Is there a scientific method? Physics, Biology and Beyond

post by Jimdrix_Hendri · 2019-12-05T16:50:01.105Z · LW · GW · 4 comments

Contents

4 comments

Among the general public and, frequently, in the educated media, one comes across naïve and uncritical praise for the “scientific method”. Often, accusation of violating the “method” is wielded to denigrate the viewpoint of a political opponent who supposedly offended against some prestigious, generally accepted norm of reasoning. I want to question the cogency of these arguments. My point is there is no agreed scientific method, as different sciences apply very different criteria in deciding what counts as valid explanation.

The paradigm approach for life scientists, for instance, begins by subjecting a phenomenon of interest to patient and thoroughgoing observation. They carefully describe key features, categorise functional and structural commonalities, then organise the material into a cladogram of some sort, after which they feel satisfied claiming they understand the phenomenon. I come across this approach again and again in Aristotle, who started his intellectual adventure as a zoologist.

For a physical scientist a biologist’s explanation is unconvincing. The physical sciences, with its strong emphasis on aetiology and relentless reductionism, raises questions not taken up in biology. Physicist are uncomfortable saying they know something until they can identify a small number of (ideally a single) exogenic factor/s giving rise to almost all the characteristics.

I need to emphasise that these are generalisable templates for explanation that are found outside of physics and biology. Thus, from the point of view of exogency, Jared Diamond’s explanation for the early ascendancy of western Eurasia as the result of geographical advantages (prevalence of domesticable animals and highly nutritious plants, etc.) is deeply satisfying to a physicist, since further pursuit for a cause moves the discussion outside the original domain of from human differences into geography. Compare this with a statement like: 'France’s preference for a strong centralised government is a natural continuation of the same policies having been implemented in the Roman Empire'. The statement passes the “holds water test”, but it is less satisfying, since it begs the question as to why Rome was like that. I.e., we are still in the realm of politics.

Explanations that are prefaced with “there are many reasons why ..” occur very frequently in biology are equally unsatisfying against the standards applied in physics. We also find cases where the physics template has been successfully applied within biology. The virulent arguments between Richard Dawkins and Stephen Jay Gould is (apart from some differences over political agenda) largely about differences a physics and biological view of evolution.

The physics template has enormously superior explanatory power, since its strong theoretical orientation enables predictions over a wide range of phenomena. This fact has not gone unnoticed in other disciplines (especially economics) where there continued, mostly unsatisfactory attempts to achieve the same deep theoretical understanding.

I’ve concentrated on the templates applying in the physical and life sciences, but there is a further paradigm for explanation found chiefly in the so-called soft or social sciences (although you meet it sometimes in the life sciences as well). These are disciplines that rely nearly exclusively on empirical results, since theory is either undeveloped or not trustworthy. A typical social science explanation is that a double-blind test of hypothesis X carried out over a population characterised by Y produced a positive outcome with a p-value 90%. This amounts to saying: given a certain input, we can anticipate a certain output most of the time. And we may even have some concepts that guided us in designing the experiment. But we are far from confident of its general applicability.

4 comments

Comments sorted by top scores.