This post is a short summary of The Era Beyond Eisemann et al. (1984): Insect pain in the 21st century, a peer-reviewed, open-access publication on insect welfare in the Quarterly Review of Biology. The paper and supplemental information can be accessed here. The original paper was written by Meghan Barrett and Bob Fischer; the research conducted in the paper was funded by Rethink Priorities.
This post was written by Abraham Rowe and reviewed by Meghan Barrett and Bob Fischer. All information is derived from the original publication, and some text from the original publication is directly adapted for this summary.
Introduction
In 1984, Craig Eisemann and several co-authors published Do insects feel pain? — A biological view in Experientia. The paper and its arguments have become the primary evidence presented that insects do not feel pain. It’s been cited hundreds of times, continues to be cited to this day, and is influential in and beyond entomology.
Specifically, Eisemann et al. examine evidence for pain, defined as distinct from nociception. While nociception is the ability to detect and respond to noxious stimuli, pain is the aversive, subjective experience that accompanies many noxious stimuli in humans and, plausibly, many other animals. Eisemann et al. claim that insects probably don’t experience pain because: 1) they lack nociceptors (sensory cells that detect noxious stimuli); 2) the presence of endogenous analgesic opioid peptides can be explained by other mechanisms; 3) they have small and simple brains; 4) they don’t protect injured body parts; 5) they behave normally after traumatic injuries; and 6) they have sufficiently inflexible behaviors such that pain would not be adaptive.
Barrett and Fischer, the authors of The Era Beyond Eisemann, show that the empirical evidence that has been discovered since the publication of Eisemann et al. undermines most of its arguments. Further, they contend that some of the philosophical assumptions in Eisemann et al. are hard to defend. Barrett and Fischer do not present a positive argument that insects do feel pain; instead, they point out the errors in Eisemann et al. and call for a more careful evaluation of this topic based on all the developments since 1984.
Evaluating the claims in Eisemann et al.
Do insects have nociceptors?
Nociceptors are cells that respond to noxious stimuli and, in mammals, are essential to the perception of noxious stimuli (a prerequisite for experiencing those noxious stimuli as painful). As of 1984, nociceptors had not been found in insects. Further, Eisemann et al. wrote that even if they were discovered in insects, they might only be present to enable reflexive behaviors in response to noxious stimuli.
Eisemann claim: Insects lack nociceptors, and if insects do end up having nociceptors, they aren’t necessarily evidence of pain sensation.
Barrett and Fischer response: Nociceptive ion channels that are homologous to mammalian nociceptors have been identified in multiple insect species. These ion channels are found in cells—nociceptors—that are morphologically similar to mammalian nociceptors and mediate the perception of noxious stimuli in insects. While Barrett and Fischer agree that nociceptors aren’t strong evidence on their own of pain sensation, they are evidence of its possibility.
Do endogenous opioid peptides and their receptor sites in insects indicate the capacity for pain perception?
In 1984, endogenous opioid peptides with putatively analgesic functions had recently been discovered in insects. These peptides help modulate pain in mammals. However, Eisemann et al. argued that the mere presence of these peptides and receptor sites doesn’t indicate pain experience because they could also have other functions (and may have other functions in mammals).
Eisemann claim: The presence of endogenous opioid peptides and their receptor sites in insects does not necessarily indicate the capacity for pain perception.
Barrett and Fischer response: In fact, further genetic evidence suggests insects do not have an endogenous opioid system with analgesic functions; this system only evolved in vertebrates. Because this system only evolved in vertebrates, it is tricky to evaluate the studies on which Eisemann are basing their claim, which are largely pharmacological interventions demonstrating analgesic behavioral responses of insects to exogenous opioid delivery, and the counteraction of that analgesic effect when insects were given naloxone. There is likely an as-of-yet unknown mechanism that causes the appearance of endogenous, analgesic opioid peptides and their receptors in insects, as documented in these older studies. For instance, they may have a different type of receptor that is responsive to these exogenous, putatively analgesic opioids.
However, endogenous opioid peptides and receptors are not the only way that mammals regulate pain. It is possible that if insects experience pain, then other mechanisms provide a similar analgesic effect as the one that opioid systems provide for mammals. For instance, GABAergic neurons are known to mediate insect responses to chronically “painful” stimulation, and the exogenous delivery of GABA agonists results in an analgesic effect for which there is much greater mechanistic understanding.
Beyond the shift in scientific understanding about opioids in insects, there is a separate problem with Eisemann et al.’s argument here. Eisemann et al. claim that even if these peptides and receptors are present, it isn’t an indicator of the capacity for pain perception. This claim relies on a 1902 comparative animal psychology precept known as Morgan’s Canon: In no case is an animal activity to be interpreted in terms of higher psychological processes if it can be fairly interpreted in terms of processes which stand lower in the scale of psychological evolution and development.
Barrett and Fischer point out that this standard is unreasonable if it is interpreted in a way that confuses possible explanations for best explanations. And that’s how Eisemann et al. interpret it. Again, insects have multiple potential analgesic systems, some of which they share with distant or closely related species. The fact that the presence of any one of them could be explained without positing the capacity for pain doesn’t mean that that’s the best explanation, since that would involve a more complex hypothesis where all those potential analgesic systems have other, unrelated functions despite their phylogenetic relatedness. Granted, this point doesn’t show that these systems do modulate pain; they just add to the case that pain is possible. Still, this point reminds us that we need to assess the possibility of insect pain based on a comprehensive understanding of the evidence, rather than considering individual pieces of evidence in isolation.
Do insects have simple brains with relatively few neurons? If so, does that indicate they cannot generate the experience of pain?
Eisemann claim: Because insects have simple brains with relatively few neurons compared to mammals, they cannot generate the experience of pain.
Barrett and Fischer response: There is some evidence from vertebrate species that different brain structures can lead to pain experiences (for instance, the neocortex in mammals vs. the pallium in birds). There is even some evidence within a singular brain type (such as a human brain) that there can be multiple mechanisms through which painful experiences occur (e.g., in humans, the cortex is often viewed as necessary for pain experience, but some studies have demonstrated pain can occur without it). So, we shouldn’t think that insects not having the same brain regions as mammals makes pain implausible, given the multiple realizability of many functions in animals.
Additionally, insect brains are not necessarily smaller or less complex than the smallest vertebrate brains. The smallest mammal brain known (Etruscan Shrew) is only around 5.5 times the mass of the largest insect brain studied to date (tarantula hawk wasp). The smallest known vertebrate brain (Algerian sand gecko) is around the same mass as that species of wasp\ and contains an identical number of neurons (~1.8 million). There is also good reason for thinking that we may discover even larger insect brains (by both mass and neuron number): there are insects with body masses 50-100 times larger than that of the tarantula hawk wasp, and basic allometric scaling principles would suggest they may have larger brains (whether by mass or by neuron number).
Finally, Barrett and Fischer argue that brain size broadly should not be taken as a measure of a capacity, such as the capacity for pain, pointing to evidence that brains of varying sizes are able to accomplish certain tasks equally well. Moreover, some behaviors, such as reversal learning, can even occur more quickly in smaller-brained animals. Further, leaning heavily on neuron counts and brain size has unusual implications. For example, elephants have around three times as many neurons as humans, but it would be odd to conclude they are much more likely to experience pain than humans are. In general, Barrett and Fischer concur with other authors who have written previously that bigger brains are not necessarily better brains. Instead, many larger brains may mostly be composed of larger numbers of repeating modules that are necessary to maintain larger bodies (without changing the capacities of the animal overall).
Do insects protect injured body parts?
Eisemann et al. claim that “protecting” behaviors, where an animal takes steps beyond reflexive avoidance or withdrawal to prevent reinjury or otherwise protect a wound, is important evidence of pain experience. An example of protecting behaviors in vertebrates might include limping, where an animal changes its gait to reduce the pain from a leg injury, or grooming an injury site.
Eisemann claim: Insects do not protect injured body parts.
Barrett and Fischer response: Some insects do protect injured body parts. Crickets have been documented declining to mate or feed after injury, instead grooming the site of injury. Some studies suggest that in mantids, motivation to mate declines immediately after injury. Cockroaches may groom the site of a puncture wound on their abdomen. There is now evidence that some insects alter their gait in response to injury, potentially to reduce the weight load on an injured area. Further, physical injury is only one kind of damage animals may experience as painful. Bumblebees respond to heat damage (from a heated probe) with site-specific grooming of the injured antenna. Mantises respond to the injection of noxious chemicals from the stinger of a velvet ant by releasing the velvet ant and grooming the specific injury site.
Do insects behave normally after severe injury?
Eisemann et al. claim that insects will continue their behavior unchanged after a severe injury. They give examples of male mantises who continue to mate as they are eaten by their mating partners, ants who walk “normally” on a crushed leg, insects that cannibalize their own open wounds, and insects who do not change behavior as they are consumed by large internal parasites.
Eisemann claim: Insects often behave normally after severe injuries.
Barrett and Fischer response: First, Eisemann’s evidence for this claim is entirely anecdotal, and not based on a quantitative, methodologically rigorous survey of insects’ responses to different injuries. Quantitative analysis might indicate that many of these behaviors are actually unusual and shouldn’t be assumed to be taken as generalizable evidence of insect behavioral responses to pain. For instance, one study found that only 0.63% of black soldier fly larvae that received a grievous injury attempted any auto-cannibalism of the injury site, suggesting the behavior is uncommon among insects. As another example, male mantis cannibalization during mating only occurs in a minority of mating pairs in the wild, and there is evidence that male mantises are cautious about which females they approach for mating, trying to avoid females that appear hungrier (and thus might be more likely to cannibalize them). And, in species where cannibalization occurs much more frequently, males often fight back.
Second, Eisemann et al. don’t consider alternate explanations: we regularly see injured vertebrates who, due to context-specific motivations or trauma-induced physiological changes, do not change their behavior in response to injury. For example, injured humans in a crisis might continue behaving normally for an extended period, and injured prey animals are known to hide injuries (e.g., walk without a limp) to avoid appearing vulnerable to predators. Strange behaviors in insects that appear contra-adaptive might actually have an unexplained adaptive function.
Nevertheless, it is clearly the case that insects don’t always respond the same way to injuries as vertebrates and shouldn’t be viewed as “miniature mammals.” However, because insect bodies are physiologically very different from vertebrate ones, we should expect their behavioral changes in response to injuries to also be different. For example, mammals with a leg injury might limp to reduce the load on that injury. However, for animals with exoskeletons, it isn’t obvious that the same protective response is required. To date, minimal data have been published on the kinematics of insect injury and response; however, understanding the biomechanics of injury in insects will be essential to contextualizing when behaviors would actually be protective.
Finally, Eisemann mainly discusses insects’ behavioral responses following mechanical injuries to the exoskeleton. But insects have been shown to have behavioral responses to many non-mechanical injuries, such as electrical shocks or heat. In response to these injuries, their behavior is often much more like what we’d expect in mammals—they immediately recoil, attempt to escape, and often alter their behavior long-term to avoid further noxious stimulation (e.g., via learning). In fact, these responses are so consistent and widespread that they serve as the basis of an entire field of aversive learning paradigms in insects. Looking at just mechanical injuries with only anecdotal evidence misses the significant evidence that insects do alter their behavior in response to many injuries.
Is insect behavior sufficiently inflexible such that pain would provide little adaptive value?
Eisemann et al. suggest that having plastic behavior (i.e., learning and changing responses) in response to painful stimuli is important evidence for pain sensation and that insect behavior is not flexible in response to injury.
Eisemann claim: Insect behavior is inflexible in a way that suggests that pain would have little adaptive value for them.
Barrett and Fischer response: We now have significant evidence that many insects do have plastic behaviors, including in response to painful stimuli.
Fruit flies have demonstrated a higher willingness to experience shocks for different rewards—they’ll cross a shocking barrier with a higher voltage for ethanol than they will for sucrose. A 2022 paper found that in five orders of insects (out of six studied), there was strong evidence that insects learned associations with painful stimuli. For example, fruit flies learned to avoid shapes paired with heat exposure, associating the visual cue with the negative stimulus despite the shape being harmless by itself. This learning occurred even when researchers introduced distracting stimuli during the training process. Additionally, while associative learning has been demonstrated in plants and unicellular organisms, insect learning appears to be significantly more flexible—including reversal learning, trace conditioning, and delay conditioning. There is now sufficient evidence that insects have flexible behavior such that pain could play an adaptive role.
Conclusion
Barrett and Fischer argue that Eisemann et al.’s claims are now generally inaccurate despite still being widely cited in the scientific literature and beyond. They are clear that their arguments are not meant to suggest that insects are capable of feeling pain. Instead, they mean to suggest only that the question is still open and demands further empirical and philosophical research.
Finally, Barrett and Fischer point out that people often ignore one major conclusion of Eisemann et al.: that a precautionary approach to insects is warranted even given the Eisemann et al.’s skepticism about insect pain. Barrett and Fischer agree with this recommendation that we treat insects cautiously until we know more about their status as moral patients.
Acknowledgements
This research is a project of Rethink Priorities. It was written by Abraham Rowe, summarizing work by Meghan Barrett and Bob Fischer. Thanks to Meghan Barrett and Bob Fischer for helpful feedback and Shane Coburn for copy editing. If you like our work, please consider subscribing to our newsletter. You can explore our completed public work here.