Might Intense Agony Be Widespread?

Published on July 30, 2026 3:21 PM GMT Humans are capable of experiencing very intense agony. But how widespread is this capacity in the animal kingdom? How many animals can experience intense pain? In this essay, I will argue, from both theoretical and experimental considerations, that capacity for feeling intense agony might be quite widespread. It may be that even the simplest creatures capable of suffering can suffer intensely. This essay was largely based around a piece I entered in an essay contest, hence the higher-effort citations. 1 Introduction Humans can experience quite intense agony. If a human is stabbed, set on fire, or drowned, they suffer a great deal. But are other creatures capable of similarly intense agony, and if so, which ones? This is a question that has been surprisingly neglected, despite considerable research into which animals feel pain. Here, I will argue for what can be called the widespread agony thesis. This thesis simply holds that the capacity to feel intense pain is quite widespread, being present in crustaceans like shrimp and lobsters; insects like ants, flies, and bees; and fish. If this thesis is correct, humans constitute only a tiny fraction of the conscious organisms capable of intense pleasure and pain. Intense pain could be present in every species whose consciousness is seriously debated. This thesis is quite important if true. Humans kill around 25 trillion wild shrimp every year for food and farm about 440 billion (Waldhorn & Autric, 2023). About a trillion insects are farmed and killed annually (Rowe, 2020). Tomasik (2016a) estimates that the average human prevents about 14 million years of insect life through their annual environmental impact. If the creatures we affect in staggering numbers are capable of very intense pain, that has quite significant implications. Indeed, it may mean that the impact that we have on the suffering of non-humans far exceeds the impact that we have on the suffering of humans. I do not claim that this thesis is overwhelmingly likely. It could, for all I know, be false. But I claim that given the huge implications that this thesis has if it is true, it deserves quite a bit more investigation. So long as the thesis is merely plausible—which I shall argue it is—this has quite significant ethical implications. This article will have several sections. Section 2 will present general considerations favoring the presence of ubiquitous intense pain in animals and address general arguments against the notion that intense pain is widespread. Section 3 will provide a case study, analyzing evidence of intense pain in fish. Section 4 will analyze evidence of intense pain in crustaceans. Section 5 will analyze evidence of intense pain in insects. Then, section 6 will conclude. 2 General considerations Typical practice towards shrimp, fish, and insects seems to take for granted that they do not suffer very much. Few think that when they swat a fly, there is a remote possibility that the fly experiences anywhere near the amount of pain that a human would from being crushed to death. Yet what justifies this confidence? In my view, there are several independent considerations which make this confidence unwarranted. The first concerns our general uncertainty about consciousness. We do not have an agreed upon theory of consciousness. There isn’t even agreement about whether consciousness is physical or non-physical (Schneider & Velmans, 2017). Consciousness is arguably the thing that science has given us the least insight into, and some have even daringly proposed that we will never solve the mystery of consciousness (e.g. McGinn, 1989). This uncertainty gives us some reason to be suspicious of very low estimates of animal consciousness. As an analogy, if we do not know much about some particular class of aliens, we should not confidently declare them to be very small. Similarly, if we are highly uncertain about consciousness—as, indeed, we are—that is good reason to not be very confident that various creatures only suffer mildly. This analysis is somewhat complicated by two-envelope questions. I plan to write a post at some point explaining why I don’t think that those considerations, but for now, see St. Jules, 2024. Our uncertainty should also make us suspicious of arguments that proceed from a specific theory—say, the integrated information theory of consciousness—to specific judgments about the pain-capacity of animals. The standard theories of consciousness are largely silent about how intensely animals suffer and even whether creatures with very different brains from ours are conscious (see Birch, 2022). In light of this and our high degree of uncertainty about which theory of consciousness is correct, it would seem premature to let this majorly influence our judgments about the intensity of valenced experience in animals. More general views of consciousness like dualism and panpsychism don’t have any direct implications on the intensity of suffering in non-humans. The second argument is behavioral: even simple animals, when injured, behave as if they’re in a lot of pain. Suffocating fish thrash around with great force. Lobsters being boiled alive struggle, rather as one might expect you or I to do if we were boiled alive, as David Foster Wallace (2005) famously noted in Consider The Lobster . Lobsters show heightened neural activity for quite some time after being boiled (Birch et al., 2021)—once again, as one might expect if they were in pain. Though we typically don’t notice because they are small, insects often behave similarly (Qu et al., 2022). If a creature behaves dramatically, as would be expected if it were in intense pain, we should think it’s reasonably likely that the creature does, in fact, feel intense pain. If we came across aliens struggling and thrashing when boiled, it would be reasonable to think with decent odds that they feel quite intense pain. If this is so, we should accept a similar inference with regards to simple animals. A third consideration is evolutionary: there is every evolutionary reason for simple creatures to feel intense pain. Pain in humans causes them to avoid various behaviors that are likely to be harmful. Pain in fish, insects, and decapods serves similar functions. We should expect intense pain in these creatures to be adaptive for the same reason that pain in humans is adaptive. If a creature just felt mild discomfort, that would be unlikely to sustain the kind of behavior needed to safeguard survival. Intense pain could be adaptive for eliciting rapid and dramatic behavior. For this reason, we should expect evolution to take various shortcuts to allow even simple creatures to feel rather intense pain. Thus, if it’s in principle possible for simple creatures with few neurons to feel intense pain, then it’s decently likely that actually-existing simple creatures feel intense pain. Given that humans can feel very intense pain despite only a small share of our neurons being responsible for its production, we should expect a brain efficiently geared towards producing pain in response to noxious stimuli to efficiently perform such a feat. A fourth reason simple creatures might suffer intensely has been proposed by Dawkins (2011): the purpose of pain is to teach a creature a lesson, to get it to avoid certain activities. If a creature is simpler and has less great cognitive capacities, it might need more pain to teach it a lesson. It may be that intense pain is needed for a creature to remember its agony in the long-term. Only intense pain would leave a lasting impression on such a simple creature. This would also accord with findings that simple creatures like crabs (Fernandez-Duque et al., 1992), fish (Sneddon, 2015), and insects (Giurfa, 2015) can remember unpleasant experiences and avoid them. A fifth argument for pain being intense in simple creatures: perhaps pain intensity is more a function of the share of an organism’s cognitive architecture spent on pain, rather than the absolute processing power spent on pain (Tomasik, 2016b). For instance, imagine a creature with only 10,000 neurons, but all of those neurons were spent processing pain signals. Plausibly, such a creature would feel pain more intensely than a creature with 10 million neurons, where only 10,000 neurons were spent on feeling intense pain. For the first creature, pain would occupy its entire consciousness, while for the second it wouldn’t. The most intense pain occurs when pain occupies the entire conscious field of the sufferer, like during brutal torture, so that they can focus on nothing else. If simple creatures routinely have almost their entire cognition taken up by pain, then perhaps their pain is of similar unbearable intensity. To give an analogy from Merker (2016), the mechanism of replication works as well in DNA with short sequences of base pairs as with long and complex sequences. Perhaps the neural structure underlying consciousness is similar, and can be produced with significant intensity despite fairly simple neural architecture. Or to give another analogy, while simple creatures likely tend to have less developed higher-order processing related to vision, they still have the ability to see objects with decent proficiency. Perhaps conscious abilities are similar—simple conscious creatures can feel reasonably intense pain states, just as they can have reasonably developed visual faculties. A sixth consideration: historically, humans have underestimated sentience in animals. Prior to 2002, when nociceptors were discovered, it was widely believed fish had no nociceptors (Sneddon, 2015, p.968). Throughout the 1980s, it was widely thought that animals weren’t conscious, and veterinarians routinely ignored their distress. Similarly, it was widely thought that newborns were not conscious (see Anand & Hickey, 1987 as well as Birch, 2024, p.194). If we notice a widespread pattern of error, we should worry that we are making the same error. This gives us a reason to suspect that we underestimate the consciousness of simple creatures. If consistently humans have underestimated the consciousness of things different from us, that should lead us to suspect that we might be underestimating the consciousness of animals (particularly when the main reason people think simple creatures aren’t intensely conscious is from direct intuition). A seventh argument (see Tomasik, 2016b) is a variant of the argument from marginal cases. The argument from marginal cases was popularized by Singer (1975) as an argument for taking animals seriously. The core idea is as follows: suppose there’s some trait used to argue that animals should not be afforded significant moral status. If we imagine a human with that trait, then presumably we wouldn’t deny them moral status. For instance, if it is suggested that animals are on the menu because of their diminished mental capacities, this would seem to imply that humans with similarly diminished mental capacities (babies, for instance) were similarly morally unimportant. But clearly, this is an unacceptable result! A similar argument can be employed against many of the objections to significant pain in animals. Animals have relatively simple brains. But if we came across a human with a comparatively simple brain, who seemed to respond aversively to external damage the way animals do, we would be hesitant to be confident that its conscious experience is only mild. We should have similar skepticism directed towards arguments for such conclusions about animals. The eighth and final argument for taking seriously the suffering of simple creatures: the most detailed report to date which attempted to estimate intensity of valenced experience in animals guessed that even simple creatures were, in expectation, intensely conscious. This report came from the team of Fischer et al. (2024). Using a mixed function taking into account various different behavioral proxies, the final estimate (p.233) was that shrimp suffer 8% as intensely as humans and carp 24% as intensely. The median estimate was that shrimp suffered 5% as intensely as humans. Using a more integrated model that took into account even more inputs (see Duffy, 2024, table 2) their mean estimate was that shrimp suffered a whopping 19% as intensely as humans, bees 14.8%, and carp 17.5%. The median estimate from that other model was 7.1% for bees, 8.1% for carp, and 2.9% for shrimp. And while the authors emphasize that you shouldn’t take the numbers super literally or treat them as highly precise, they do provide rough order-of-magnitude estimates. The methodology for the report was as follows: they first tried to ascertain the evolutionary function of pain. There are three main theories of the function of pain (Schukraft et al. 2024, p.164): Pain functions to enable organisms to learn. If some experience is pleasurable, an organism repeats it. If the experience is painful, the organism learns to avoid it. Pain serves as a common currency. When comparing different actions, pain can give a single axis along which to compare them. Pain might disincentivize behaviors that are bad for fitness. If an organism dies, for instance, that will be bad for its fitness, so evolution makes death painful. On each of these theories, there should be various proxies of pain. For instance, if an organism made motivational trade-offs between pain and reward, then that seemed to indicate that the organism had a cognitive representation of painful stimuli. If an organism demonstrated greater associative learning in response to painful stimuli or greater anxiety after a painful stimulus, that seemed to be behavioral evidence that it was in more intense pain. If, for instance, pain serves as a common currency, then we should expect organisms with greater associative learning to feel more pain, because their brain more vividly represents external damage. Taking into account a range of these behavioral proxies, as well as various proxies for the brain’s information-processing capacity, they compiled a series of estimates of the intensity of valenced experience in animals. Now, of course, these proxies were far from perfect. But this is by far the most detailed estimate of animal consciousness to date. In light of this, we shouldn’t find results broadly consistent with its findings to be wildly implausible. There are various considerations that might lead one to be skeptical of this thesis. A first argument purports to show that these animals aren’t conscious at all. Brian Key (2015) and James Rose (2002) have suggested that fish (and this argument would generalize to crustaceans) are not conscious because they don’t have a cortex. Key and Rose argue that a cortex is needed for human pain and that because other animals lack a cortex, we should assume they do not feel pain. The literature has not been kind to the Key and Rose thesis (for criticisms see, among others Braithwaite & Droege, 2016; Broom, 2016; and Dinets, 2016). First, many have argued that a cortex isn’t needed for pain in humans or other mammals (see, e.g., Merker, 2007). Panksepp et al. (1994) show that rats without a cortex go on to play mostly normally and display seemingly conscious behavior. Such a theory would also imply that octopi aren’t conscious, despite their complex behavior, which includes a tendency to play (Kuba et al., 2003) and likely dream (Godfriey-Smith, 2017). Octopi also behave, in various ways, as if they’re in pain (see Birch et al., 2021). Second, even if a cortex is responsible for human pain, there could very well be other structures that give rise to pain in other creatures. Arms and legs are needed for humans to swim, but this doesn’t mean fish can’t swim because they don’t have arms and legs. In fish, for instance, the hippocampal homologue might sustain consciousness (see Zacks & Jablonka, 2023). Similar things may be true of the mushroom body in insects (see Heisenberg, 1998) and the hemiellipsoid body in decapod crustaceans (see Crump et al., 2022). For these reasons, it seems clear that one should have significant doubt that the cortex is needed for consciousness. But if one is uncertain in this thesis, then it does not significantly undermine the probability that simple animals suffer intensely. A second reason people often doubt that simple animals feel intense pain is that they have very few neurons (see, e.g. Budolfson & Spears, 2020 for discussion of this method). The argument goes that because animals have vastly simpler brains than humans, with many fewer neurons, they are probably not very conscious. Now, it’s unclear exactly why one would be confident that few neurons leads to mild pain. As discussed before, intense pain might be the sort of thing that doesn’t take many neurons to maintain, just as a basic visual system doesn’t take many neurons to maintain. Such an argument would seem to have odd consequences. Elephants have more neurons than humans, though not more impressive cognitive capacities (see Hart et al., 2008). Despite this, presumably they do not have more intensely valenced conscious experience. And when humans lose many neurons, so long as their brains perform similar functions, we don’t normally think they are vastly less conscious. Shriver (2024) provides a number of reasons to be skeptical of neuron count proxies. The biggest challenge is that it doesn’t fit with evidence from neuroscience (p.114). Neuroscientists haven’t identified anything like a robust correlation between neuronal activation and intensity of valenced experience. In various cases (p.116) there was an inverse correlation between intensity of valenced experience and thickness or activation level of brain regions. The more significant effect comes not from the number of neurons but instead which patterns of neurons are active—thus, in simple creatures, so long as their brains execute the right sorts of functions, experience could be intense. Indeed, neuronal activation is not empirically adequate to settle the question of which of two people has more intense pain (p.125). Shriver additionally notes (p.119) that it would be quite strange if merely adding extra neurons without changing the broad functions they performed increased valenced experience. If redundant neurons were added that didn’t change which functions were performed, it would be surprising if this increased the intensity of valenced experience. Any amount of neuronal activation can, in principle, produce precisely identical behavior and computations (p.127). For this reason, it would be odd if neuron count on its own, without producing new functions or different behavior, significantly affected the intensity of valenced experiences. Neuron counts might particularly underestimate small minds, which have advantages over large minds in terms of computation per neuron (Chittka & Niven, 2009). For this reason, perhaps a more appropriate measure of intensity of experience is the natural log of the number of neurons. But such a method typically turns up much more modest results. Log₁₀(86 billion)/log₁₀(100,000) ≈ 2.19. Another reason one might doubt that simple animals can suffer intensely is that they are incapable of higher-order reflection on their pain. They cannot appreciate the injustice of their suffering, nor can speak about it. But babies cannot do these things either, and few think that babies are incapable of intense pain. While higher-order reflection often makes pain worse, the mere fact that a creature is incapable of it should not lead us to doubt they feel intense physical pain. There is one last reason to suspect simple animals suffer less intensely than more complicated animals. What we know of evolutionary processes is that they tend to begin at low levels, and then expand in more complicated creatures. For instance, various creatures have extremely rudimentary eyes that are little more than light censors. It would be surprising