Recently, I polled y’all on various ecological and evolutionary laws and rules: Kleiber’s law, Haldane’s rule, the link-species scaling law, etc. Respondents were asked to rate each law on a 3-point scale, from 1 (rarely or never true, or very far from the truth, or true in only very restricted circumstances, or etc., and so definitely not a law) to 3 (definitely a law). Respondents were also asked to indicate their level of expertise on each purported rule/law, and to skip any rules/laws about which they knew nothing. (UPDATE: here’s a view-only link to the spreadsheet of the poll results, so that you can see all the laws that were included in the poll and all the responses.)
I was interested in this for a couple of reasons. One was as a novel starting point for talking about “laws” in ecology and evolution. There is of course a lot of discussion of scientific laws in the ecological and evolutionary literature, much of it from a philosophical point of view. Which is totally fine. But another way to start thinking about laws is just to note that, in practice, ecology and evolutionary biology have various claims that are called “laws” (or “rules”). It’s right there in the name! One of the most basic questions you can ask about those laws/rules is if they’re true (or close to true, or true in a wide range of circumstances, or etc.). Because if they’re not true (or not close enough to true, or etc.), then surely they can’t actually be laws. Looking at the claims that ecologists and evolutionary biologists actually do refer to as “laws”, and at which of those claims are (seen to be) true, might help us see those philosophical discussions in a new light.
The second reason I was interested in this was because I’m interested in scientific agreement and disagreement. Disagreement is a normal part of science–even in contexts in which you’d think there wouldn’t be any scope for disagreement. Such as questions like “Is Kleiber’s law actually a law?”, which you would think would have an objective answer on which all sufficiently knowledgeable experts would agree. We know from an old poll that there are many controversial ideas in ecology on which the experts actually disagree with one another more than the non-experts do. I was curious whether the same was true with regard to ecological laws. Perhaps by looking for patterns in scientists’ agreements and disagreements with one another, we can learn something about why those disagreements persist.
Unfortunately, what with the COVID-19 pandemic having stomped our traffic like a vat of grapes at a winemaking festival, we only got 57 responses. So the results are even more anecdata-y than is usual for our polls. But still, there are a few interesting tidbits that I’m pretty sure are robust to our modest sample size, and so are worth talking about.
Ecologists and evolutionary biologists don’t all agree with one another about anything. We know that from past polls, and this one reinforces the point. Every law on the list had at least a couple of respondents who think it’s definitely a law, and at least a couple who think it’s definitely not. And there was no law on the list which received the same score from >65% of respondents. Now, I suppose it’s possible that a bit of that disagreement is due to different people having different notions of what a “law” is. Like, if you think “law” means “universally applicable, exceptionless true statement”, then you probably won’t think any of the laws in the poll are actually laws. But just eyeballing people’s responses, it doesn’t seem like that explains much of the disagreement.
Most ecological and evolutionary laws are widely seen to have at least some basis in fact. Most of the laws on the list were scored as 2 or 3 by a majority of respondents. Many of these laws were “stylized facts” at the time they were first proposed. Apparently, they mostly still retain something like that status in the eyes of the majority of respondents. Maybe that means that ecologists and evolutionary biologists are pretty good at identifying “stylized facts” that continue to serve as useful research foci long after they’re first proposed. Or maybe it means that ecologists and evolutionary biologists are too slow to give up on purported “stylized facts” even as contrary data accumulate. Discuss.
Some purported laws definitely are laws, at least in many respondents’ eyes. Respondents mostly agreed that the latitudinal species richness gradient, power law species-area curves, Kleiber’s law, and Haldane’s rule are laws. For each of those claims, 40-55% of respondents scored them as “3” (definitely a law), and the majority of the remaining respondents scored them as “2”. All of these laws had a mean score of 2.3-2.4. (aside: yes, yes, I know these are really ordinal scale data and so we really shouldn’t be taking means. But this is just a blog post, means are fine here as quick-and-dirty data summaries.)
Some purported laws definitely are not laws, at least in many respondents’ eyes. Farr’s law and Cope’s rule were the lowest-scoring laws on the list, with mean scores just below 1.5. Both Farr’s law and Cope’s rule got a “1” from >60% of respondents. I have to say, in light of the COVID-19 data, I was very surprised that Farr’s law–the claim that epidemics rise and fall in a symmetrical, bell-shaped curve–didn’t score even worse. Not because COVID-19 is the only counterexample to Farr’s law–it’s not–but just because it’s a counterexample that’s been so prominent in the news lately. I’m genuinely curious to hear comments from anyone scored Farr’s law as a 2 or a 3. Not out to criticize you (although I do respectfully disagree with you, given what I know about disease dynamics). Just honestly curious to hear your thinking.
The most controversial laws. There was no law on the list for which the distribution of opinion was bimodal. But there were three for which the distribution of opinion was fairly close to uniform: Dollo’s law, the link-species scaling law, and Damuth’s law. Dollo’s law received a “3” from 48% of respondents, a “2” from 27%, and a “1” from 24%. The link-species scaling law received a “3” from 23% of respondents, a “2” from 37%, and a “1” from 40%. Damuth’s law received a “3” from 27% of respondents, a “2” from 38%, and a “1” from 35%. It’s interesting to me that the three most controversial laws include one that’s infamously vague (Dollo’s law), one that’s a precise quantitative claim (Damuth’s law), and one that’s intermediate (the link-species scaling law, which says that one variable is proportional to another but doesn’t specify the constant of proportionality). Just eyeballing these admittedly-crude data, I don’t see much sign that the vagueness of a law is an important determinant of how controversial it is.
Experts don’t “fight the law”. This was the most surprising and interesting result to me. For 13/18 laws in the poll, the respondents with the most expertise* scored them higher on average than did all respondents. And the difference in mean scores was pretty substantial–the average expert score for a given law was about 0.7-0.8 higher than average score from all respondents. Here’s a figure:
Now, it’s possible this is a blip that wouldn’t hold if we polled again. For many laws, we had only 2-3 expert respondents, so there’s a lot of sampling error in the mean expert response. And the respondents who indicated that they were experts on one law often indicated they were experts on some other law(s), so the expert responses for different laws aren’t all independent of one another. But it’s a strong enough trend that it’s maybe worth talking about. I’m particularly intrigued because the trend here runs in the opposite direction to one in our our old poll on controversial ideas in ecology. In that poll, the experts on any given idea were more likely to disbelieve that idea (or less likely to believe it), as compared to all respondents.
Maybe part of what’s going on here is that anyone who goes to the trouble of learning enough about a purported law to become an expert on it presumably thinks there’s something to it. It’s surely rare for anyone to go to the trouble of becoming an expert on some claim that they think is just rubbish,** or to come to believe that some claim they’ve spent years studying is just rubbish.
Can’t really tell if there’s more disagreement among experts than among non-experts. Another striking result from that old poll on controversial ideas was that those ideas were more controversial among experts than non-experts. That is, experts on any given idea disagreed with one another more than did the non-experts. Unfortunately, the sample sizes in this poll are too small to tell if the same is true with regard to opinions about ecological and evolutionary laws.
Looking forward to your comments.
*i.e. the self-described “experts”, except for the two laws for which we had <2 expert respondents. For those two, I also included the respondents with “some” expertise.
**Though I can think of a few examples. For instance.