In this, the second of three posts on features potentially relevant to invertebrate sentience, we assess 5 drug responses, 5 motivational tradeoffs, and 5 feats of cognitive sophistication. Here are some high-level takeaways:
- Research that analyzes the effects of analgesics, antidepressants, and anxiolytics on invertebrates—especially self-administration studies—has the potential to reveal important evidence about various invertebrates’ capacity for valenced experience.
- Studying motivational tradeoffs can help us distinguish reflexive, pre-programmed behaviors from more plastic responses.
- Comparing cognitive sophistication across dissimilar taxa is extraordinarily difficult.
- Notwithstanding (3), many invertebrates, especially arthropods and cephalopods, appear surprisingly intelligent.
- The relationship between cognitive sophistication and the capacity for valenced experience is unclear.
Introduction and Project Overview
This post is the fourth in Rethink Priorities’ series on invertebrate welfare. In the first post we examine some philosophical difficulties inherent in the detection of morally significant pain and pleasure in nonhumans. In the second post we discuss our survey and compilation of the extant scientific literature relevant to invertebrate sentience, as well as the strengths and weaknesses of our approach to the subject. In the third post we explain some anatomical, evolutionary, and behavioral features potentially indicative of the capacity for conscious experience in invertebrates. In this post we explain some drug responses, motivational tradeoffs, and feats of cognitive sophistication potentially indicative of the capacity for conscious experience in invertebrates. In the fifth post we explain some learning indicators, navigational skills, and mood state behaviors potentially indicative of the capacity for conscious experience in invertebrates. In the sixth, seventh, and eighth posts, we present our summary of findings, both in narrative form and as an interactive database. In forthcoming work, to be published in late July, we analyze the extent to which invertebrate welfare is a promising cause area.
Affected by analgesics in a manner similar to humans
An anesthetic is a substance which induces a general lack of feeling in the target area. In contrast, an analgesic specifically reduces the sensation of pain without eliminating other feelings. In humans various types of analgesics are used as painkillers. Studying the effects of analgesics on invertebrates may help us better understand if they possess the capacity for valenced experience. Of course, there’s no way to directly measure if an analgesic reduces pain sensation in a nonhuman animal, so we have to look for proxies. One such proxy is a reduction in nociceptive reflexes and avoidant behavior. If nonhuman animals react to analgesics with diminished nociceptive reflexes and avoidant behavior, then, in the absence of defeaters, that is evidence that the analgesics are acting to reduce pain in the animals.
One must be careful to distinguish between cases in which nociceptive reflexes and avoidant behavior are reduced from cases in which activity overall is reduced. There is a huge variety of analgesics and their effects on phylogenetically distant animals is largely unknown. If a particular analgesic functions more like an anesthetic, reducing overall sensation, nociceptive reflexes and avoidant behavior may be diminished--but for the wrong reasons. Similarly, an analgesic might merely damage an animal, leading to reduced activity without any suppression of pain. One possible way to circumvent these worries is to test cases in which, if the analgesic did provide pain relief, the animal would increase rather than decrease activity, for example when the sensation of pain might caution an animal against taking a particular action.
Self-administration studies may help us better understand whether an analgesic reduces pain sensation or merely diminishes response to stimuli in general. In (the relevant subset of) self-administration studies, injured animals are given the opportunity to feed or drink from a source laced with an analgesic or a control source with no added drugs. Self-administration studies have been performed on a variety of animals from rats to chickens to honeybees. When the animals in these studies favor the analgesic source, that is mild evidence that their injuries cause them pain for which they seek relief. When the animals show no preference for the analgesic source, that is mild evidence that either the injured animals are not in pain or that the analgesic does not reduce their pain. Of course, confounding factors must be considered. Perhaps the analgesic alters the taste of the food/water so much that it makes the food/water unpalatable despite its pain-relieving properties. And as noted above, there is an astounding array of available analgesics. The fact that one type of analgesic does not reduce the pain associated with one type of injury does not prove that analgesics in general don’t relieve pain or that the animals in the study are incapable of valenced experience.
Affected by recreational drugs in a manner similar to humans
A wide variety of recreational drugs have been tested on invertebrates, including amphetamine in crayfish, cocaine in honey bees, marijuana and caffeine in spiders, and MDMA in octopuses. In humans, recreational drugs often alter the valence of experience. If when exposed to these drugs nonhuman animals exhibit behavior relevantly similar to the behavior of humans exposed to the same drugs, then, in the absence of defeaters, we have some evidence that those nonhuman animals undergo similarly valenced experiences. Of course, not all changes in behavior are the result of changes in valenced experience. Recreational drugs are often toxic, and some changes in behavior, both in humans and nonhumans, are the result of the body reacting to a poison. It’s possible that neurochemical markers can help us distinguish behavioral changes caused by changes in valenced experience from behavioral changes caused by changes in physiology.
Like many other features we investigated, this feature comes in degrees, in this case along at least two dimensions. The first and most obvious dimension is similarity to human behavior. Similarity is not binary. Some reactions to recreational drugs will be more similar to human reactions than others. Even setting aside the problem of measuring similarity, it’s difficult to know what degree of similarity should be sufficient for a creature to be said to possess this feature.
The other dimension concerns the number of drugs that affect a creature in a manner similar to humans. A creature may react to some recreational drugs in a manner similar to humans and react to other recreational drugs in a totally non-human manner. Creatures that react to more recreational drugs in a manner similar to humans might be said to better satisfy this feature than creatures that react to fewer recreational drugs in a manner similar to humans. However, for purposes of this project, we counted a creature as possessing this feature so long as it reacts to at least one recreational drug in a manner similar to humans.
This feature is considerably more ambiguous than many of the other features we investigated. We have not attempted to define “recreational drug” with much precision, and hence there will inevitably be edge cases that are difficult to classify. (For instance, should we include codeine and other medicines that are abused recreationally? We’re not sure.) More worryingly, we have not delineated in advance which behavioral and physiological changes would qualify as a reaction relevantly similar to a human reaction. Given the huge number of recreational drugs and their incredible array of potential effects, it would be difficult for us to specify exactly what reactions to look for. Instead, we’ve had to rely on attributions of similarity in the literature. So, for example, in a recent study on the effects of alcohol on the nematode C. elegans, the authors report that the roundworm displays disinhibited locomotion and feeding behaviors after being exposed to ethanol. They write, “Similar behavioral disinhibition is also seen in many animal models of ethanol response, from invertebrates to mammals and primates.” Thus, we concluded that C. elegans is affected by (at least one) recreational drug in a manner similar to humans.
Self-administers recreational drugs
Self-administration studies may help us better understand whether recreational drugs alter behavior due to changes in valenced experience or merely due to changes in physiology. In (the relevant subset of) self-administration studies, animals are given the opportunity to feed or drink either from sources laced with recreational drugs or unaltered control sources. If the animals consistently prefer the source laced with recreational drugs, that is some evidence that they find the effects of the drug pleasurable, and hence have a capacity for valenced experience.
Of course, there are a number of complications which must be accounted for when evaluating the importance of this feature. First, the recreational drug may alter the taste of the food/water so much that it makes the food/water unpalatable despite its pleasure-inducing properties. Thus, a creature’s failure to self-administer a recreational drug does not preclude the possibility that the creature finds the drug pleasurable. Second, some recreational drugs are highly addictive. If an organism begins a self-administration study consuming equal quantities of the experimental and control source, it may quickly become addicted to the experimental source and alter its behavior accordingly. Importantly, the process of altering one’s behavior in response to an addictive substance need not involve any conscious experience--even in humans. In laboratory settings human addicts will display a subconscious preference for intravenous doses of cocaine too small to produce conscious effects. The addicts do not report any differences in subjective feeling and deny that they are expressing a preference at all. Hence, self-administration behavior need not be driven by the capacity for valenced experience.
Affected by antidepressants or anxiolytics in a manner similar to humans
The effects of antidepressant and anxiolytic drugs have been explored in a number of invertebrate species, including fruit flies, nematodes, and crabs. In humans, antidepressant and anxiolytic drugs alter the valence of experience. If when exposed to these drugs nonhuman animals exhibit behavior relevantly similar to the behavior of humans exposed to the same drugs, then, in the absence of defeaters, we have some evidence that those nonhuman animals undergo similarly valenced experiences. Of course, not all changes in behavior are the result of changes in valenced experience. Antidepressant and anxiolytic drugs may be toxic to some animals, and thus some behavioral changes in nonhumans may be mediated solely by physiological changes. Hopefully, neurochemical analysis can help us distinguish behavioral changes caused by changes in valenced experience from behavioral changes caused by changes in physiology.
An example will help illustrate this feature. According to a 2014 study, stress-induced avoidance behavior in crayfish bears a striking resemblance to mammalian anxiety. The authors demonstrate that shocked crayfish develop an extended, context-independent aversion to light. (The shocks were not associated with levels of illumination.) In contrast, unshocked crayfish, though preferring the dark, were happy to explore both illuminated and unilluminated areas of their environment. Most importantly, injecting the crayfish with the anxiolytic drug chlordiazepoxide (used to treat anxiety in humans) eliminated the aversion to light. In humans, anxiety is often associated with danger that is perceived to be unavoidable or situations in which the threat is ambiguous or unknown. The electric shocks applied to the crayfish fit this description. In humans, anxiety is associated with generalized fear, that is, increased fear of unrelated stimuli. The shocked crayfish appeared to exhibit increased fear of light that is unrelated to the source of stress. In humans, anxiety is reduced by anxiolytic drugs. In crayfish, anxiety-like behavior is reduced by anxiolytic drugs. One explanation of this phenomenon is that crayfish, like humans, are capable of experiencing negatively valenced emotional states.
Like many other features we investigated, this feature comes in degrees, in this case along at least two dimensions. The first and most obvious dimension is similarity to human behavior. Similarity is not binary. Some reactions to antidepressant and anxiolytic drugs will be more similar to human reactions than others. Even setting aside the problem of measuring similarity, it’s difficult to know what degree of similarity should be sufficient for a creature to be said to possess this feature.
The other dimension concerns the number of drugs that affect a creature in a manner similar to humans. A creature may react to some antidepressant and anxiolytic drugs in a manner similar to humans and react to other antidepressant and anxiolytic drugs in a totally non-human manner. Creatures that react to more antidepressant and anxiolytic drugs in a manner similar to humans might be said to better satisfy this feature than creatures that react to fewer antidepressant and anxiolytic drugs in a manner similar to humans. However, for purposes of this project, we counted a creature as possessing this feature so long as it reacts to at least one antidepressant and anxiolytic drug in a manner similar to humans.
Finally, it should be noted that we were not able to find enough studies to include the feature self-administers antidepressant or anxiolytic drugs on the table.
Paying a cost to receive a reward
An animal possesses this feature if it can and will sacrifice something of immediate value in order to earn a positive stimulus. Many animals, including at least one species of invertebrate, are willing to pay a cost to receive a reward. For example, fruit flies will endure electric shock in order to attain the cue associated with ethanol, indicating that they are prepared to tolerate punishment to obtain the drug. This behavior suggests fruit flies are capable of weighing competing interests. In general, motivational tradeoff behavior is evidence that the animal in question possesses some kind of unified utility function in which benefits and risks of different types are processed and evaluated. This type of integration may be evidence for the capacity for valenced experience.
Of course, some tradeoff behavior is sure to be purely reflexive, instinctual, or otherwise pre-programmed. The context in which an animal displays this trait is important for determining the importance of the feature. Tradeoff behavior that occurs in novel situations demonstrates the plasticity that one would expect of a creature with the capacity for valenced experience. Tradeoff behavior that occurs in response to situations repeatedly encountered in the wild does not require the same level of plasticity and hence may not require the capacity for valenced experience.
Paying a cost to avoid a noxious stimulus
An animal possesses this feature if it can and will sacrifice something of immediate value to prevent or halt a noxious event. Many animals, including invertebrates, are willing to pay a cost to avoid a noxious stimulus. For instance, shore crabs generally avoid well-lit areas, preferring to hide in dark environments such as those found under rocks. Given the choice between two chambers in a laboratory setting, one brightly lit and the other dark, the crabs will universally choose the dark chamber. However, if the dark chamber is rigged to deliver a mild shock, the crabs will begin to opt for the normally avoided well-lit chamber. The crabs do so in increasing numbers (and increasingly quickly) as the number of trials increases. A compelling explanation of this behavior is that the crabs feel pain, then learn to avoid the pain by choosing the opposite, otherwise undesirable chamber.
Of course, some tradeoff behavior is sure to be purely reflexive, instinctual, or otherwise pre-programmed. The context in which an animal displays this trait is important for determining the importance of the feature. Tradeoff behavior that occurs in novel situations demonstrates the plasticity that one would expect of a creature with the capacity for valenced experience. Tradeoff behavior that occurs in response to situations repeatedly encountered in the wild does not require the same level of plasticity and hence may not require the capacity for valenced experience.
For our purposes, self-control is the ability to consistently choose a large delayed reward over a small immediate reward. Self-control requires some form of processing in which different needs and desires are weighed against each other. Self-control probably cannot be exercised reflexively. Thus, self-control is evidence of behavioral plasticity, which is itself modest evidence for the capacity for valenced experience.
Self-control has been studied extensively in vertebrates, including humans. Pigeons and rats consistently demonstrate a preference for immediate small rewards over large delayed rewards. This impulsiveness may have an evolutionary explanation. In environments where animals face uncertain futures, opting for a more certain small reward may be more beneficial than waiting for a larger reward which may not come. Self-control has only recently been studied in invertebrates. Eusocial insects, such as bees and ants, would appear to benefit from self-control. Because workers are infertile, propagating workers’ genes requires the long-term survival of the hive or colony. Initial indications suggest that honey bees choose large delayed rewards over small immediate rewards at much greater rates than rats and pigeons.
Predator avoidance tradeoffs
To stay alive, animals must avoid predators. To stay alive, animals must eat and drink. Sometimes, the need to avoid predators conflicts with the need to eat and drink. In these cases, the animal must weigh competing demands and come to a decision. Such predator avoidance tradeoffs have been widely studied, including in invertebrates. Earthworms will feed in high-light conditions (exposing them both to visual predators and the threat of dessication) when they are starved and half-starved but not when they are fully nourished. Overhead shadows (indicating a potential predator) cause fully nourished fruit flies to disperse, but starved fruit flies remain feeding. In a laboratory setting, hermit crabs will abandon their shells if they are subjected to a mild shock. However, the crabs are significantly less likely to abandon their shells after shock if the odor of a predator is present.
Like other motivational tradeoffs, some predator avoidance tradeoff behavior is likely to be reflexive and hardwired. Nearly all animals, even very simple creatures, must balance the risk of starvation and the risk of predation. Take no risks when nearing starvation and the animal will starve to death. Take too many risks when already nourished and an animal may die without reducing risk from starvation. Thus, it could be argued, having some varied response in these scenarios would be expected given evolutionary selection pressures regardless of the presence or absence of mental states of any kind.
Some contexts, however, can provide evidence for the capacity for valenced experience. When considering shocked hermit crabs leaving their shells, it is tempting to conclude that the behavior is purely reflexive. However, the fact that the crabs remain in their shells when the odor of a predator is present suggests that the behavior is not reflexive. A natural explanation is that the crabs weigh the pain of the shock against the fear of a predator. In the absence of defeaters, it may be more parsimonious to prefer this explanation to alternative explanations.
Selective attention to noxious stimuli over other concurrent events
The degree to which a noxious event commands the attention of an animal can potentially tell us something about whether that event is experienced as painful or not. In humans the conscious sensation of pain is distracting. Mild pain can be an annoyance, and intense pain can be so overwhelming that it effectively incapacitates the sufferer. If a nonhuman animal devotes more attention to a noxious event than other potential objects of attention, especially if doing so reduces rather than increases an animal’s fitness, that is evidence that the noxious event is experienced as painful. In contrast, if an animal completely ignores a noxious event, that is evidence that the noxious event is not experienced as painful. For example, male mantids will continue to mate even as they are devoured by their partners. This behavior has a clear evolutionary explanation. Only by mating successfully can the male mantid’s genes propagate. Nonetheless, it is hard to imagine any mammalian male continuing to mate while it is being eaten alive; the pain would be too distracting. Hence, it seems the male mantid doesn’t experience the process of being eaten alive as painful. Perhaps it doesn’t experience anything as painful.
An animal engages in uncertainty monitoring when it is aware, in a rough and minimally functional sense, of its credence in a particular proposition. Humans engage in uncertainty monitoring with relative ease. Asked to name both the 22nd president of the United States and the 1st president of the United States, most humans have little problem identifying which answer they are more confident in. Even three-year-old children are capable of uncertainty monitoring.
For most of the 20th century, it was thought that uncertainty monitoring was a skill unique to humans. In the mid-1990s, researchers began applying the so-called “uncertain response” paradigm to nonhuman animals. In an uncertain response paradigm, animals are trained to associate a correct answer to some question with a reward and an incorrect answer with a punishment. A third answer choice is then introduced: uncertain. If an animal responds with the uncertain answer, it receives neither punishment nor reward. When presented with a series of increasingly difficult questions, animals capable of uncertainty monitoring will select the uncertain response where appropriate, thereby improving their overall reward-to-punishment ratio. Using this paradigm, uncertainty monitoring has been conclusively demonstrated in both dolphins and rhesus monkeys.
Recently, uncertainty monitoring has been studied in eusocial insects. Using the paradigm outlined above, one study tasked honey bees with discriminating between two stimuli. A correct answer earned a reward (sucrose) and an incorrect answer earned a punishment (quinine). When given the choice to opt out, honey bees were found to opt out more often when the trial was difficult (when the honey bee was proportionately more likely to receive a punishment than a reward). The honey bees on average improved their success-to-failure ratio when given the option to opt out of trials.
Another study found that ants upregulate pheromone trail deposition in response to changes in the location of food. In an experiment with a T-maze, researchers trained ants to a feeder location, then altered the environment by changing the feeder location to the other arm of the T-maze. After finding the new food source, ants upregulated pheromone deposition if they had made a wrong choice. Additionally, the researchers found that outgoing ants which went on to make an error deposited less pheromone. This seems to imply that the ants can measure the reliability of their own memories, and respond accordingly, depositing more or less pheromones.
Uncertainty monitoring is a form of metacognition. According to many higher-order theories of consciousness, metacognition is a necessary condition on conscious experience. More speculatively, metacognition might be construed as a type of self-awareness. And self-awareness is plausibly a sufficient condition on conscious experience. (One must be conscious in order to be self-conscious.) Thus, uncertainty monitoring is a particularly important feature to investigate. Indeed, it has been said that “the comparative study of metacognition potentially grounds the systematic study of animal consciousness.”
However, one ought to interpret the alleged examples of uncertainty monitoring in invertebrates with caution. Even the authors of such studies are unsure what their research represents. Regarding their findings on the upregulation of pheromone deposits in ants, Tomer Czaczkes and Jürgen Heinze write, “it’s hard to believe that such tiny-brained animals are capable of such an advanced cognitive feat” and “one could conceive of several alternative explanations for our findings, which do not invoke metacognition.” At the same time, they argue that their findings, “alongside similar results from honeybees (Perry & Barron, 2013), are suggestive of metacognitive abilities in social insects.” As with many other features, more research is needed.
Self-recognition (mirror test)
Self-recognition is an important feature to investigate because if self-recognition implies self-awareness, then on some theories of mind, self-recognition is fairly direct evidence of consciousness. Self-recognition in nonhuman animals is generally measured via the mirror test. The test assesses self-recognition by determining whether an animal can recognize its own reflection in a mirror. This is accomplished by secretly marking the animal with a small dot that is only visible by looking in the mirror. If the animal sees the dot in the mirror then touches the dot on its body (indicating it understands the relationship between itself and the creature in the mirror), it passes the test.
The mirror test has several well-known methodological shortcomings. For example, because the test uses visual perception, creatures which perceive the world primarily through other sense modalities seem to be at a disadvantage. (Dogs, who have repeatedly failed the classic mirror test, navigate the world mostly by a combination of olfaction and audition.) Creatures for whom eye contact is a sign of aggression also seem to be at a disadvantage because they will either refuse to directly investigate their reflection, or, if they do make protracted eye contact, will move to counteract the perceived aggression of the ‘foreign’ creature before they have an opportunity to notice the dot. (Gorillas, who are mostly reported to fail the mirror test, fit this profile.)
These flaws lead to false negatives. But false positives are also possible. The mirror test is an imperfect measure of self-recognition. But even if the mirror test were a perfect measure, it would still be difficult in some cases to definitively say whether some animal passes the test. For instance, according to a recent study, a species of bony fish (the bluestreak cleaner wrasse, Labroides dimidiatus) passes the mirror test, the first animal outside birds and mammals to do so. From our non-specialist perspective, it is unclear how to evaluate such a new, controversial finding, especially given the small sample size of the study. Thus, for this feature and many others, we have been forced to make some difficult judgment calls. Again, we have tried to indicate as much in comments attached to the relevant cells.
Deceptive behavior in animals comes in a wide variety of forms. Many species use mimicry or camouflage to avoid predators or ambush prey. Some species temporarily adopt more aggressive behavior during times of vulnerability to deter attackers. And some species, notably humans, intentionally misrepresent their beliefs in order to achieve some goal.
For purposes of this project, we are only concerned with deceptive behavior insofar as it bears on the possession of phenomenal consciousness. Thus, genuine deception (also sometimes called “tactical” or “intentional” deception) must be distinguished from passive adaptations, such as mimicry or camouflage. If possible, genuine deceptive behavior also ought to be distinguished from behaviors that appear to be instances of deception but are in fact innately specified or automatically elicited by external stimuli. In practice, of course, it may be difficult to make this distinction.
Genuine deception is a complex cognitive and social skill. To deceive, one must be capable of appreciating the distinction between pretense and reality. One must also be capable of representing, to some extent, the mental states of the deception target. To deceive one must have a cognitive grasp on how things appear, both to oneself and to one’s target. To be aware of appearances, one must experience them. A complex form of behavior in which an animal uses appearances to deceive other animals is thus potentially fairly direct evidence that the animal is phenomenally conscious. The absence of deceptive behavior, however, doesn’t tell us much. Higher-order cognition is plausibly not a requirement for phenomenal consciousness.
For a plausible example of genuine deception in invertebrates, consider cuttlefish. Cuttlefish deploy a wide variety of chromatic, textural, postural, and locomotor elements to communicate with predators, prey, and conspecifics, and there is evidence that they possess a sophisticated theory of mind. During courtship male cuttlefish engage in a remarkable form of tactical deception with rival males: simultaneous dual gender signalling. A male will position itself between a rival male and a potential mate. On the side of its mantle facing the potential mate, the deceptive male will produce typical chromatic courtship patterns. But on the side of its mantle facing the rival male, it will mimic typical female displays, thus confusing the rival and significantly reducing the odds that the rival male will attempt to disrupt copulation.
It was once thought that only humans use tools. Although the definition of “tool” admits of borderline cases and the phenomenon “using a tool” is similarly vague, it has nevertheless become clear that this old thought is false. A wide range of nonhuman animals, including mammals, birds, even some fish, use a variety of tools to great effect.
Tool use is often considered a benchmark of cognitive sophistication. Genuine tool use requires the acquisition of a foreign object to be utilized at some later time. (As such, the makeshift shells sometimes acquired by hermit crabs do not qualify as tool use because the shell is effectively in use the whole time the crab inhabits it.) In most cases of tool use, the acquisition or transportation of the tool inflicts some cost on the animal, for which the animal is compensated by some benefit when the tool is later deployed. Thus, tool use also requires a certain degree of foresight and planning.
For purposes of this report, we separate mere tool use, which is often instinctive and narrowly-structured, from so-called “flexible” tool use. An animal is capable of flexible tool use when it can utilize objects with which it does not have an evolutionary history. Mere tool use, in contrast, is confined to natural objects and normally occurs only in response to specific environmental cues in order to achieve a small number of specific goals. Although evidence for flexible tool use in invertebrates is only just emerging (see below), simple tool use has long been documented in insects, for example the many species of ant that use leaves to transport food.
Flexible tool use
Among nonhuman animals, tool use is perhaps most striking in chimpanzees, who use tools for everything from nut-cracking to insect-gathering. (Indeed, the subtle variations in constructed tools among different chimpanzee populations are sometimes said to constitute “cultural” differences among the groups.) This sort of tool-making and tool-using is probably too complex to be the direct product of evolutionary forces. Rather, it is a byproduct of chimpanzees’ general intelligence.
New evidence has recently emerged that flexible tool use is not restricted to primates. Though controversial, flexible tool use is now sometimes attributed to invertebrates. For example, Finn et al. 2009 reports soft-sediment dwelling octopuses retrieving coconut shell halves discarded by the local human population and later assembling the shell halves into protective shelters. The awkward manner in which the octopuses must move while carrying these shells (the authors describe it as “stilt-walking”) represents a cost in terms of energy and increased predation risk, which is only recouped later when the shelves are successfully assembled into a surface shelter or encapsulating lair. Importantly, the only known source of these clean and lightweight shells is the coastal human communities, and thus the octopuses have not interacted with these items on an evolutionary timescale.
It has even been suggested that bees are capable of using tools flexibly. In a recent study Loukola et al. 2017 trained bumble bees to see that a ball could be used to dispense a reward. In subsequent iterations of the experiment, the bees independently learned to solve the task more efficiently by using a ball positioned more closely to the target, even when the ball was a different color. Such cognitive flexibility is the hallmark of sophisticated tool use.
Finally, it must be noted that tool use, no matter how ingenious, is not direct evidence of valenced experience or phenomenal consciousness. Rather, tool use is positively correlated with cognitive sophistication. For those that believe cognitive sophistication is a necessary condition on phenomenal consciousness (or that phenomenal consciousness is a byproduct of cognitive sophistication), tool use is then indirect evidence for such states.
This essay is a project of Rethink Priorities. It was written by Jason Schukraft with contributions from Max Carpendale. Thanks to Kim Cuddington, Marcus A. Davis, Peter Hurford, and Daniela Waldhorn for helpful feedback. If you like our work, please consider subscribing to our newsletter. You can see all our work to date here.
Vertebrates constitute a subphylum in the phylum Chordata. Cladistically, it would be more precise to speak of ‘chordates’ and ‘non-chordates.’ In using the terms ‘vertebrates’ and ‘invertebrates’ we defer to common usage. However, the number of invertebrates in the phylum Chordata is trivial compared to the number of invertebrates outside Chordata, so common usage is not wholly inaccurate. ↩︎
We use the terms ‘sentience,’ ‘phenomenal consciousness,’ and ‘subjective experience’ interchangeably. An organism is sentient just in case there is something it is like to be that organism. ‘Valenced experience’ denotes a proper subset of conscious experience in which experiences take on a positive or negative affect. All creatures with the capacity for valenced experience are necessarily sentient, but not all sentient creatures necessarily have the capacity for valenced experience. Note: ‘sentience’ gets used in different ways by different philosophical communities. In philosophy of mind, the term is normally used in its broad sense, to mean ‘phenomenal consciousness.’ (See, inter alia, this SEP article on animal consciousness.) In moral philosophy, the term is normally used in its narrow sense, to mean ‘valenced experience.’ (See, inter alia, this SEP article on the grounds of moral status.) We have adopted the philosophy of mind usage. ↩︎
For an example of this in crabs, see Stuart Barr and Robert W. Elwood. 2011. “No evidence of morphine analgesia to noxious shock in the shore crab, Carcinus maenas.” Behavioural processes 86, no. 3: 340-344. ↩︎
Alexandra L. Whittaker and Gordon S. Howarth. 2014. “Use of Spontaneous Behaviour Measures to Assess Pain in Laboratory Rats and Mice: How Are We Progressing?” Applied Animal Behaviour Science 151: 1-12. ↩︎
T.C. Danbury, C. A. Weeks, J. P. Chambers, A. E. Waterman-Pearson, and S. C. Kestin. 2000. "Self-Selection of the Analgesic Drug Carprofen by Lame Broiler Chickens." The Veterinary Record 146: 307-311. ↩︎
Julia Groening, Dustin Venini, and Mandyam V. Srinivasan. 2017. “In Search of Evidence for the Experience of Pain in Honeybees: A Self-Administration Study.” Scientific Reports 7: 45825. ↩︎
Udita Datta, Moira van Staaden, and Robert Huber. 2008. "Crayfish Self-Administer Amphetamine in a Spatially Contingent Task." Frontiers in Physiology Volume 9, Article 433. ↩︎
Eirik Søvik, Jennifer L. Cornish, and Andrew B. Barron. 2013. “Cocaine Tolerance in Honey Bees.” PLoS ONE 8 (5): e64920. ↩︎
D.A. Noever, R.J. Cronise, and R.A. Relwani. 1995. “Using Spider-Web Patterns To Determine Toxicity.” NASA Tech Briefs, 19 (4), 82. ↩︎
Eric Edsinger and Gül Dölen. 2018. “A Conserved Role for Serotonergic Neurotransmission in Mediating Social Behavior in Octopus.” Current Biology 28: P3136-314.E4. ↩︎
See, for example, Brian V. Entler, J. Timothy Cannon, and Marc A. Seid. 2016. “Morphine Addiction in Ants: A New Model for Self-Administration and Neurochemical Analysis.” The Journal of Experimental Biology 219.18: 2865–69. ↩︎
Stephen M. Topper, Sara C. Aguilar, Viktoria Y. Topper, Erin Elbel, and Jonathan T. Pierce-Shimomura. 2014. “Alcohol disinhibition of behaviors in C. elegans.” PLoS One 9, no. 3: e92965. ↩︎
This is the most common paradigm for testing self-administration, but there are other paradigms that are sometimes used. For example, to test the self-administration of amphetamine in crayfish, researchers used a spatially contingent task paradigm in which the animals had to distinguish a quadrant of the arena with a particular textured substrate in order to earn the drug. Udita Datta, Moira van Staaden, and Robert Huber. 2008. "Crayfish Self-Administer Amphetamine in a Spatially Contingent Task." Frontiers in Physiology Volume 9, Article 433. ↩︎
Kent C. Berridge and Terry E. Robinson. 2011. “Drug Addiction as Incentive Sensitization.” in Poland and Graham (eds.) Addiction and Responsibility MIT Press: 21-54. ↩︎
An anxiolytic drug is a medication used to treat anxiety. ↩︎
Ariane-Saskia Ries, Tim Hermanns, Burkhard Poeck, and Roland Strauss. 2017. “Serotonin modulates a depression-like state in Drosophila responsive to lithium treatment.” Nature Communications 8, Article number: 15738. ↩︎
Alex T. Ford and Peter P. Fong. 2015. “The Effects of Antidepressants Appear to Be Rapid and at Environmentally Relevant Concentrations.” Environmental Toxicology and Chemistry, 35(4), 794-798. ↩︎
Trevor James Hamilton, Garfield T. Kwan, Joshua Gallup and Martin Tresguerres. 2016. “Acute fluoxetine exposure alters crab anxiety-like behaviour, but not aggressiveness.” Scientific Reports volume 6, Article number: 19850. ↩︎
Even in humans, antidepressants and anxiolytics can sometimes cause a wide range of unpleasant side effects. See James M Ferguson. 2001. “SSRI antidepressant medications: adverse effects and tolerability.” Primary care companion to the Journal of clinical psychiatry 3, no. 1: 22-27 and J. Guy Edwards. 1981. “Adverse effects of antianxiety drugs.” Drugs 22, no. 6: 495-514. ↩︎
Pascal Fossat, Julien Bacqué-Cazenave, Philippe De Deurwaerdère, Jean-Paul Delbecque, and Daniel Cattaert. 2014. “Anxiety-like behavior in crayfish is controlled by serotonin.” Science 344, no. 6189: 1293-1297. Note that the popular introduction to the article mistakenly states that the crayfish were afraid of the dark, not the light. A corrected title for the popular introduction is available here. ↩︎
Separately, the injection of serotonin in unshocked crayfish induced light-aversion behavior that was also eliminated by chlordiazepoxide. In humans, elevated levels of serotonin are associated with anxiety. ↩︎
This feature is often investigated using the conditioned preference paradigm, which is usually used for studying drug addictions in animals. These experiments have been widely developed with rats. See, e.g., Thomas M. Tzschentke. 1998. “Measuring reward with the conditioned place preference paradigm: a comprehensive review of drug effects, recent progress and new issues." Progress in Neurobiology 56, no. 6: 613-672. ↩︎
Karla R. Kaun, Anita V. Devineni, and Ulrike Heberlein. 2012. “Drosophila melanogaster as a model to study drug addiction.” Human genetics 131, no. 6: 959-975. ↩︎
The contingencies under which a reinforcer is presented, rather than (or in addition to) its physical properties, can influence whether or not an association is generated. This is called “selective association.” See, inter alia, Stanley J. Weiss, Leigh V. Panlilio, and Charles W. Schindler. 1993. “Selective associations produced solely with appetitive contingencies: The stimulus‐reinforcer interaction revisited.” Journal of the Experimental Analysis of Behavior 59, no. 2: 309-322. Additionally, in studies with vertebrates, it has been seen that an animal's performance in a conditioning experiment may be affected by the choice of particular combinations of conditioned or discriminative stimuli and reinforcers. In other words, certain combinations of stimuli and reinforcers may fail to elicit expected behaviors, whereas other combinations from the same set of stimuli do result in the expected performance. This is called “stimulus-reinforcer interaction.” See, inter alia, Donald D. Foree and Vincent M. LoLordo. 1973. “Attention in the pigeon: differential effects of food-getting versus shock-avoidance procedures.” Journal of Comparative and Physiological Psychology 85, no. 3: 551. ↩︎
Barry Magee and Robert Elwood. 2013. “Shock avoidance by discrimination learning in the shore crab (Carcinus maenas) is consistent with a key criterion for pain.” Journal of Experimental Biology 216: 353-358. ↩︎
Adele Diamond. 2013. “Executive Functions.” Annual review of psychology 64: 135-168. ↩︎
Young children and chimpanzees routinely fail tests of self-control. Children begin to outperform chimps around age six. Esther Herrmann, Antonia Misch, Victoria Hernandez‐Lloreda, and Michael Tomasello. 2015. “Uniquely human self‐control begins at school age.” Developmental science 18, no. 6: 979-993. ↩︎
Ken Cheng,, Jennifer Peña, Melanie A. Porter, and Julia D. Irwin. 2002. "Self-Control in Honeybees." Psychonomic Bulletin & Review 9: 259-263. ↩︎
Pawandeep Sandhu, Oskar Shura, Rosalind L. Murray, and Cylita Guy. 2018. “Worms make risky choices too: the effect of starvation on foraging in the common earthworm (Lumbricus terrestris).” Canadian Journal of Zoology 96, no. 11: 1278-1283. ↩︎
William T. Gibson, Carlos R. Gonzalez, Conchi Fernandez, Lakshminarayanan Ramasamy, Tanya Tabachnik, Rebecca R. Du, Panna D. Felsen, Michael R. Maire, Pietro Perona, and David J. Anderson. 2015. “Behavioral responses to a repetitive visual threat stimulus express a persistent state of defensive arousal in Drosophila.” Current Biology 25, no. 11: 1401-1415. ↩︎
Barry Magee and Robert Elwood. 2016. “Trade-offs between predator avoidance and electric shock avoidance in hermit crabs demonstrate a non-reflexive response to noxious stimuli consistent with prediction of pain.” Behavioural Processes 130: 31-35. ↩︎
It should also be noted that for some species, such as sheep and cattle, anti-predator responses are no longer needed, and thus are no longer adaptive, due to domestication. Cornelia Flörcke and Temple Grandin. 2013. “Loss of anti-predator behaviors in cattle and the increased predation losses by wolves in the Northern Rocky Mountains.” Open Journal of Animal Sciences 3, no. 03: 248. ↩︎
Sam C.C. Chan, Chetwyn C.H. Chan, Anne S.K. Kwan, Kin-hung Ting, and Tak-yi Chui. 2012. “Orienting attention modulates pain perception: an ERP study.” PLoS One 7, no. 6: e40215. ↩︎
C. H. Eisemann, W. K. Jorgensen, D. J. Merritt, M. J. Rice, B. W. Cribb, P. D. Webb and M. P. Zalucki. 1984. “Do Insects Feel Pain? - A Biological View.” Experentia 40: 164-167. ↩︎
Alternatively, it might experience the process of mating as so immensely pleasurable that the pain of being eaten alive is outweighed. ↩︎
Kristen E. Lyons and Simona Ghetti. 2011. “The development of uncertainty monitoring in early childhood.” Child Development 82, no. 6 : 1778-1787. ↩︎
Michael J.Beran, Justin J. Couchman, Mariana VC Coutinho, Joseph Boomer, and J. David Smith. 2010. “Metacognition in nonhumans: methodological and theoretical issues in uncertainty monitoring.” In Efklides A., Misailidi P. (eds)Trends and Prospects in Metacognition Research, pp. 21-35. Springer, Boston, MA. ↩︎
Clint J. Perry and Andrew B. Barron. 2013. “Honey bees selectively avoid difficult choices.” Proceedings of the National Academy of Sciences 110, no. 47: 19155-19159. ↩︎
Tomer J. Czaczkes and Jürgen Heinze. 2015. “Ants adjust their pheromone deposition to a changing environment and their probability of making errors.” Proceedings of the Royal Society B: Biological Sciences 282, no. 1810: 20150679. ↩︎
J. David Smith, Wendy E. Shields, and David A. Washburn. 2003. “The comparative psychology of uncertainty monitoring and metacognition.” Behavioral and Brain Sciences 26, no. 3: 317-339. ↩︎
Czaczkes and Heinze 2015: 5. ↩︎
False negatives are even found in six-year-old humans. At least one study indicates the average age at which children first pass the mirror test varies considerably across cultures. One hypothesis is that non-Western children use mirrors less frequently and less conspicuously than their Western peers. Thus, although non-Western children acquire the self-concept at roughly the same development stage as their Western counterparts, mirror use, especially public mirror use, confuses them. Tanya Broesch, Tara Callaghan, Joseph Henrich, Christine Murphy, and Philippe Rochat. 2011. “Cultural variations in children’s mirror self-recognition.” Journal of Cross-Cultural Psychology 42, no. 6: 1018-1029. ↩︎
Masanori Kohda, Takashi Hotta, Tomohiro Takeyama, Satoshi Awata, Hirokazu Tanaka, Jun-ya Asai, and Alex L. Jordan. 2019. “If a fish can pass the mark test, what are the implications for consciousness and self-awareness testing in animals?.” PLoS Biology 17, no. 2: e3000021. ↩︎
For a critical response, see Frans B. M. de Waal. 2019. “Fish, Mirrors, and a Gradualist Perspective on Self-Awareness.” PLoS Biology 17, no. 2: e3000112. ↩︎
Stan Kuczaj, Karissa Tranel, Marie Trone, and Heather Hill.. 2001. “Are Animals Capable of Deception or Empathy? Implications for Animal Consciousness and Animal Welfare.” Animal Welfare 10: 161-173. ↩︎
Michael Tye. 2017. “Do Fish Have Feelings?” in Kristen Andrews and Jacob Beck (eds.) The Routledge Handbook of Philosophy of Animal Minds 169-175. ↩︎
Amber Thomas and Christy MacDonald. 2016. “Investigating body patterning in aquarium-raised flamboyant cuttlefish (Metasepia pfefferi).” PeerJ 4: e2035. ↩︎
Culum Brown, Martin P. Garwood, and Jane E. Williamson. 2012. “It Pays to Cheat: Tactical Deception in a Cephalopod Social Signalling System.” Biology Letters 8: 729-732. ↩︎
Giacomo Bernardi. 2012. “The use of tools by wrasses (Labridae).” Coral Reefs 31, no. 1: 39. ↩︎
Christopher Baber. 2003. Cognition and Tool Use: Forms of Engagement in Human and Animal Use of Tools. London: CRC Press. ↩︎
William C. McGrew. 1992. Chimpanzee Material Culture: Implications for Human Evolution. Cambridge University Press. ↩︎
Julian K. Finn, Tom Tregenza, and Mark D. Norman. 2009. “Defensive Tool Use in a Coconut-Carrying Octopus.” Current Biology 19: PR1069-0R1070. ↩︎
Olli J. Loukola, Clint J. Perry, Louie Coscos, and Lars Chittka. 2017. “Bumblebees Show Cognitive Flexibility by Improving on an Observed Complex Behavior.” Science 355: 833-836. (See here for an inexplicably adorable video clip of the bees in action.) ↩︎
Another comment about uncertainty monitoring: Central place foragers tend to spend extra time memorizing the visual landmarks around their nest at times, there is a recent paper on ants describing how this correlates to uncertainty in some detail. As an insect moves further from its nest the accuracy of its knowledge of the nest decreases (errors accumulate in its path integration), and there is evidence that the magnitude accumulated error influences which search strategy an ant will use if it gets lost.
I also have a feeling that insects will start to ignore a sensorimotor cue that provides starts to provide unreliable information. For instance, airflow and visual motion are usually correlated to movement direction and used to control parameters like flight speed. If wind is artificially manipulated such that it is no longer correlated to visual motion or flight speed (it should be random, not negatively correlated), then I think the insect would stop using it as a cue to control flight speed. I can't find a reference for this quickly, but I can look further if it's of interest. I recall something similar also occurs in the case where two cues are initially paired with a reward during associative conditioning but only one turns out to be consistently rewarded (the distractor is called a confound) - after a while a bee can learn to ignore the confound and increase its accuracy. Again I don't have a reference at hand for this but could look later.
Gavin, if you have the time, I'd love to see references for those abilities. We've got another round of invertebrate posts coming out in mid-to-late July, and most if not all of your examples can be used to bolster the case that arthropods in general and insects in particular deserve a close look from the effective animal advocacy movement. Thanks for your contributions!
Hi Jason, I look a bit more into the idea of uncertainty modelling in both sensorimotor and learning activities. I must admit I couldn’t find much on learning to ignore completely random cues (maybe I picked this idea up at a talk or discussion, not a paper, or I can’t get the right search terms for it), but I did come across a few extra studies associated with sensory processing, navigation and foraging that you might be interested in.
In terms of sensorimotor learning, the idea of humans doing probabilistic weighting of sensory cues according to their reliability was proposed of by Wolpert quite recently. There aren’t many insect studies looking at this directly, although some have considered this indirectly as part of the study design. Other useful search terms are probabilistic/dynamic/bayesian/optimal/multisensory reweighting/integration, reliability, uncertainty, and this all links closely to adaptive motor control.
A basic example is the idea of neural correlates that works to determine when several different sensory measurements agree - this was found a long time ago in locust neurons which fire most strongly when visual motion (as seen by both ocelli and compound eyes) and air flow indicate the same direction of motion. More recently moths have been found to increase their visual integration time at lower light levels, which would allow them to see more accurately at the expense of moving slower. Aphids have also been found to respond to indicators of predators approaching by temporally correlating multiple sensory cues in the case where individual cues may be unreliable. The above are probably examples of systems that innately deal robustly with uncertain information. However, I once did a study where I waxed or amputated honeybee antenna, in both cases removing their main sensory perception of airflow. The honeybees reacted to airflow differently in both cases, and I (very speculatively) suggested this could be because the honeybees that still had intact antenna believed the information from it, while amputated honeybees then tried to use alternate cues (air flow on body hairs and legs for instance). I don't think my study is confirmation honeybees do sensory re-weighting (it wasn't intended to be), but such proof may already exist in a similar context using cue conflicts or ablations.
In terms of navigation, I also found studies claiming optimal usage of navigation cues by ants and Drosophila.
There are also studies looking at foraging decisions made by bumblebees, which suggests that they prefer flowers that provide consistent nectar rewards, and they change their visitation rate to flowers depending on how likely the flower is to provide a reward.
Finally, there is a study suggesting that individual ants assess their uncertainty when deciding how to contribute to colony level decisions.
Hope these are helpful references with regards to uncertain insects!
Thanks for the references, Gavin! You truly are an inexhaustible resource. The paper on uncertainty monitoring in ants looks particularly impressive and relevant. I hope to give it a full read later this week. The ability of eusocial insects to incorporate diverse streams of information into an integrated decision-making framework is, to my mind, decent evidence that they are conscious.
(Also, I'm getting a session timed out error on the aphid link.)
No worries! Yes, eusocial insects certainly are quite amazing creatures. There are actually studies looking at facultatively social bee species (whereby females can nest individually or in hives with multiple reproductive females) that suggest sociality leads to increase in brain volume. Besides cognitive demands, sociality also appears to lead to other things like increased hygiene and immune function prevent disease spread in a colony.
Actually, it could be interesting to include naked mole-rats as a vertebrate comparison specific to social insects in this study. I'm not really familiar with their biology but they are generally considered eusocial , particularly that there is division of reproductive labour that creates queen and worker castes within colonies. Maybe impressive feats seen in social insects also appear in mole-rats more than you would expect compared to normal rats? In fact, there are also eusocial species shrimps from the Synalpheus genus which would probably display different traits to the other groups of crustaceans you're looking at.
I also updated the Aphid link, it should work now, but the link is below if it doesn't.
My comments are certainly biased towards bees because of my background. I hope there are relevant examples available for other invertebrates groups, although it may be that a lot of these concepts have mostly been tested in Drosophila or eusocial insects.
A few thoughts about the categories in this article:
-Deception: There are some species of cuckoo bees that will sneak into the hive of another (in this case a solitary) bee, eat the owners eggs and then lay there own. As is the case with cuckoo birds, the owner then happily raises them as her own.
More extreme cases of nest parasitism occur in bumblebees when a cuckoo bumblebee invades a newly established hive of a true bumblebee, kills its queen, and then uses the original queen's workers to raise her own offspring (the cuckoo bumblebee can only lay fertilised eggs, not workers). The later is more complex than a passive act of deception, although it's also not clear to what extent the original workers are completely deceived or just being dominated by the invader.
-Self-control: I'm not sure that comparing self-control between feeding and reproductive contexts is really appropriate. Maybe a better choice would be fungus gardening or aphid herding by ants: In the former case the ants don't eat the leaves they collect in order to grow fungus on them (although I am not sure the ants could actually digest the raw leaves), in the later case they don't eat the aphids so they can milk them (this needs a video).
The self-control of bees is kind of imposed by most workers being sterile and the queen dominating them. This is also not universally true, and the weird relatedness between bee colony members and the occasional presence of workers with developed ovaries mean that it is advantageous for workers to lay male eggs if they have the opportunity (unfertilized honeybee eggs produce male clones of their mother; so a bee is most related to her sons, potentially more related to her sisters ((if they have the same father)) and their sons than her mother, and least related to her brothers - I'm not sure this is true for all social bee species). In bumblebees this can result in a worker revolts where the workers in an established colony kill their queen and all start laying male eggs.
-Paying a cost to receive a reward: Aphid herding ants defend their aphids from predators/competitors and it seems that they make a cost-benefit type decision about if they will defend them.
-Tool use: I think that prolonged nest construction kind of fits in here. External resources need to be collected over time (different bees use combinations of mud, resin, cotton, flower petals, small rocks, and other items to build their nests) in specific sequences, the cost is lost time foraging for food, and the benefit of the nest might not be realized until it is finished (or gained progressively during construction, it's not useful straight away like the hermit-crab's shell).
Thanks for the examples; keep them coming! Whether or not they possess the capacity for valenced experience, eusocial insects truly are remarkable creatures. Do you have an easy reference for the cuckoo bumblebee behavior? I’ve got a running list of amazing things different invertebrates do, and I’d love to add it to the list.
(On the subject of videos, check out the video I’ve linked in footnote 53. It always brings a smile to my face.)
No worries Jason, happy to keep posting the examples that come to mind (finally my knowledge of obscure insect behaviours is useful in EA!). This is a recent review of bumblebee cuckoos that could be useful. I also found another study indicating bumblebee cuckoos actively change their odor profiles to maintain control over the hives workers.
I agree, bumblebees look amazingly cute when rolling balls around! The string pulling experiment done by the same lab also has a nice video.
To extend the tool use point a bit, I recall that primates have been found to have extra neurons in sensorimotor brain regions that are most active when the animal is using a tool, and essentially provide extra capacity for the brain to extend sensory and motor mappings/homunculus to include external artifacts (apparently also quite useful when learning to control of things with BCI). I'm not sure if this type of latent neural capacity has been found in rodents and strongly suspect it wouldn't be present in insects (they tend to be quite frugal with their neurons!), although tool using birds like crows may have been studied as a comparison. Having neural circrity for tool use should be a sufficient (but perhaps not necessary) criteria for flexible tool use and its quite an objective (if difficult) test.
I read this in Beyond Boundaries by Miguel Nicolelis (good book although a bit long winded and fanciful) which should have some academic references.
Actually, Nicolelis's BCI work also has some relevance to self-recognition. You can put electrodes into a monkey's motor cortex, measure the neural activation associated with, say, arm movement and then decode those signals to control the motion of a robot arm (that the monkey is is not aware of) pretty well. However, if you show the monkey the arm and it is rewarded for moving the robotic arm, it often stops moving its own arm while continuing to use the disembodied arm (with pretty much the same motor cortex activity). I'd never thought of this in the context of awareness before, but suggests it is somewhat analogous to a mirror test and overcomes some of the limitations you mentioned. A fair bit of a work has been done around insect neural interfaces (probably more invasive and extreme than anything an ethics board would let you do to a mammal to be honest) and you might find that similar tests have been performed but not labeled as a self-recognition tests.
Thanks Gavin! I've added Beyond Boundaries to my reading list.
The potential connection between BCI and self-recognition is fascinating. Offhand, do you know any references for insect neural interface studies that might be comparable to the monkey example you describe?
The example that first springs to mind is the work of Kanzaki's group who study odour plume tracking in silk moths. They have made robots controlled by both a moths walking action (also a movie) and also by its measured neural activity. However, when doing electrophysiology on insects it is common to completely wax their body in place and amputate their legs/wings to minimize electrical noise caused by muscle movement (which they did in the moth case). I'd forgotten this, and it does make it a bit harder for insects to demonstrate self-awareness in a similar way to the monkeys. Still, it's recently become more common to make recordings from actively behaving insects, as active behaviour has been found to modulate many neural responses (such as optic lobe processing of visual motion), so some more relevant examples might have been published recently.