This is based on a short piece I wrote a couple of years ago about what I think is the nature of science, what makes something ‘scientific’, when scientific reasoning is reliable, and how it can be justified. I have updated some element to reflect my most recent thinking on the matter.
A Model of Science
- Science is the process of carefully collecting objectively verifiable empirical data about the world, categorizing this data to help better understand and access it, and using such data to construct rigorous, testable models about the way various phenomena behave.
- There is no philosophically interesting ‘fundamental’ methodological difference between the natural sciences and the social sciences. All are equally scientific to the degree to which they adhere to point 1.
- Confidence in a given scientific model is increased to the degree that it is based upon objectively verifiable and carefully collected empirical observations, to the degree that it makes specific verifiable predictions, and to the degree that it appeals to fewer entities or processes that are not otherwise known to exist (i.e. to the degree that it makes fewer new assumptions).
- A scientific model is said to explain some system to the extent that the model is able to make specific predictions about the behaviour of the system under a given set of circumstances. These ‘predictions’ need not necessarily be of future events or experimental results, but they must be empirical in some form.
- Scientific models are relevant or applicable in a particular situation to the degree to which they are based on observations from similar circumstances/phenomena, and to the degree that they have made successful predictions under similar circumstances.
- This methodology is self-consistent as it presents a descriptive and normative model for science which is based on objective facts about the history, practice and past successes of science, and makes predictions about what sorts of scientific methodologies will generate useful results.
Preempting Some Rebuttals
- Whatever the nature of the ‘ultimate reality’ we inhabit, so long as it exhibits at least some degree of predictability or regularity, then this methodology would be expected to arrive at reliable beliefs about that reality. Thus the ‘Problem of Induction’, while not directly answered, does not seem to hold a great deal of force against Science.
- Science has been very successful in the past at making accurate predictions and generating technological advances, therefore it is reasonable to suppose that the method has some degree of validity.
- Science does not require perfect objectivity or rationality of its practitioners, nor does it require or assume complete separation of observers from their observations.
- Science does not presuppose naturalism or materialism, or require the existence of universal natural laws. It requires only the sorts of regularities described in 1.
A Controversial Claim
- Science, as described here, is the only method of arriving at reliable, justifiable knowledge about any aspect of the world outside of our own direct, everyday experience (a concept I will define more rigorously elsewhere).
- Intuition, commonsense, philosophical reasoning, subjective experience, emotion, tradition, religious dogma, and political ideology are not reliable methods of arriving at knowledge of the external world outside of our own direct, everyday experience.