----- Original Message -----
From: "John D. Giorgis" <[EMAIL PROTECTED]>
To: "Brin-L" <[EMAIL PROTECTED]>
Sent: Thursday, August 02, 2001 9:30 PM
Subject: Re: What is science.
> At 07:39 PM 7/28/01 -0500 Dan Minette wrote:
> >I thought about the easiest way to do this, and have
> >decided to start with considering the paradigm science: classical
> >mechanics/universal gravitation (CM/UG). It has the advantage of being
the
> >first real science, and containing pretty much all that one finds in
> >science.
> >
> >Without going into a great deal of history, let me just make the
following
> >observations about CM/UG.
> >
> >1) It assumes that the objects under study act according to laws of
nature,
> >not according to volition
>
> I don't think that this is valid, because to me, it excludes biologists,
> especially zoologists. Animals act according to volition, also, as has
> been demonstrated in numerous experiments, yet nobody has yet suggested
> that Biologists, Zoologists, and Animal Behaviorists are not scientists.
I guess that I should probably modify this point to make it a bit clearer.
The objects under study should act according to laws of nature, such that
any volition that may exist can be ignored by the scientist. Most of the
biologists I know are in microbiology or biochemistry. Very little
discussion of free will here. When larger scale behavior Animal
behaviorists are a bit problematic, for what its worth. There is evidence
of strong cultural biases in the work that has been done on animal behavior.
I think it is possible to do science in this area, treating the animals as
"black boxes" and tabling the any discussions of the wills of the animals.
But, if the consideration of the animals as individuals that make choices
cannot be tabled, then, IMHO, science cannot be done.
I don't think this tabling is possible for humans. Humans are political
animals. A single charismatic leader can effect vast changes in society.
Belief systems, for example, can be responsible for significant changes,
including economic changes. If one were to argue in 1700, one would argue
that serfdom and/or slavery was an absolute requirement for any complex
organized civilization. One would be able to cite scores of examples, with
no real counter examples given. Today's society would, if described, be
considered a violation of "laws".
> Additionally, there is no "why" for this criterion. UG/CM applies
> perfectly well to human beings, even when acting under our own volition.
> Just because I choose to jump out of a plane without a parachute, UG/CM
> still applies.
OK, but whether or not you choose to jump is not a matter for science. This
is quite a nit-pick there John. I'll try to be perfectly precise in my
language, but I don't think that giving me no quarter at all for lack of
perfect precision will get us anywhere.
>
> Additionally, although Economists spend a great deal of time thinking,
> there is a huge emphasis on data collection. I use introspection a great
> deal on this List, because the List format is more suited for it.
>
> Moreover, was UG/CM developed completely without introspection? Pure
> empiricism is nice once you have the Laws to test, but how are these Laws
> developed in the first place? Or was the development of UG/CM not part
of
> science?
>
> Finally, I think that you and I both agree that there are a lot of
> Economists who engage in a lot of empirical observation. So, there seems
> to be a large sub-sector of Economists that meet this requirement.
Oh, there is empirical observation. But all the data that go into economics
are not empirical. Its not that thought experiments are done as a means of
thinking about a particular model. Its that internal experiences are used
for data.
Look at the questions you asked: "what decision would you make?" That is an
invitation to reflection for primary data. The reason for this is the basic
component of economics, one individual, has data available about human
beings from introspection as well as from empirical observation. This is
not true for physics, chemistry, molecular biology, etc.
>
> >3) It is experimentally based.
>
> See above. The empiricists are also engaged in experiments.
>
Well, can you give examples of economic experiments? The ability to do
experiments is rather limited, as far as I can see. In particular, can you
give examples of experiments when one has control of the variables involved?
> >4) It makes broad predictions from a few basic principals
>
> In a way, all of Economics is essentially reducible to the Law of Supply
> and Demand.
>
> You can't get much more broad than that.
Lets see some real predictions from that law. Lets take something real
simple. What will
be the price of oil for each month from now till 2003?
>
> In fact, I found this category funny, because one of the great attractions
> of Economics to me, when I decided to switch my focus from Geology to
> Economics was how much of Economics could be derived from so little.
OK. Show it to me with some numerical precision. Not arm waving general
explanations, but precise predictions. As I said offline, the top companies
energy companies primarily use technical models (which are close to Tarot
cards as a tool) to predict future prices. This is not done in science.
>
> >5) It makes precise predictions
>
> Actually, Economics does make precise predictions when it has precise
>data. In computer simulations, the predictions are very precise. I don't
>think that you want to argue that Economics is not a science simply because
it
>is hard. ....in this case, just hard to get precise data. After all,
the
> lack of precise data has troubled many other areas of science since its
> inception.
>
Well, there are very precise data available. There are a zillion financial
numbers out there. I cannot see the lack of precision in the data being the
problem. Much better science was done with fewer numbers than are available
to economists.
I see the problem as not being able to separate causality from correlations;
the inability to come up with a model
> >6) It is reductionistic
>
> See #4.
>
> Indeed, Economics is famous for assessing large economic interactions by
> simulating the actions of an economy with only one or two people. In
>this case, individuals are our atoms.
Simulating aren't experiments. One can simulate all sorts of unreal stuff.
I'm
talking about real experiments where variables can be isolated enough to
actually do experiments. Look, I do simulations for a living, and I know
that they are a way to apply known theory make predictions. The Monte Carlo
models of photon and neutron transport that I work with are excellent for
this.
They can also be used as a means of thinking through a model. That's fair
enough. But they are not experiments.
>
> At any rate, #6 is almost identical to #4.
>
Not really. It is a hallmark of science going into an area, such as fluid
dynamics, where precise predictions are not immediately available but there
is still a solid scientific basis underlying the study. It is assumed that
fluid dynamics is reducible to statistical mechanics. Using that, one can
do a great deal.
> >7) It is synthetic
> >7) Great theories of physics put together different fields that are not
> >always clearly linked.
>
> Although this is nice, I don't think that this can be a *requirement* of
> science. After all , if it is reductionistic, then in theory, science
> could simply have made its discoveries from the bottom up. The fact that
> different fields were not clearly linked is simply an artificat of
>history, and would not necessarily be repeated in other worlds that
developed
> science.
>
It is not a requirement, remember this is a paradigm example. Think of
getting graduated points for each answer and having a fuzzy science/not a
science line. Actually, this is where Zimmy's arguments for hard science,
not so hard science, and not science has some validity.
> >8) It is falsifiable by additional experiments
> >Not very well. A good example of this is the argument over the factors
> >involved in the Great Depression.
>
> In part, that is because this is a single historical event, and we do not
> have perfect data from that historical period. Putative causes of
> depression (small "d"), can, however be falsified by additional
>experiements.
OK, what experiments have been done with control of all variables? Maybe I
missed something.
>
> Another great example is Amartya Sen, who took the theory that famines are
> random events produced by crop failures, and falsified it by examining the
> data.
>
Can you give this example, his data set, and how he proved it?
> >9) It survives as a special case of the newer theory
> Well, given that EM was only derived from QED in the last 50-75 years, I'm
> not going to worry myself too much that we haven't gotten that far yet.
>
> #9 fails as a requirement of science, because under this criterion, you
can
> never know if you are doing science until the field is very mature.
>
I was thinking about how new ideas replace old. Economics has had many more
paradigms in the last 50 years than physics has had in the last 400. That
is not indicative of sciences. If there has only been one well established
theory, then its not a problem. But, economics has had many paradigms, many
schools already. Physics has had three if you include the Aristotelian
Physica, two if you do not.
> Indeed, a great many of your requirements Dan, seem to me to be possibly
> linked to the fact that the Universe began as a single point in the Big
> Bang. Thus, it is not surprising that there are a few underlying
> principles to the entire Universe.
>
No hard feelings, but you don't know what you are talking about here.
Singularities do not necessitate simple physics.
> I wonder, however, how closely any of the other Sciences might be able to
> match these nine criterion.
>
I'll give an example of molecular biology fitting these criterion in a later
post.
> As I have been thinking about this discussion, it occurs to me that there
> is a profound difference between myself, and the deniers of Economics'
> status a science. This difference can be expressed through a corny
> analogy - I view science as a process, you view science as a destination.
Well, by your definition, anything can be in process. I've seen this argued
for all sorts of nonsense. Since nothing cannot be proven to not have
predictable values sometime in the future, then anything you want can be a
science.
> You seem to judge science based mostly on "post facto" criteria.
>
You mean shouldn't worry about whether economics does what science does well
at all? As long as people are trying with earnest efforts, then they are
what they want to be?
> This poses some strange difficulties for your definition, however.
>After all, was Newton a scientist before he developed CM/UG? It seems to
me
> silly to argue that he was not - but your entire presentation would leave
> us with that conclusion. Newton in the days before CM/UG would have had
>a hard time meeting most of the criteria on your list.
But, if things didn't work for 200 years, would he have been doing science?
If you cannot accomplish the tasks that science accomplishes, why call
yourself a scientist.
Many very good things are not science. Raising a child isn't a science.
Understanding other people is not a science. Writing music is not a science.
Politics is not a science.
Economics is definitely a worthwhile field. But, if it were a science, it
would be a poor one, because it cannot match the understanding of the
paradigm sciences. Look at the predictive power of physics vs. the
predictive power of economics.
Why do we do science anyways? We want to model, predict,and manipulate
phenomenon. Physics does a very good job of this. Economics does a fairly
poor job.
Yet, if you take it as what it is, it is worthwhile. Like politics. Even
though the writing of the US constitution was not scientific, it was a great
accomplishment.
It is true that economics uses some of the techniques of science. It also
uses other techniques and descriptions. My closest guess is that it is a
hybrid between the humanities and sciences.
Dan M.
Dan M.
