Also this week: a rare Javan rhino link from Brian 🙂 , ecology vs. strong inference, the rules of the (status) game, 150 years of The Descent of Man, interspecies money, fear vs. cicadas, Jeremy continues to inflict share music, and more.
From Brian (!):
For those readers who are interested in processes of science and how science does (and does not) work, this report on the history of the debate over whether COVID is airborne or not is fascinating. It’s got lots of classics. A mysterious number that is taken as gospel even though nobody knows where it came from. Dominance of a single personality over the debate. Terminology that obfuscates more than it enlightens. A conflation of ideas that seem intuitive until you break it apart and then see they never should have been conflated. Outsiders who are ignored for a long time before they are listened to. (NB: I think the authors of the piece oversell the implications of COVID being airborne even if the scientists reported on didn’t – the real point is airborne/droplet is an artificial dichotomy and it is a continuum – COVID has an airborne component but it still (fortunately) behaves much more like influenza in terms of travel patterns than say measles or tuberculosis although the early WHO/CDC emphasis on cleaning surfaces over masks was a bad mistake due at least in part, although probably only in part, to sloppy thinking on this topic).
From Jeremy:
A while back, I asked why ecologists don’t often use “strong inference”: developing alternative hypotheses and then collecting the data that would distinguish among those hypotheses. I think ecologists could and should use strong inference more often–but maybe I’m wrong about that! Here’s Martin Johnsson arguing that it’s fine that ecologists don’t use strong inference. He notes that the inference that ecology would be improved by more frequent use of strong inference isn’t itself a strong inference. (Aside: I only just stumbled across Martin’s blog; it looks very interesting. Added it to the blogroll. I’m enjoying digging through the old posts.)
Jessica Hullman and Andrew Gelman on feeling like an academic pariah even when you’re not. Related post on whether being involved in a scientific controversy will help or hurt your scientific career. Also related (aside: that last post is one of my better efforts, I think.) Also related, from Meghan (it might not seem like it’s related, but it definitely is.)
A piece of career advice that worked well for former marine biology grad student Ananya Sen: find something you’re good at, that makes you happy 80% of the time.
Two contrasting looks back on Darwin’s The Descent of Man and its influence, 150 years after its publication.
Sabine Kleinert and Elizabeth Wager with some ideas on how universities and journals can better coordinate to address cases of possible scientific misconduct. Personally, I’m coming round to the idea that more countries should follow Sweden’s lead on this. Have an independent national agency charged with investigating cases of possible scientific misconduct. Those investigations should be run by scientists, I think, not lawyers. But it’s not something I’ve thought much about yet, and so I’m pretty hazy on the details of how such an agency would work. Would welcome comments from those who know more than me about the Swedish experience on this.
Philosopher Agnes Callard on status games. A couple of sample quotes, to encourage you to click through:
In an academic context, I’ve noticed that complaining about how busy one is hits a sweet spot of oppression—I cannot manage my life!—and importance—because I am so in demand!
[W]e need worth to come for free, and we also need it to be acquirable. And no philosopher—not Kant, not Aristotle, not Nietzsche, not I—has yet figured out how to construct a moral theory that allows us to say both of those things. This is a giant unsolved problem, and it touches us all. We may not explicitly articulate it, but we feel it, and project the psychic turmoil it generates onto our interactions with one another
And here’s Ezra Klein interviewing Agnes Callard.
And here’s Agnes Callard on publish and perish.
xkcd vs. Muller’s ratchet. 🙂
Interspecies money. I’m pretty sure this is satire. But these days, the line between “satire” and “not satire” is…blurry.
And finally, continuing my little series of Cover Versions That Are Different Than, And Even Better Than, The Original, here’s “I Still Haven’t Found What I’m Looking For” as gospel soul:
Have a good weekend. 🙂
Dynamic Ecology 
Every time I hear people argue for Platt’s strong inference in ecology, I flash back to a story Art Winfree shared with a seminar of first year grad students (including me). A certain species of fungi was known for creating a striped growth pattern. He came up with 18 possible causes/hypotheses. He proceeded to devise and run experiments for all 18, and rejected all 18 hypotheses! He gave up on ecology and went into theory of temporal patterning in petri dishes, hearts, and other things and had an amazing career (my judgment, not his).
Strong inference is great when there are only two proposed carriers (DNA, proteins) for the genetic code and they are completely distinct (the necessity for strong inference being hypotheses that are non-overalapping and comprehensive). But in a mutlicausal world (https://dynamicecology.wordpress.com/2016/03/02/why-ecology-is-hard-and-fun-multicausality/) neither are true.
I’m personally much more sympathetic with Chamberlain’s much older paper arguing for multiple competing hypotheses – with an emphasis on countering human psychology by forcing us to explore the realm of possibilities and not get overly attached to a pet theory rather than Platt who really emphasizes this ability walk down a branching path of hypotheses that requires a set of comprehensive, non-overlapping hypotheses (the forks in the branching path).
But either way, yeah ecologists don’t do the multiple hypothesis thing very often. Probably too afraid the same thing will happen to them as happened to Art. At least if you only test one hypothesis you can publish the paper either way (and yes of course I agree that rejecting 18 hypotheses should be even more publishable than rejecting one, but not sure our world works that way).
Agreed, Brian. Although I’d interject that it is npossible to develop non-overlapping & comprehensive hypotheses in ecology. From my experience, it requires a lot more work in terms of study design, but I found it well worth the effort.