JoeDawg said:
One of the big problems in the philosophy of science is called the 'demarcation' problem.
http://en.wikipedia.org/wiki/Demarcation_problem
If philosophy is about generalizations and science is about specific measurments, then categorizing species in biology, categorizing stars in astronomy, quite a lot of archeology, paleontology, and geology; are philosophy.
Add to that, measuring the positions of stars, in order to create a horoscope, is science.
I don't think most scientists would agree with that.
Demarcation problem is really about finding reasons to put science on a pedestal which is quite the wrong kind of defence I believe.
The better approach (and here I adhere most closely to Rosen's Modelling Relations epistemology) is to see modelling as a democratic exercise. Sit back, let the models compete, and the most effective (for their purposes) will emerge from the fray.
So astrology to the point that it models the world in some accurate sense is doing the job we would want of any "science". And yes, it is a modelling approach that attempts to extract general principles and then relies on measurements to drive the "equations". It relies on local inputs such as your birth date.
Where astrology falls down in the modelling relations view is that while the models make predictions, there is a lack of feedback about the observed outcome. As a field, it is not systematic in closing the measurement loop so that the model - the generalisation - is subject to learning.
So when I say science is measurement, I mean precisely the kind of predict and test loop that we expect of science.
Of course, this is not just how good modelling works, it is how brains work. Our minds anticipate the next moment's sensory input and then responds to the errors in prediction, paying attention to what was not expected, what was surprising, and then updating a running model of the world. When everything is being smoothly predicted, they call it the flow experience.
The point I was making is that - as modelling relations, a modern approach to epistemology, says - the business of modelling is naturally dichotomous. It divides strongly into formalised ideas and informal measurements. Generals and particulars. And then having divided, the two different aspects are better able to interact. The more clearly we can say this is our model that makes the guesses (the anticipations) the more clearly we can say well here is a prediction that didn't work out and so perhaps the model needs more work.
For some reason, philosophy and science have become disjointed activities in the sociology of human knowing. I am explaining why a division would emerge, but then also why the two have to work together.
We have similar divides at lower levels of course. In particle physics, there are the theory groups and the experimental groups. Once you get a big gang like at Cern, it seems natural to divide. But academic philosophy and science have floated right apart.
So forget the current disfunctional academic scene. I am saying look at how minds know worlds, look at some sophisticated epistemology like modelling relations. Then you can understand both the value of a divide, and also the way the divided are meant to interact.
And probably I am not making myself clear on what is meant by generals and particulars. In biology for example, darwinian selection would be a general, the categorisation of species would be closer to a particularisation activity. Evolution gives us the general idea that all species would be related like branches of a common tree. Then we would predict and measure to place all observable species on such a tree.
Or just using a straight cognitive example, you will have a general idea of what is a cat. It may include lions, Sylvester or Garfield, etc. Then you will also have specific cat impressions. You will have things come into your field of view and make some kind of concrete local act of interpretation. A prediction that may be confounded if the cat starts barking.