing.
Suppose you are contemplating starting a war.
Can you _prove_ it will work out as planned?
In situations like this, asking for proof is
the wrong question. A timely decision has to
be made. Insisting on absolute proof just
means that the default decision (inaction)
will be made, without regard to the merits.
Sometimes the default is good, sometimes it's
not.
For sure it's totally wrong to use a decision-
making process that is biassed one way or the
other. A bias toward inaction is as bad as
any other bias.
There is a reasonably sensible formalism for
decisionmaking under uncertainty. It is
somewhat alarming how non-widely known the
techniques are.
One technique of considerable value is called
scenario planning. There are entire books on
the subject, but to strip it down to the barest
caricature, the elements are:
a) write down all the plausible scenarios you
can think of;
b) estimate the likelihood of each scenario;
c) come up with procedures to cover all of the
reasonably-likely scenarios;
d) evaluate the costs and benefits of the
various outcomes.
You iterate on steps (c) and (d) to come up
with actions that you can take (now!) to
ensure that you will be in a good position,
no matter which of the scenarios comes to
pass.
I emphasize that you have to take action
now, before you know which scenario will
actually come to pass.
=============
There are sometimes things you can do to
reduce the uncertainty. You wouldn't want
to risk your life on one high-stakes gamble,
but if you get a chance to gamble a little
every day, with modest stakes and favorable
odds, you can do quite well.
The manager can hire several researchers.
The toy company can bring out several new
products and see which one(s) do well in
the marketplace.
Often, though, the risk-reduction comes at
a price, and eventually you have to just
make a decision under considerable remaining
uncertainty.
Things get very nonlinear when there is a
small chance of a terrible outcome. This
is particularly true in cases where you
can't "average out" the risk by running
many copies of the experiment (the way an
insurance company does).
The global warming question places extraordinary
burdens on the decisionmakers. The question
has several nasty features:
-- a not-provably-zero chance of a really
bad outcome
-- considerable uncertainty as to the
probability of each scenario, and as to
the cost of each scenario
-- uncertainty as to the effect of some of
the proposed actions
-- long timescales and truly global length-
scale, so that you can't repeat the
experiment or otherwise average out the
risk.
=================
As an almost-unrelated item, note that you
cannot logically answer the global-warming
question merely by saying that natural long-
term climate various are occurring. That
does not answer the question of whether human
activities are also affecting the climate,
for better or worse.
=================
Perhaps a good risk-reduction strategy would
be to have a multi-component portfolio
containing ocean-front property in both Tuvalu
and Manitoba. Depending on which way the
climate-change goes, one component or the
other is going to be non-existent.