Wednesday, 19 September 2012

Do Insects Feel Pain?

As you get older, do you sometimes find yourself standing in the kitchen wondering what it was that you were looking for? Do you find yourself struggling to remember the name of that movie, the one with that guy in it, where something happened at the end? When constructing your web to catch your food, do you find you are making more errors in terms of the length of the capture spiral, the number of anomalies per cm, and in four parameters of web regularity?

Okay, maybe the last example applies more to spiders than humans but researchers have recently looked at the behavioral effects of ageing on spiders and found similar patterns of cognitive decline that are often observed in humans and other "higher" animals.

Ageing alters spider orb-web construction (M. Anotaux, J. Marchal, N. Châline, L. Desquilbet, R. Leborgne, C. Gilbertb, A. Pasquet):
ABSTRACT: Ageing is known to induce profound effects on physiological functions but only a few studies have focused on its behavioural alterations. Orb-webs of spiders, however, provide an easily analysable structure, the result of complex sequences of stereotypical behaviours that are particularly relevant to the study of ageing processes. We chose the orb spider Zygiella x-notata as an invertebrate organism to investigate alterations in web geometry caused by ageing. Parameters taken into account to compare webs built by spiders at different ages were: the length of the capture spiral (CTL), the number of anomalies per cm, and four parameters of web regularity (the angle between radii, the number of spiral thread units connecting two successive radii, the parallelism and the coefficient of variation of the distances between silk threads of two adjacent spiral turns). All web parameters were related to ageing. Two groups of spiders emerged: short- and long-lived spiders (with a higher body mass), with an average life span of 150 and 236 days, respectively. In both short- and long-lived spiders' webs, the CTL and the silk thread parallelism decreased, while the variation of the distances between silk threads increased. However, the number of anomalies per cm and the angle between radii increased in short-lived spiders only. These modifications can be explained by ageing alterations in silk investment and cognitive and/or locomotor functions. Orb-web spiders would therefore provide a unique invertebrate model to study ageing and its processes in the alterations of behavioural and cognitive functions.
The details of the study can be found in the link above but one of the more interesting aspects of the article to me was the discussion on the current scarcity of animal models of ageing using invertebrates. I think this is (at least partially) a result of a pervasive belief that humans are distinct from animals, and even though over the last century we have chipped away at the dividing line between man and brute with a greater understanding of how humans and related species share commonalities, there seems to still be some resistance to extending this understanding to insects. To suggest that insects are not just automata that respond to their world based entirely on instinct can elicit surprise and incredulity despite the fact that much of our knowledge of how learning works in the brain comes directly from insects1. Once we begin to investigate this topic, finding the similarities of cognitive functions between insect and man, finding the similarities of learning processes between insect and man, we inevitably find ourselves posing a scientifically and ethically difficult question: do insects feel pain?

There are a number of problems with trying to tackle this issue, and Bateson does a great job of broadly describing some of the issues surrounding this question in his paper on "Assessment of Pain in Animals"2. Specifically, he looks at how the traditional understanding of 'pain' is centred around the subjective awareness of a painful stimulus and this is, of course, outside the realm of scientific methods due to the impossibility of finding objective evidence for an experience. In humans we can make reasonable assumptions about somebody being in pain by simply gathering verbal reports; if somebody says, "That is painful" then we can conclude that the stimulus that provoked the response is, in fact, painful. However, the objections to this approach are obvious - the first is that the person could be lying (or misunderstanding their experience, like interpreting the sensation of cold as painful), and the second is that verbal reports in animals are relatively rare for anyone without the abilities of Dr. Doolittle.


In order to get around this, scientists like Bateson suggest that we should use indirect methods to determine the presence of pain in animals. One such approach is perhaps best described by Marian Stamp Dawkins, who has discussed the topic a number of times but presents a good summary of her position in "The Science of Animal Suffering"3. Dawkins begins by looking at the different ways in which the term 'suffering' can be used to describe human emotions but argues that the common element among them all is that it refers to a state of unpleasantness and this can be identified behaviorally by observing the animal attempting to remove themselves from a painful situation; in much the same way a thirsty animal will attempt to obtain water.

By describing behaviors in this way, Dawkins has effectively conceptualised pain as a 'punisher' and removal of pain as a 'reinforcer' (where the formal definitions of these terms refers to something which creates a decrease in behavior, and something which causes an increase in behavior). By viewing the situation in this way, we can fit "suffering" and "pain" into our framework of behavioral knowledge in order to escape the problems associated with the nebulousness of the common terms.

The advantage of this approach is that it avoids using ambiguous biological data or naturalistic claims that can sometimes lead to unreasonable leaps in logic. Dawkins gives the example of how egg-laying hens who have been exposed to an enriched environment will often exhibit a rise in corticosteroids (a hormone correlated to stress) and malformations in the shape of their eggs. With these findings we could mistakenly conclude that hens are in fact happier in barren wire-floored cages but when we perform behavioral tests to assess which environments they prefer, we find that the hens will work harder and longer in order to have access to the enriched environments (as we would expect). So what we might have interpreted as extra stress in the chickens was in fact a correlate for another emotion - most likely a form of excitement or happiness, in the same way we would see a rise in corticosteroids of someone going on a first date but we would be wrong to interpret it as an unpleasant sensation and experience.


The first question that we need to address when looking at pain in insects is whether it is possible for them to experience pain, as if there is no physiological mechanism that could allow for the perception of pain in insects then any behavioral data we might gather would be useless. A review from Eisemann and co4 looked at this and at the time they concluded that there seemed to be no strong evidence to support the existence of nocireceptors in insects, which are the nerve endings that send the pain signals to the brain to communicate the presence of some injury or damage. To further compound the problem, there are reports of behaviors in insects which are inconsistent with how we would intuitively expect an organism in pain to react; for example, there are common examples of locusts continuing to eat whilst currently being eaten by praying mantises, and male praying mantises attempting to continue mating whilst the female eats him.

Despite this, as the Eisemann paper itself discusses, there are difficulties in concluding that insects do not feel pain based on the lack of corresponding biological structures that are found in mammals, or because they do not react in ways that other organisms might. Sherwin5 termed these the problems with the 'argument-by-analogy' and notes the inconsistency of using analogies to determine the existence of pain in animals; that is, a dog writhing in response to a noxious stimulus is deemed to be in pain but an insect reacting similarly in response to a noxious stimulus is often dismissed as not representing a demonstration of pain.

If we allow ourselves to consider the possibility that the squirming of insects in these situations could be indicative of experiencing pain rather than a simple stimulus-response reaction, then we can perhaps make sense of the effect of opioids on the behavior of a number of insects. Opioids are substances which modify the signals transmitted from nocireceptors, and generally this is interpreted as an evolutionary adaptation to reduce pain in the organism. This means that when we observe honey bees receiving an electrical shock and that naloxene (an opioid blocker) decreases their 'stinging response threshold', we can argue for an opioid system in the honey bee that affects the functioning of nociception6. In other words, it is possible that findings like this indicate a biological system designed to alleviate the perception of pain.

Similar results have been found in fruit flies, and when looking at their responses to heated probes researchers even discovered a gene that was dubbed painless7. The importance of this finding is discussed in more detail in the paper "A genome-wide Drosophila screen for heat nociception identifies α2δ3 as an evolutionarily conserved pain gene"8 but the main thrust of the finding is that they were able to find a line of fruit flies that did not react to the noxious stimulus and from there they isolated a specific gene which appeared to control the 'pain' sensation and was absent in this group of fruit flies. This existence of this gene (and the effect it has when it is activated) suggests that there is nocireceptor system in fruit flies, which could be applied to closely related insects and thus used to infer the existence of pain in these organisms.


So do insects experience pain? As with many things in science, our conclusion is currently: we don't know. There is a range of evidence for and against each conclusion and unfortunately the answer for now is the oft-quoted science trope of "more research is needed". But from what we do know, I personally think that it is reasonable to leave the question open and that it is important to not reject the possibility out of hand. It has been over 30 years since Wigglesworth wrote his article "Do Insects Feel Pain?"9, and although the answer to the question was no less ambiguous than it is now, his conclusion still seems reasonable; that we have enough evidence and reason to suppose that insects might feel pain to make it worthwhile to implement laboratory policies that restrict experimentation on insects that could cause pain (for example, dissection without analgesia). For now, I think agreeing on that conclusion would be a positive step forwards.


1. Brembs, B. (2003). Operant conditioning in invertebrates. Current Opinion in Neurobiology, 13(6):710–717.

2. Bateson, P. (1991) Assessment of pain in animals. Animal Behaviour, 42:827–839.

3. Dawkins, M.S. (2008). The science of animal suffering. Ethology, 114:937-945.

4. Eisemann C.H., Jorgensen W.K., Merritt D.J., Rice M.J., Cribb B.W., Webb P.D., Zalucki M.P. (1984) Do insects feel pain? A biological view. Experientia (Basel), 40:164–167.

5. Sherwin, C.M. (2001). Can invertebrates suffer? Or how robust is argument-by-analogy? Animal Welfare, 10:S103-S118.

6. Nunez, J., Almeida, L., Balderrama, N., Giurfa, M., (1997). Alarm pheromone induces stress analgesia via an opioid system in the honeybee. Physiology & Behavior. 63, 75–80.

7. Tracey, W.D., Jr, Wilson, R.I., Laurent, G., Benzer, S. (2003) painless, a Drosophila gene essential for nociception. Cell. 113:261–273.

8.  Neely, G.G., Hess, A., Costigan, M., Keene, A.C., Goulas, S., Langeslag, M., Griffin, R.S., Belfer, I., Dai, F., Smith, S.B., et al. (2010) A genome-wide Drosophila screen for heat nociception identifies α2δ3 as an evolutionarily conserved pain gene. Cell. 143:628–638

9. Wigglesworth, V.B. (1980). "Do Insects Feel Pain?" Antenna. 4:8-9.


  1. "Specifically, he looks at how the traditional understanding of 'pain' is centred around the subjective awareness of a painful stimulus and this is, of course, outside the realm of scientific methods due to the impossibility of finding objective evidence for an experience."

    There's a difference between 'pain' in the 'pain physiology' or 'pain behaviour' sense and 'pain' in the 'pain experience' sense that you can't just ignore because it's "outside the realm of scientific methods".

    'Pain physiology' or 'pain behaviour' can tell us if a being could have 'pain experience' but not that a being does have 'pain experience'.

    And even if we go so far as to assume materialism (which I personally don't do) there's no reason to also assume that the same physiological features that cause 'pain behaviour' also cause our (humans, or mammals, or 'higher' animals) 'pain experience'.

    1. "There's a difference between 'pain' in the 'pain physiology' or 'pain behaviour' sense and 'pain' in the 'pain experience' sense that you can't just ignore because it's "outside the realm of scientific methods"."

      Certainly! I had planned to go into detail on this topic but, as you can probably see, my post was quite lengthy even without it. However, I don't think that ignoring it on scientific grounds necessitates a rejection of it in a metaphysical sense. That is, the point is just that if a certain method for determining pain was good enough for judging humans and mammals, then the same methods must be adequate enough for determining it in insects.

      Of course, when we rely on physiological and behavioral methods, we can question whether they actually experience pain, or have qualia as we do, and that's not a question science can answer - which is what I was trying to get at in the section of my post that you've quoted. It's a philosophical issue but just as we can question whether insects share these qualia, I see no reason why the same arguments couldn't be applied to other animals and even humans.

      So what animal welfare scientists try to do, and what I've tried to do in my post above, is look at it from a purely pragmatic perspective. In other words, we simply ask whether we have any evidence to make it reasonable to suspect that certain individuals or species could experience pain. In my opinion, the physiological and behavioral experience we have in humans and various animals is enough to make it reasonable to, for example, implement laws that prevent people from mistreating them.

      "'Pain physiology' or 'pain behaviour' can tell us if a being could have 'pain experience' but not that a being does have 'pain experience'."

      Of course, and this is why I think the question is outside the realm of scientific methods. No matter how much objective evidence we gather, ultimately we will be no closer to answering the question of whether an individual is having subjective experiences. Essentially, I guess this brings us back to the problem of p-zombies.

      "And even if we go so far as to assume materialism (which I personally don't do) there's no reason to also assume that the same physiological features that cause 'pain behaviour' also cause our (humans, or mammals, or 'higher' animals) 'pain experience'."

      I don't think we need to assume materialism (I personally don't either, and I don't think many scientists do - at least it's not explicitly necessary in their work), but I do agree that relying on physiological mechanisms to determine the existence of pain is problematic. It's an issue in the positive sense that you've highlighted (i.e. the existence of these components do not guarantee the sensation of pain), but it's also an issue in the negative sense, in that the lack of these components do not exclude the possibility of sensing pain.

      That's why I think the behavioral evidence is far more convincing, as it rests on fairly uncontroversial assumptions - basically that pain is something that is avoided. Whether it proves in a philosophical sense that insects (or even animals, or other humans) actually feel pain is certainly another question, and one that I don't think it does, or intends to, answer. But when deciding whether lab animals need to be anaesthetised before surgery or not, I think the evidence is convincing enough for us to determine relevant laws.