by Brian Tomasik
First published: 5 Jul. 2017; last update: 5 Jul. 2017


In debates regarding whether insects and other invertebrates can feel pain, conflicting evidence can be raised on either side. For example, it seems clear from many studies that invertebrates can learn to avoid electric shock, heat, certain chemicals, and so on. Meanwhile, there are examples of invertebrates apparently unconcerned by physical injury. In my opinion, this collection of evidence suggests that invertebrates plausibly suffer in response to some stimuli but maybe not others. If so, it seems useful to further explore which particular stimuli are unpleasant to invertebrates to what degrees, in order to inform ethical treatment of invertebrates.

Nociceptors vs. suffering

In some discussions of whether insects feel pain, we encounter a common canard: "Since insects lack nocireceptors, they cannot experience pain" (Garcia 2013).

One problem with this statement is that some insects, such as fruit flies, do have nociceptors. Ballenger (2016):

The best studied nociceptor is the Drosophila nociceptor TRPA1, which is pretty closely related [to] a gene in mammals which has an identical function. It senses pressure and heat, and reacts to the chemicals which make peppers hot (capscaicin) and mustard spicy (isothiocyanate). I’m not sure how widespread this protein is, because there have been gains and losses in some nociceptors. It’s also found in beetles, so painless appears to be somewhat widespread.

However, the more important problem with the "no nociceptors => no pain" argument is that it ignores other ways in which sensory inputs can be transformed into suffering experiences. Consider the saying: "sticks and stones may break my bones, but words will never hurt me". Sticks and stones can activate nociceptors, while hurtful words cannot. But obviously words can still give rise to suffering, which is what's morally relevant.

Artificial reinforcement-learning agents illustrate a barebones outline of pain and pleasure learning, yet these agents often don't have any specific structures analogous to nociceptors. Reinforcement-learning agents can receive arbitrary reward signals in response to arbitrary states of the world and actions taken.

Ultimately, "suffering" is the suite of responses that an agent has in response to "negative" events (where the fuzziness of this definition is part of my point), and those responses can be triggered by a variety of kinds of cognitive processing. Meanwhile, nociception is not sufficient for suffering. A nociceptor firing in isolation is just a bare signal being passed along. What matters is what the rest of the cognitive system does with that information.

relaximanentomologist (2013) agrees with me that the issue of insect pain is not as simple as the "no nociceptors => no pain" argument and adds: "nociception is not pain. The current definition of pain requires an emotional response. Humans can feel pain without any physical stimulus and are capable of emotions associated with pain". Likewise, Eisemann et al. (1984) say (pp. 165-66):

The apparent absence from insects of any known or likely candidate nociceptors is of doubtful significance. Whilst it could be argued that the evolution of specific nociceptors accompanies a developing capacity to experience pain states, alternative mechanisms could possibly subserve this function. For example, nociceptive information could be decoded from abnormally high-frequency discharges from non-nociceptive mechano-, chemo-, and thermo-receptors, such as has been proposed for nociception in some mammalian viscera[...].

Stimuli that are plausibly painful to invertebrates

Stimuli such as heat, electric shock, and quinine or salt are commonly used in conditioning studies on invertebrates. For example, all of those punishments and more have been used for learning experiments on fruit flies (Pitman et al. 2009). You can also search for words like "heat" and "shock" in Wikipedia's article "Pain in invertebrates".

Perry et al. (2013) provide an amazing summary of studies on invertebrate learning, of which the following figure (from p. 564) shows a sample. I underlined what appear to be negative stimuli.

Species differences in what's painful

Whether a given stimulus is found painful by a given type of organism depends on the adaptive landscape faced by that organism's ancestors. For example, this page reports: "Diving animals such as mink and burrowing animals, such as rodents and rats, are sensitive to low-oxygen atmospheres and (unlike humans) will avoid them". Ballenger (2016): "Different animals use different receptors to sense different things. Drosophila is repelled by the taste of wasabi, but crayfish are not. Additionally, crayfish do not react to the type of burn that dry ice produces when you get a wart burnt off at the doctor’s office."

Sneddon et al. (2014) give more examples (p. 203):

Whereas mammals find the burning sensation noxious and avoid eating chilli peppers, capsaicin does not activate the avian receptor so birds can ingest these and act as an aid to dispersal of the seeds. [...]

rainbow trout nociceptors are not responsive to cold temperatures below 4 °C (Ashley et al., 2007). This is intuitive since these fish may frequently encounter such low temperatures and it would not be adaptive to perceive them as noxious.

Not all harmful stimuli are painful to humans. Inert-gas asphyxiation provides one example of this point. Mather (2001) provides another (p. 153): "Because nuclear radiation can kill us without our feeling a thing, humans too do not always respond with pain to possible tissue destruction." It would be wrong to conclude from these examples that "humans can't feel any pain".

Is bodily injury painful to insects?

Eisemann et al. (1984), p. 166:

No example is known to us of an insect showing protective behavior towards injured body parts, such as by limping after leg injury or declining to feed or mate because of general abdominal injuries. On the contrary, our experience has been that insects will continue with normal activities even after severe injury or removal of body parts. An insect walking with a crushed tarsus, for example, will continue applying it to the substrate with undiminished force. Among our other observations are those on a locust which continued to feed whilst itself being eaten by a mantis; aphids continuing to feed whilst being eaten by coccinellids; a tsetse fly which flew in to feed although half-dissected; caterpillars which continue to feed whilst tachinid larvae bore into them; many insects which go about their normal life whilst being eaten by large internal parasitoids; and male mantids which continue to mate as they are eaten by their partners. Insects show no immobilisation equivalent to the mammalian reaction to painful body damage, nor have our preliminary observations of the response of locusts to bee stings revealed anything analogous to a mammalian response. Wigglesworth has provided additional examples of insect non-response to treatment which would certainly produce both pain and violent reactions in humans.

Mather (2001), p. 152: "Insects, for instance, can walk normally with a couple of broken-off legs and survive with apparent unconcern as a parasite is eating them up inside, when presumably we would be in excruciating pain."

These anecdotes do seem to suggest that insects are at least much less bothered by certain types of physical injury than vertebrates would be. One unlikely but possible reply to these examples could be that, while injured insects do feel pain, continuing on as normal is the best they can do under the circumstances. This view would be strengthened if it were discovered that physical injury could be used as a punishing unconditioned stimulus in conditioning studies. I'm not aware of any such studies, though it seems like they'd be pretty easy to do. For example, if a bee lost a leg in the presence of a particular odor, perhaps the bee would register that "This is bad, and I should avoid similar conditions in the future, but I should carry on as best I can despite the damage." Of course, it seems more likely—and consistent with Morgan's Canon—that many insects don't have the machinery to register (at least certain types of) physical injury as bad at all.

Insect predation videos

This video shows a male mantis continuing to mate after his head has been bitten off, although he does later try to escape.

However, in my opinion, most videos showing insect predation are more ambiguous regarding pain. For example, this fly being eaten does struggle to escape, as does this locust. Rather than concluding that insects probably don't mind bodily injury, these examples lead me to think that we don't know if insects mind bodily injury.

Eisemann et al. (1984) would probably argue that a fly's attempts to escape a mantis predator may be merely reflexive: "Some insect behavior, such as [...] the struggling of restrained living insects[...], although superficially resembling that of higher animals responding to painful stimuli, no more requires the presence of a pain sense than do reflexive withdrawal responses" (p. 166).

Injured ants

While it provides limited evidence, a Reddit post says:

today I was sitting in my backyard and saw an ant. It was walking across the blanket I was sitting on. So, I picked it up to move it off and realized that I could actually hold it by one of it's limbs/legs. Out of curiosity, I pulled a leg off, wondering if it was actually possible for humans to torture insects by removing theirs legs.

Well, I pulled it off. Then, it started to walk away like nothing had happened. Still curious, I proceeded to pull off each of it's limbs or legs until it was almost unable to walk. Every time I did so it didn't really show any signs of being in pain.

I asked this Reddit person: "What made you think the ant wasn't in pain? Because it kept running away rather than flailing around? For a slower-moving worm, wiggling about might be the best way to escape from a predator, while for an ant, perhaps running away is best. Both might still cause pain—if not immediately then perhaps after the 'heat of the moment' has passed? It's hard to say without more info." The person replied: "It didn't even appear to run away in a traumatized way. It just started to walk away in the same manner it was walking before I had even disturbed it. Though, after I had pulled off several of it's legs, it did seem to move faster."

While I've never pulled the legs off of an ant, I have moved them out of harm's way on many occasions, often by letting them crawl onto a piece of paper. I find that the ants get "scared" when they, e.g., realize that the paper underneath them is moving. They start running several times faster than normal in various directions in an apparent attempt to escape. Likewise, I've seen similar "startled" behavior in many other invertebrates, from worms to spiders. Is this behavior accompanied by long-lasting cognitive changes? For example, can these "fear responses" be used as punishments in classical/operant conditioning? Such studies would help elucidate the extent to which these "fear" behaviors are shallow reflexes vs. how much they're processed in more depth by an invertebrate's nervous system.

On 1 July 2017, I happened to see my housemate accidentally step on a carpenter ant. The stepping wasn't enough to significantly crush the ant, but it apparently caused some kind of injury, because the ant was writhing and running in circles until I put it out of its (apparent) misery by crushing it as hard and thoroughly as I could a few seconds later. In general, my anecdotal experience with insects is that they react to bodily injuries more dramatically than the "insects don't feel pain" articles tend to suggest.

Groening et al. (2017)

Groening et al. (2017) found that injured honeybees didn't consume a greater proportion of morphine relative to sucrose than non-injured honeybees:

Foragers were subjected to two different types of injuries: (i) a clip that applied continuous pressure to one leg and (ii) amputation of one tarsus. The bees were given a choice between two feeders, one offering pure sucrose solution, the other sucrose solution plus morphine. We found that sustained pinching had no effect on the amount of morphine consumed, and hence is unlikely to be experienced as painful. The amputated bees did not shift their relative preference towards the analgesic either, but consumed more morphine and more solution in total compared to intact controls. While our data do not provide evidence for the self-administration of morphine in response to pain, they suggest that injured bees increase their overall food intake, presumably to meet the increased energy requirements for an immune response caused by wounding.

Of course, this null result isn't conclusive. For example, maybe the two types of physical injury tested (both involving the legs) don't cause pain, but other physical injuries do. Or maybe leg injury does cause pain not mediated by morphine. But morphine does seem to reduce responses to other noxious stimuli, as Groening et al. (2017) note: "It has been demonstrated in praying mantis, crickets and honeybees that morphine injection reduces their defensive response to a noxious stimulus in a dose-dependent way and that this analgesic effect can be blocked by naloxone."

Groening et al. (2017) mention regarding their bees: "We observed[...] that some individuals were stepping on their clip and pushing it down with another foot, presumably in an attempt to remove it." So clearly at least some bees could feel the clip and tried to get it off. However, this behavior needn't indicate pain, in a similar way as a human removing a leaf from her hair isn't in pain. Eisemann et al. (1984) might suggest that bees' attempts at clip removal are an extension of grooming behavior? Regarding similar behaviors, they write: "Most such examples can be explained adequately as adaptive behavior patterns elicited reflexly by certain kinds of stimulation; for example, the activation of neural circuits for escape and cleaning behavior" (p. 166).

A chicken self-administration study

Groening et al. (2017) mention by way of introduction: "Danbury et al. showed that lame broiler chickens selectively chose food that contained carprofen as an analgesic, suggesting that they were in pain. The wounded birds consumed more drugged food than sound birds, and the consumption of the analgesic increased with the severity of the injury."

Danbury et al. is also cited by Freire et al. (2008), an examination of beak-trimmed chickens. Freire et al. (2008) say (p. 446): "The pullets beak trimmed at 10 weeks of age pecked more gently than intact pullets, consistent with the hypothesis that these recently beak trimmed pullets were guarding a painful beak from further contact." However (p. 446): "No evidence was found that birds from the four beak trimming methods [including untrimmed controls] consumed different amounts of analgesic-treated feed when both feeds were presented simultaneously". So this appears to be an instance where apparent pain wasn't detected by a self-administration experiment. Hence, it seems possible (if not necessarily probable) that Groening et al. (2017) also failed to detect actual pain in their animal subjects.

Freire et al. (2008) weren't sure why beak-trimmed chickens didn't consume more analgesic-treated feed, but one possibility was "that carprofen had no analgesic effect on the type of pain arising from beak trimming" (p. 447). The authors continue (p. 447):

Support for this latter explanation is provided by Kupers and Gybels (1995). They found that rats with neuropathic pain, induced by partial sciatic nerve injury, did not consume more fentanyl (an opioid analgesic) than control rats. In contrast, rats with nociceptive pain, adjuvant induced monoarthritis, consumed significantly more fentanyl than control rats. They concluded that fentanyl has a good analgesic effect on neuromuscular pain but a poor analgesic effect on neuropathic pain. It may be that carprofen has a similar selective analgesic effect in chickens, in being effective for neuromuscular pain in lameness as reported by Danbury et al. (2000) but having no effect on neuropathic pain following beak trimming in the present study. Unfortunately, the mode of action of carprofen in birds is unknown.

Is bodily injury painful to decapods?

Sneddon et al. (2014) explain (p. 207): "Decapods also show prolonged rubbing or guarding of an affected area as seen in vertebrates (Weary et al., 2006)." Sneddon et al. (2014) cite several examples of this (pp. 207-08). Table 2 (p. 204) of Sneddon et al. (2014) says that (at least some) decapods do show protective behavior in response to injury, while insects don't. If true, this is a bit surprising because insects and decapods are both in the phylum Arthropoda, while other animals that clearly show protective behavior in response to injury are in other phyla: Mollusca (cephalopods)a and Chordata (vertebrates).

Diarte-Plata et al. (2012)

Diarte-Plata et al. (2012), p. 176:

Pain in crustacean decapods is openly debated (Elwood et al., 2009; Elwood and Appel, 2009; Newby and Stevens, 2008; Ponnuchamy et al., 1980; Taylor et al., 2004), but growing evidence indicates that these animals experience adverse sensory effects caused by a stimuli that result in tissue damage (Dunlop and Laming, 2005; Sherwin, 2001; Zimmerman, 1986). In our study, we observed that a trauma caused by injury to [the freshwater prawn] M. americanum yielded behaviors related to pain, including tail flicking as a reflex response to allow escape (Barr et al., 2008) and rubbing the affected area (Stasiak et al., 2003). In addition, five other variables were altered relative to the standard behavior of these organisms: sheltering (Balasundaram et al., 2004; Mariappan and Balasundaram, 2003), disorientation, recoil, and feeding behavior (Newby and Stevens, 2008; Taylor et al., 2004).

Diarte-Plata et al. (2012) reported that an anesthetic could reduce some prawn pain behaviors (p. 177):

Eyestalk ablation and ligating the eyestalk are traumatic procedures that are performed as part of crustacean aquaculture. [...]

The use of Xylocaine as an anesthetic helped to alleviate some responses to the treatment (such as flicking) and had the tendency to improve coping with rubbing, non-sheltering and recoil.

That said, the study used only 10 animals per treatment (p. 173), which makes me think the results may be fairly noisy. Many of the apparent reductions in pain behaviors due to Xylocaine aren't statistically significant, as you can see in Figs. 1-3. The authors acknowledge (p. 177) that "This Xylocaine effect occurred under specific circumstances and not for all stimuli."

Next steps

As Groening et al. (2017) conclude, there's a "need for continued investigation of the possibility of pain in invertebrates." They say: "Future experiments could include the measurement of physiological changes in response to injury, but to distinguish nociception from pain, the greatest insights can be gained by behavioural observations. In particular, avoidance learning or protective motor reactions indicate a more complex response which implies central processing, rather than a simple nociceptive reflex."

One simple approach could be to ask whether a given stimulus (including various forms of physical injury and stress) can be used as punishment in associative-learning experiments. Such studies can be done without fancy technology. Of course, such experiments alone aren't sufficient, and we should explore insect reactions to harmful stimuli from a variety of standpoints.

In the meantime, I think it's prudent to err on the side of caution by avoiding causing physical injury to insects. Such injury may cause some pain-like cognitive changes (we don't know yet). Even if injury isn't directly painful, it may lead to hungerb or other possibly painful states. And physical injury often frustrates the (implicit) preferences of insects. Carruthers (2007), p. 296: "Invertebrates believe things, want things, and make simple plans, and they are capable of having their plans thwarted and their desires frustrated."c

And some stimuli, like excessive heat, seem quite likely to be painful to a wide range of invertebrates. While I can't find the full text of Wigglesworth (1980), Eisemann et al. (1984) say regarding that article: "Wigglesworth has, by inference from his observations of insect behavior, concluded that while most of the manipulations to which insects are commonly subjected probably do not cause them pain, certain stimuli, such as high temperature and electric shocks, apparently do so." This suggests that, e.g., cooking insects alive before eating them is cruel.


  1. Sneddon et al. (2014), p. 207: "squid do not appear to show targeted wound-tending behaviour (Crook et al., 2011), although increased sensitivity and prolonged behaviour directed at the site of a wound has been observed in the octopus Abdopus aculeatus (Alupay et al., 2014)."  (back)
  2. Is hunger painful? Here's one data point, from an extremely simple invertebrate with only 302 neurons. Mori (1999) reports (p. 416) that if C. elegans are well fed, they prefer the temperature that they were in recently, while if they're starved for 2-4 hours, they avoid the temperature they were in recently.   (back)
  3. Note that Carruthers (2007) himself doesn't accept direct moral obligations to invertebrates (p. 296): "it is a fixed point for me that invertebrates make no direct claims on us, despite possessing minds in the sense that makes sympathy and moral concern possible."  (back)