Why are all zombie ideas simple?

Possibly-weird question, inspired by this tweet: why are all zombie ideas simple? Why is it that, whenever many people working in a field believe X despite a lack of good evidence and arguments for X, or even despite good evidence and arguments against X, X is invariably some quite simple claim?

Is it that complicated ideas are never widely believed by anyone, so the only ideas that can possibly become zombie ideas are simple? Is it that zombie ideas invariably are intuitively appealing (that’s what protects them from contrary evidence and arguments), and to be intuitively appealing an idea has to be simple? Is it that ideas become zombies by being misunderstood or misapplied, and that misunderstandings and misapplications invariably involve simplifying complexity and nuances? Something else?

Or am I just wrong about zombie ideas invariably being simple claims? Maybe I’m just vastly overgeneralizing from a small sample size?

You tell me, because I dunno. Looking forward to your comments, as always.

p.s. To be clear, I don’t think that most or all ecological ideas are zombies. You should not infer from the fact that I often write about zombie ideas that I think everybody but me is wrong about everything! Zombie ideas are interesting and important to think about even though they’re rare, because they’re cases of entire scientific fields getting stuck.

p.p.s. As always, no personal criticism of anyone is intended or implied by the suggestion that some widely-held scientific ideas are incorrect and ought to be abandoned. Science is hard and disagreement is a normal part of it. And it’s handy and fun to have shorthand phrases like “zombie ideas” and “Buddy Holly ideas” for different ways in which the collective scientific process occasionally breaks down. 🙂

16 thoughts on “Why are all zombie ideas simple?

  1. Zombie ideas are all simple for the same reason that most popular ecological theories are simple: they require very little knowledge or understanding of the field to be understood, and the conclusions to be drawn are evident from the LIMITED set of assumptions and phenomena that are presented by the author. IMHO, Grime’s CSR theory is a great example of a zombie idea that became (and, for many people, still is) a major ecological hypothesis. The beauty of it is that you don’t need to know ANY details of physiological ecology or biomechanics to understand its rationale, and the (highly flawed) data advanced by the author appear to be compelling, Throwing the theory over completely (and, in my opinion, that has yet to be done) requires an understanding of plant physiology and biomechanics and the experimental literature in plant ecology. That – and the key fact that many “predictions” of CSR theory comport with known ecological patterns, even if they are the right conclusions drawn for the wrong reasons – make CSR theory a zombie.

    • “CSR theory is a zombie idea” is one of those posts that probably everyone thinks I’ve already written, even though I haven’t. I don’t feel like I know enough to write it, although based on the modest amount I do know I share your view that it’s a zombie idea.

  2. Zombie ideas are not limited to ecology, for example the number one zombie idea in molecular biology is that Ethidium Bromide is dangerous. Fits your description perfectly, simple intuitive idea that anything that makes DNA glow must be dangerous yet the preponderance of primary literature and veterinary evidence says otherwise. As we all know, this particular one has even wormed its way into regulatory bodies in a big way.

    I suspect that agreement with orthodoxy would be another big factor. Anything that would be uncomfortable to deny either socially or professionally certainly increases the probability of Zombization..

  3. Dear Jeremy,
    I’m not sure zombie ideas fall on a simple vs. complicated axis. Your post implies that simple ideas are more likely to become zombies. What then is the fate of complicated ideas? If a zombie idea is one that aims to be general, but fails in specific cases, then what is an idea that aims to be specific, but fails to generalize? A Medusa idea? There should to be two sides to this coin.

  4. For a great many seemingly simple ecological ideas, it’s hard to pin down what would constitute rejection to the point of assigning the zombie classification. CSR theory is a good example. The core asserts that stress (anything that limits plant growth) and disturbance (anything that destroys plant biomass) are two dominant factors driving the evolution of plants and therefore what plant traits are associated with different environmental conditions. A lack of either means that the main constraint on fitness will be interactions with other plant species (competition). Traits that confer high fitness in one condition are different than those that confer high fitness in other conditions (i.e., there are tradeoffs), so we get trait-environment correlations across species. There may have been particular things proposed within the context of CSR that don’t hold up to scrutiny, but is the overarching perspective wrong? I suppose I’m wondering if one person might call “zombie” because theory X is useless in context Y, while someone else finds it very useful in context Z.

    Island biogeography is another example. Some people think of it as no more than the idea that richness results from a balance between colonization and extinction (true by definition in the absence of speciation), others think it’s the crossing-curves graph, and others think of every sentence of the book as being part of the theory (I have had this comment – not a straw man). What would have to happen to render the theory a zombie? Not obvious to me.

    • Yes, this point came up in Twitter discussion about this post too. And Doug Sheil and David Burslem’s counterargument to my claim that the IDH is a zombie idea is basically to argue that the ideas I’m criticizing aren’t actually part of the IDH, properly understood (https://dynamicecology.wordpress.com/2013/09/03/a-thumbs-up-for-the-intermediate-disturbance-hypothesis-guest-post/). It’s a fair point. Though I’m not sure that “this idea is so vague that nobody can agree on exactly what the idea is, the circumstances in which it applies, or how to test it” is a huge improvement on “this ideas is a zombie idea”.

      There are times when vagueness is a virtue in science (https://dynamicecology.wordpress.com/2015/06/21/when-is-vagueness-a-virtue-in-science/). But FWIW (which may not be much!), I think vagueness tends to hinder ecological progress more than it helps.

      Maybe instead of trying to develop ideas that apply generally, and that achieve their generality by being vague, ecologists ought to seek generality using other approaches.

      (Note: comment edited slightly after initial posting to add a reference to Sheil & Burslem)

      • Interesting points. I’m not sure the ideas themselves are always all that vague, but different “users” draw not on the original papers/books but secondary ones, ultimately lead to ambiguity in people’s minds about what is in or out of theory X. Science can also progress by keeping the bits that work (the baby), tossing those that don’t (the bathwater), but the zombie label sometimes feels like dropping all of it and so letting the baby go with the bathwater.

      • Sheil & Burslem agree with you that I’m throwing the baby out with the bathwater when it comes to the IDH. To which my response is, attempts to narrow/redefine/elaborate the IDH have been going on for decades. If they were going to turn Connell & Huston’s original idea into a productive, progressive research program, they’d have done so by now.

        Maybe one way (the only way?) you can get people to separate the baby from the bathwater is by first threatening to throw them both out the window.

      • Mark’s point has a lot of parallels to conversations we frequently have in the comments about statistical techniques. The techniques have strengths and weaknesses of which the original proposers are aware. Its the subsequent cultural practices that inflate them into flawless techniques that should always be used.

        Which raises the hypothesis: maybe zombie ideas are simple because zombie ideas by definition must be widespread, and the only ideas that can culturally spread widely are simple ones. Can a nuanced, complex idea ever be widespread?

      • Brian,
        “Can a nuanced, complex idea ever be widespread?” Who has time for that? Partly joking, but the original proposers of theory/methods went into the subject in depth. Many people (not all) using their work read the abstract of the paper, gleaned the methods, and went from there (oversimplifying and exaggerating a bit here). So, yeah, it makes sense that people, over time, might take an originally more nuanced method/theory and morph it into something simpler (e.g., from secondary sources as Mark suggests) that spreads and becomes widely accepted.

        The take home message for me is that zombie ideas spread because we as a community are not using best practices as rigorously as we should, at least partly because of the pressures of time. Which means (groan) that I really should spend much more time using primary sources to work through the theory/methods as if they were my own ideas.

      • “So, yeah, it makes sense that people, over time, might take an originally more nuanced method/theory and morph it into something simpler (e.g., from secondary sources as Mark suggests) that spreads and becomes widely accepted. ”

        Zombie idea = game of telephone! 🙂 Hmm, could be something too this in some cases. But not in others, I don’t think. In other cases, I think the simple idea that became a zombie was there from the start. For instance, textbooks and other secondary sources actually do a good job of summarizing the really influential early papers on the IDH, and don’t leave anything out. Huston (1979) is one short paper about one simple model, not a long monograph reporting a complex set of observations plus experiments plus a bunch of different models. As another example, Hutchinson’s paradox of the plankton paper is short and the main idea (which became a zombie) is simple and briefly stated, and accurately summarized by subsequent textbooks.

  5. Slightly different framing – a Zombie idea must be simple, in order for it to be easy to recall (not just understood). Non sub-field experts are unlikely to remember really complex ideas. If you’re not a sub-field expert you may only be exposed to the zombie idea a few times, therefore it must be easy to memorize for it to take off.

    • Yes, I think this is one of the main reasons. Also, if you are the professor of some undergrad basic course and you have a restricted time (and sometimes also a restricted knowledge of the entire subject), it is reasonable to focus in simple ideas. Simple zombie ideas are not only more reminiscent but also more teachable.

Leave a Comment

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

WordPress.com Logo

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.