Posted by: Jeremy Fox | August 24, 2011

In which I try to get readers to do my job for me

I’m a little short on ideas for posts at the moment. I’ll probably do something on path analysis at some point (if only to start an argument with Jarrett Byrnes). Maybe something on Bayesian vs. frequentist statistics. And probably some more looks back at classic, or should-be-classic, Oikos papers. But for the moment, my brain is mostly empty of bloggable ideas.

So let me ask you: What should I be posting about? What are folks talking about in ecology and evolution right now? What should we be talking about? What are you just dying to hear my considered opinion on?

About these ads


  1. In the Borders fire sale I picked up a volume of the “best” of Stehpen J. Gould. It got me thinking about how much I read him before I was a graduate student and just a field assistant. Then when I started reading all about evolution during the time he was active I realized other than his Spandrels paper, I never read anything of his. The giants of the 70′s and 80′s evolution does not include him, but he was the public face of evolution for so long. It gets me thinking about the disjunct between the public and academic intellectual, and who are currently big name intellectuals, especially after reading this piece about Daniel Lieberman:

    I have some other generic research blogging things on my list, and I’m also a fan of writing some simple tutorials on methods because I’m always trying to learn new methods.

    • Cheers for this Ted.

      Another post I’ve been thinking of writing is something trashing the Spandrels of San Marco, which is in my humble opinion both deservedly famous and pretty bad. But that paper’s been so talked about I’m not sure it’s worth adding more verbiage to the pile.

      • Oh, trashing the Spandrels paper would be fun – submit it to the Carnival of Evolution, and you’ll get a lot of attention.

        Like Edmund I’ve noticed the disconnect between the fame of Gould (and Dawkins), and their contributions to evolutionary biology. I’m sure there’s a blog post in there somewhere too.

        BTW, I assume you’ve read Segestråle’s book “Defenders of the Truth”. She fills in some of the background about the PedantivesSpandrels paper.

      • Thanks Bob. I haven’t read Defenders of the Truth, though I should probably do so before posting anything on Spandrels.

        I have read Dennett’s attack on Spandrels, and the very interesting Trends in Ecology & Evolution paper (from an architectural historian) gently correcting both Dennett’s and Gould & Lewontin’s knowledge of architecture, thereby implicitly undermining both their spandrel-based arguments (briefly, it turns out spandrels are neither ‘non-adaptive’ nor what Dennett called an ‘obligatory design opportunity’). There are a number of larger conclusions which could be included in a post, such as the slippery nature of analogy-based reasoning. CORRECTION: It’s a 1996 American Scientist paper from architectural historian Robert Mark; it was cited in a letter to TREE in 1997 by Alasdair Houston.

        The disconnect between the public fame of Gould and Dawkins, and their original contributions to evolutionary biology, is something which I’m sure others have commented on, but I may add my two cents if I can come up with something new(ish) to say (e.g., about the feedback from public fame to academic science–a lot of academic research in evolutionary biology was motivated by the desire to address broad claims Gould initially, or primarily, addressed to a popular audience, such as the claim that the “tape of life” would play out differently if it were “rewound”).

      • Ok, you asked for it:

        I feel like I really tossed some bombs in this one (starting with the title!), hope I don’t end up provoking an over-intense reaction. Although the other times I’ve been afraid of that (e.g., my first “zombie” post, my post on refighting the null model wars), the reaction was actually very positive. And I can hardly imagine that anything I’ve said about “Spandrels” hasn’t already been said, and more forcefully, by someone else. We’ll see!

  2. FYI, before you hop on path analysis, be sure to read some stuff about D-separation from Bill Shipley. I’ve been digging in lately, and finding the discussion fascinating. That, and comments on semnet about Judea Pearl’s work.

    • Boy, all I had to do was threaten to post and I drew a comment from you! ;-)

      Yes, I’m aware of D-separation and the emphasis Shipley places on it. Also the fact that D-separation doesn’t work in the context of nonlinear systems, or systems with causal feedbacks (something Shipley himself emphasizes, but which doesn’t seem to have been taken to heart by the broader community as far as I can tell, though I have yet to do all the background reading I’d want to do before posting). More broadly, back in grad school I did a review paper on path analysis and structural equation modeling as applied in ecology and evolution, so although I’m not Bill Shipley they are approaches I’ve thought about for a while.

      Wasn’t aware of Judea Pearl’s work or comments about it on semnet. Indeed, I wasn’t even aware of “semnet”. ;-) Will have to look that up.

      • Jeremy & Jarrett: it would be great to read some fresh stuff on the application of SEM in biology from you; personally, a few years ago I devoted myself to reading and studying the “cosmogony” of structural equations with latent variables, including Pearl’s catching work on this, and I found it fascinating. I do think that there is still much potential in this techniques when applied to general ecological problems (but few people have explored this)!

  3. Why not write a piece (or series?) introducing ecologists to the Price equation and its applications? You’ve piqued my interest in it, but a primer could be nice before delving into mathematical/theoretical literature.

    • I haven’t done it only because I may write a paper on that at some point.

  4. Concerning d-separation: it does apply to nonlinear systems and is complete in this case. In systems with feedback d-sep still applies if the probability distribution over the variables is multivariate normal but not otherwise (i.e. it is possible for there to exist partial independencies that are not identified by d-separation).
    -Bill Shipley

    • Thanks Bill. I was going by memory in my earlier comment, I certainly plan to read up on d-separation before posting anything. Although at this rate this comment thread will end up containing everything I’d ever want to post. ;-)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s



Get every new post delivered to your Inbox.

Join 3,011 other followers

%d bloggers like this: