“Everything should be made as simple as possible, but not simpler.” — Albert Einstein
Reading your position on the status of the model reassured me somewhat. I’ve read many, many more research papers in these past few weeks than in my entire life, and, frankly, the degree of trust and respect for a lot of them is lowering. I want to model myself–my own development–only on the best scholars, but there are so many journals, so many papers that the sheer amount of text out there is overwhelming. At the same time, finding those gems, those actually helpful papers is not an easy task. Your input, Dr. A. is so welcome.
<humour>As a Scot, I’m culturally programmed to dislike Occam’s razor.</humour> Occam came a generation after Duns Scotus (after whom the English took the term ‘dunce’ to mean ‘stupid’). Duns Scotus was a bit of a Socrates-like figure who went around Europe playing the devil’s advocate by turning arguments upside down. Occam was a swing back to simplicity. The notion of parsimony has been a distractor ever since, as the above quotation from Einstein suggests.
Sagan’s requirement that “extraordinary claims require extraordinary evidence” informs my basic worldview. Taking this beyond the extraordinary, I look for evidence for everything. This leads me down the rabbit hole frequently. It doesn’t matter if any scholar says such-and-such, it’s not the scholar, it’s the quality of the claim that’s important. Hence the necessity to reference clearly, to be able to follow a thought down that rabbit hole. Often the trail leads to an opinion, like Beeler’s, or to an empirically tested model. Sometimes along the way, there may be amendments and complimentary additions. I stop searching when it is clear that the discussion is a pure ontological one.
From your use of the word parsimony, I think this morning that I got a glimpse into why scientists insist on that. Please correct me if I’m wrong, but the discussion last week on quantitative models reminded me of the techniques of multiple regression and the concept of multicolinearity–when two independent variables highly correlate. If two variables are found to be multicolinear, an underlying and as yet unstated explanation for both of them should be looked for. Is it possible that parsimony is a function of the statistical modelling underlying multiple regression more than an ideological stance about the nature of explanation? I can see how the elements in this question overlap, but if the former, there may be a principled and justified reason for qualitative researchers to bypass criticism from the positivist camp. I didn’t see this argument in the reading last week, but it strikes me as plausible.