I kind of enjoy the raging debates that you often see in the literature in the broad fields of ecology and evolution. These debates can be useful, because folks feel strongly about the points that they are making, and so often marshal all of their resources and rhetorical skills to advance their position. Although I definitely see the benefits of these debates, I often feel that many of them boil down to one person saying that Factor A is the only important thing for explaining some pattern, and the other side is saying the same thing about Factor B. But this makes me wonder why organismal biologists would think that most patterns are driven by a single causal factor, when I think the universe is multivariate and multicausal.
In my own little corner of organismal biology, I have noticed this pattern in studies of the evolution of viviparity in snakes and lizards. We don’t need to get into the weeds on this, but the dominant hypotheses for the evolution of viviparity is that viviparous species can regulate temperature of the eggs, while oviparous species cannot (squamates usually have eggs in burrows that are not attended). Hence, selection would favor egg retention, with an endpoint of viviparity. The evidence for this pattern is biogeographical- at high latitudes and elevations, which are cold, a high proportion of species are viviparous. The fly in the ointment is that myself and other scientists think that it is possible that other biophysical variables are important. In fact, most squamates species are found close to the equator, so species at high elevations in the tropics constitute a big chunk of viviparous squamates. After all, oxygen availability decreases with elevation, as does water vapor pressure, and of course both of these variables are correlated with temperature and with each other. Because these are so intercorrelated, isolating the effect of one variably is difficult statistically. However, most papers have stuck with temperature being the only variable that matters, frequently not even considering other variables. Why wouldn’t other variables matter as well? I think it is because some people must think that only one variable could matter, when it seems more plausible to me that whatever drives the evolution of viviparity must be multicausal.
There are many other great examples, such as whether scaling exponents are 0.75 or 0.67 for the metabolic relationship with body size, because the scaling exponent indicates whether volumetric scaling or the fractal nature of nutrient delivery networks drives metabolic scaling. Yet it seems obvious to me that both factors can be important.
So why do we have dogmatic arguments about which factor drives X, without a whole lot of nuance or room for other factors? I am not really sure, but there are several possibilities. Making the case that it is a single, simple factor behind some broad pattern may be appealing to our psychology. Writing papers with a single driver may be more appealing to high-impact journals. And I think it is natural, if you put it out there in your papers that Factor A is most important, that you defend that conclusion.
But I don’t think that is the best way to do science. We should be happy to prove ourselves wrong. A paper is not a static statement of truth. It is a data-based document with our best explanation at the time, which is going to be limited by the tools and information available to us when we wrote the paper. And more generally, I think our literature would be better if we moved beyond simplistic discussions of whether factor A or B matter, and take a more holistic approach that acknowledges the potential of multicausality driving biological patterns.
I should note that there are many counterexamples that do acknowledge complexity and multicausality, and I think most work in most fields is not dogmatic about a single causal driver of patterns. It is also probably true that those examples are harder to remember because they are not beset with drama and conflict. However, I think that this is a pervasive pattern in our literature, even if it is not in the majority.
What do I think should happen in my ideal world? Well, I wish journals, particularly higher impact journals, would not be so biased towards “clean” stories. I think The American Naturalist is a good example of a journal that publishes impactful work, but also tends to acknowledge complexity and potential for multicausality. I think it would be better if (for example), it would not be seen as invalidating or a threat to previous work to acknowledge that things may be a bit more complicated than we originally thought. And I hope that scientists grow more comfortable with thinking about how many biological patterns may be caused by multiple drivers.
In my own little corner of organismal biology, I have noticed this pattern in studies of the evolution of viviparity in snakes and lizards. We don’t need to get into the weeds on this, but the dominant hypotheses for the evolution of viviparity is that viviparous species can regulate temperature of the eggs, while oviparous species cannot (squamates usually have eggs in burrows that are not attended). Hence, selection would favor egg retention, with an endpoint of viviparity. The evidence for this pattern is biogeographical- at high latitudes and elevations, which are cold, a high proportion of species are viviparous. The fly in the ointment is that myself and other scientists think that it is possible that other biophysical variables are important. In fact, most squamates species are found close to the equator, so species at high elevations in the tropics constitute a big chunk of viviparous squamates. After all, oxygen availability decreases with elevation, as does water vapor pressure, and of course both of these variables are correlated with temperature and with each other. Because these are so intercorrelated, isolating the effect of one variably is difficult statistically. However, most papers have stuck with temperature being the only variable that matters, frequently not even considering other variables. Why wouldn’t other variables matter as well? I think it is because some people must think that only one variable could matter, when it seems more plausible to me that whatever drives the evolution of viviparity must be multicausal.
There are many other great examples, such as whether scaling exponents are 0.75 or 0.67 for the metabolic relationship with body size, because the scaling exponent indicates whether volumetric scaling or the fractal nature of nutrient delivery networks drives metabolic scaling. Yet it seems obvious to me that both factors can be important.
So why do we have dogmatic arguments about which factor drives X, without a whole lot of nuance or room for other factors? I am not really sure, but there are several possibilities. Making the case that it is a single, simple factor behind some broad pattern may be appealing to our psychology. Writing papers with a single driver may be more appealing to high-impact journals. And I think it is natural, if you put it out there in your papers that Factor A is most important, that you defend that conclusion.
But I don’t think that is the best way to do science. We should be happy to prove ourselves wrong. A paper is not a static statement of truth. It is a data-based document with our best explanation at the time, which is going to be limited by the tools and information available to us when we wrote the paper. And more generally, I think our literature would be better if we moved beyond simplistic discussions of whether factor A or B matter, and take a more holistic approach that acknowledges the potential of multicausality driving biological patterns.
I should note that there are many counterexamples that do acknowledge complexity and multicausality, and I think most work in most fields is not dogmatic about a single causal driver of patterns. It is also probably true that those examples are harder to remember because they are not beset with drama and conflict. However, I think that this is a pervasive pattern in our literature, even if it is not in the majority.
What do I think should happen in my ideal world? Well, I wish journals, particularly higher impact journals, would not be so biased towards “clean” stories. I think The American Naturalist is a good example of a journal that publishes impactful work, but also tends to acknowledge complexity and potential for multicausality. I think it would be better if (for example), it would not be seen as invalidating or a threat to previous work to acknowledge that things may be a bit more complicated than we originally thought. And I hope that scientists grow more comfortable with thinking about how many biological patterns may be caused by multiple drivers.
RSS Feed