Induction attempts to justify scientific statements by reference to other specific scientific statements, and these are frequently more basic or fundamental statements. We have seen how this can lead to an infinite regression of assumptions, each of which must be proved. Hume, among others has written about this problem. To avoid this problem, the concept of Foundationalism was introduced, initially by Descartes, built up by Hume, and given modern form by Newton, Russell and others. This concept says that basic, self-evident, foundational beliefs exist and that these require no proof. These, then, serve as the basis for derived beliefs. Foundationalism may seem like a cop-out because it seems to say that at some point, you can’t have any more proof. But logically, there is no good alternative regardless of your preferred metaphysics. No matter how much your system can explain, there will be something underlying your system that is unexplainable. This is true in geometry, calculus, and physics as much as in religion and mythology – that is the nature of explanation in all its contexts. This is called the Regress problem, and some people are uncomfortable with it. In the search for certainty, to have to resign after several deep iterations is unsatisfying. But the alternative (an infinity of ever more refined explanations) is unworkable.
Foundationalists responds to the regress problem by claiming that these most basic beliefs do not themselves require justification by other beliefs. Such would be the case with Russell's Five Postulates and Newton's Rules of Reasoning in Natural Philosophy (both described elsewhere in this document). Sometimes, these “foundational” beliefs are characterized as beliefs of whose truth one is directly aware, or as beliefs that are self-justifying, or as beliefs that are infallible. According to one particularly permissive form of foundationalism, a belief may count as foundational, in the sense that it may be presumed true until defeating evidence appears, as long as the belief seems to its believer to be true. Others have argued that a belief is justified if it is based on perception or certain a priori considerations. In any case, it can appear to detractors as philosophical hand-waving.
Coherentism is a competing solution to the problem of induction – infinite regression. This model of knowledge asserts that scientific statements can be said to be valid if they fit cleanly into an existing coherent system of other known facts or beliefs. In other words, if they form part of a coherent whole (such as the existing body of science), they can be said to be correct. In this view, there is no requirement that scientific statements always be supported by more fundamental statements, instead they can be said to be provisionally “true” if they successfully serve their role in a network of mutually supporting scientific disciplines. Similarly, the fundamental statements that support more complex concepts in several disciplines are buttressed by their repeated successful application. For example, it is not possible to “prove” Newton's theory of gravity:
But it plays such a consistent and predictable role in so many situations that it is considered as true as any scientific principle can be (leaving relativity and quantum gravity aside...). Supporters of this way of looking at scientific statements include Willard Quine and E. O. Wilson who popularized another word for this concept: consilience. However, it was not Wilson who came up with the concept. William Whewell coined the term in 1840 when he said, "The Consilience of Inductions takes place when an Induction, obtained from one class of facts, coincides with an Induction obtained from another different class. Thus Consilience is a test of the truth of the Theory in which it occurs." Stated differently, Consilience is an assertion of the truth of the Theory in which is occurs. However, when a new observation conflicts with the existing body of knowledge, either the observation can be said to be incorrect, or the body of knowledge (e.g. existing theories) need to be modified. This is exactly what has happened with Newton's theory in the face of Einstein's discoveries.
Innumerable scientific observations from many disciplines support each other and provide confirmation and support for each other in very convincing ways. For example, Eddington's observations of light bending during a 1919 solar eclipse is considered the first evidence to provide solid support for Einstein's theory of General Relativity. This support didn't come from physics, per se, but from astronomy. Other astronomical phenomena (gravitational redshift of light) have provided equally compelling support.
Genetic research and the discovery of DNA supported and explained a mechanism for Darwin's theory of Evolution. Plate Tectonics explained how mountain ranges formed, which is coherent with much earlier discoveries of submarine fossils atop the peaks of our tallest mountain ranges and fossil similarities on the east coast of South America and west coast of Africa. Other coherent discoveries in geophysics involving magnetic field orientations in rocks on formerly adjacent plates have added additional support.
The chief criticism of foundationalism is that it can lead to the arbitrary or unjustified acceptance of certain basic beliefs. If we can all use personal preference to arrive at our unproven axioms, then strange and divergent belief systems can, will, and do emerge. The criticism of coherentism is that it is basically circular: A explains B, B explains C,and C explains A.
The only other alternative that is generally suggested is to accept the infinite regress and move on. These three choices (foundationalism, coherentism, and infinite regress) bear a close resemblance to the three legs of Münchhausen's trilemma (so named because Baron Münchhausen supposedly pulled himself out of a swamp by his own hair). Simply put, the trilemma factors all possible proofs for a theory into three categories:
- The circular argument, in which theory and proof support each other (coherentism)
- The regressive argument, in which each proof requires a further proof (infinite regress)
- The axiomatic argument, which rests on accepted precepts (foundationalism)
0 comments:
Post a Comment