Chad’s been doing a series on non-adademic careers for scientists, and the first in this year’s batch: PNAS: Amy Young, Saponifier
It’s a noble effort , reminding people that there are many options outside of the professor-begetting-another-professor path, which is not sustainable, but the reason I really took notice was that this brought another topic onto some sharper focus, namely my position that everyday cooking/baking — when one simply follows the recipe in the cookbook — is not science. A lot of cooking, I think, doesn’t get past the level of I cooked too long and it’s burned/dried out, which is barely dipping your toe into the soufflé of science. (And Jennifer seemingly disagreed with this position, but it turns out it was mostly semantics — that cooking, done properly, is not about blindly following recipes is something with which I agree. The issue is blindly following recipes.)
Here’s the relevant part of Amy’s soap cooking approach that isn’t always followed in food preparation
If I hadn’t had the importance of keeping a proper lab notebook drilled into my head in my formative years, I would never in a hundred years be able to keep up with all the product lines I’ve got now. (Which colorant did I put in this one, again? And how much? Wait, wasn’t this the fragrance that made the soap seize up on me last time? I should probably try a lower temperature. And so on.) It may be six months or more between making batches of a given kind of soap, so keeping track is vital. Not to mention the product development phase, in which the thing just doesn’t work right, and I have fifteen different things to try varying; I’ve talked with colleagues who run similar businesses, and they seem to operate in a “just change stuff until it works” mode, rather than changing one element at a time (even if I run a dozen or more iterations simultaneously) so as to know which thing or combination of things created the desired effect. It’s invaluable in crafting the more complex items.
This really shows the systematic approach; it’s important to know what cause leads to which effect, and to quantify what you’ve done.
professor-begetting-another-professor path, which is not sustainable Social Security is precisely that. Official SS death date is 2036. Uncle Al is betting on ~2015 – and accordingly selling short. Uncle Al has golden hands in the lab.
rather than changing one element at a time (even if I run a dozen or more iterations simultaneously) Theory of Experimentation. You run all the variables at once in a surface-fitted multidimensional grid. On a good day it is much faster and cheaper toward optimization (if you can measure what is being optimized). On a bad day you have confounded variables (like hypertension vs. chronic sodium intake). Confounded variables cannot be disentangled except by simultaneous multivariable runs.
Medicine has a retrocranial infarct regarding hypertension etiology and chronic sodium intake. East Asians have an immense daily sodium intake all their lives but a low incidence of mid-life hypertension. This “protection” vanishes when they emigrate to the North America. Thousands of very expansive expensive studies come to no conclusion at all, nor should they. Chronic Na/K intake is the confounded variable – heavy on the veggies protects against fish sauce, soy sauce, and brine-pickled everything.