Hi Dawn, thanks for this! I like the approach of letting people choose their assumptions at the start.
Ng argued that evolutionary dynamics – particularly the r-strategy of producing far more offspring than survive – imply that suffering probably dominates happiness in nature.
In this paper, we find an error in Ng’s model that, when fixed, negates the original conclusion. Instead, the model offers only ambiguity as to whether suffering or enjoyment predominates in nature.
But I’m not a fan of Ng’s approach to this. His original argument was based on a weird and very crude assumption that if you don’t manage to reproduce, evolution would probably give you more negative experiences than positive ones. And this just doesn’t have to be true. I think Tomasik made the case for net-negative welfare better, but Browning & Veit present some pretty good reasons to be doubt that death will outweigh the other potentially positive experiences in wild animals’ lives.
I think most of those working on WAW are now tending to agree that we don’t know enough to make this judgement with any degree of confidence. Given how little we know about the experiences of most wild animals, this is just a hard call to make. For that reason, interventions that don’t depend on this assumption tend to be favored. But I’m wondering if you have particular reasons for favoring the net-negative view?
Ohhh! I remember skimming that paper ages ago but I didn’t make the connection! Probably because I forgot the name of the author… But that has been very fruitful to think about. In particular I’m surprised I hadn’t made the connection between plants and nematodes – both fairly unable to react to harmful stimuli, much less so than a fly, so pain energy is much more likely wasted for them!
What do you think about this addition to the article?
While Groff and Ng’s revised mathematical model elegantly demonstrates that high infant mortality does not automatically guarantee a predominance of suffering, the model relies on abstract “evolutionary economics” that may oversimplify the gritty realities of biology. A primary critique is that the model assumes evolution will readily dial down the intensity of nociception (pain) to conserve biological resources when an organism is statistically likely to fail. However, because natural selection operates at the level of the individual gene, it is entirely possible that maintaining an intense, energy-demanding pain response is highly adaptive if it marginally furthers that specific individual’s survival. Furthermore, for organisms with extremely simple nervous systems consisting of only a few hundred neurons, the metabolic savings of “cheapening” pain receptors are likely negligible. It is arguable that evolutionary pressures would favor much more straightforward energy-conservation strategies – such as lethargy, cannibalism, or producing fewer offspring – long before micromanaging the nuances of neuronal pain processing. The authors explicitly acknowledge this vulnerability in their model, conceding that the actual biological “cost” of suffering remains entirely unknown; they hypothesize it could be metabolic (e.g., glucose for neurotransmitters), structural, or perhaps not physical at all, but rather the evolutionary hazard of being dangerously distracted by intense pain.
A second major critique centers on the model’s failure to account for an organism’s physical agency and behavioral flexibility. The model posits that if an animal is highly likely to die young, intense pain is a wasted biological investment. This logic holds up for immobile or slow-moving organisms, where registering intense pain offers no actionable escape mechanism. However, for highly mobile creatures with rapid reaction times – such as flies – intense pain or stress is a profoundly profitable evolutionary investment because it instantly triggers life-saving evasion maneuvers. In these high-agency animals, the capacity for intense suffering is precisely the mechanism that keeps them alive, meaning evolution would vigorously preserve it regardless of the species’ overall mortality rate.
Fortunately, Groff and Ng anticipate this limitation, noting that theories linking the evolution of emotions directly to “flexibility in behavior” are highly relevant and more amenable to objective study. They further concede that if the primary evolutionary advantage of negative affective states is to force an organism to “focus” on actionable threats, rather than acting purely as a behavioral reinforcement mechanism, their equation based on reproductive failure rates becomes less informative. Ultimately, the authors agree that a single abstracted equation cannot capture the full picture of wild animal welfare, concluding that a combination of mixed, empirical models will be necessary to truly understand how different species experience nature.
But that’s very zoomed in. The dad of an acquaintance of mine has had some 15–20 children. They don’t know each other and can barely estimate the numbers. The ones my acquaintance knows about have every personality disorder and substance addiction under the sun and are in and out of prison all the time. Even if someone should find out that in aggregate their lives are net positive because at least some of them have net positive lives, that would still not be enough for me to advocate for this kind of profligate behavior.
Hi Dawn, thanks for this! I like the approach of letting people choose their assumptions at the start.
Are you aware that Groff & Ng subsequently published a correction to this claim? Specifically:
But I’m not a fan of Ng’s approach to this. His original argument was based on a weird and very crude assumption that if you don’t manage to reproduce, evolution would probably give you more negative experiences than positive ones. And this just doesn’t have to be true. I think Tomasik made the case for net-negative welfare better, but Browning & Veit present some pretty good reasons to be doubt that death will outweigh the other potentially positive experiences in wild animals’ lives.
I think most of those working on WAW are now tending to agree that we don’t know enough to make this judgement with any degree of confidence. Given how little we know about the experiences of most wild animals, this is just a hard call to make. For that reason, interventions that don’t depend on this assumption tend to be favored. But I’m wondering if you have particular reasons for favoring the net-negative view?
Ohhh! I remember skimming that paper ages ago but I didn’t make the connection! Probably because I forgot the name of the author… But that has been very fruitful to think about. In particular I’m surprised I hadn’t made the connection between plants and nematodes – both fairly unable to react to harmful stimuli, much less so than a fly, so pain energy is much more likely wasted for them!
What do you think about this addition to the article?
But that’s very zoomed in. The dad of an acquaintance of mine has had some 15–20 children. They don’t know each other and can barely estimate the numbers. The ones my acquaintance knows about have every personality disorder and substance addiction under the sun and are in and out of prison all the time. Even if someone should find out that in aggregate their lives are net positive because at least some of them have net positive lives, that would still not be enough for me to advocate for this kind of profligate behavior.