15
Do fish have feelings?

Michael Tye

It is obvious that some things in nature are conscious, for example, human beings. It is equally obvious that some things in nature are not, for example, blades of grass. What is not obvious is where to draw the line. The conscious/nonconscious division patently does not coincide with the living/nonliving one. So where does it lie? And how are we to decide? In this essay, I shall focus on the case of fish and one conscious state in particular, that of pain.

I want to begin by making some general remarks about how to proceed with respect to the question of animal consciousness. Consider the case of other human beings. What makes it rational for me to believe that you have similar experiences to me? Not the old argument from analogy, I suggest, but rather an inference to the best available explanation of your behavior. This is an application of a rule formulated by Sir Isaac Newton in his Principia to the effect that we are entitled to infer like cause from like effect unless there is defeating evidence. This rule is best seen, in my view, as providing the basis for rational preference instead of rational belief, but I shall pass over that for present purposes.

What is a defeater here? Well, suppose I find out that your head is empty and that you have only an organic exterior. Your movements are controlled by Martians. You are a Martian marionette. This new evidence defeats my entitlement to prefer the view that you have experiences and feelings like me, even though you behave in very similar ways.

Alternatively, suppose I find out that you have only a silicon chip in your head with a vast look-up table inscribed on it, a table that controls your every move. Again, I have a defeater.

These points can be applied to the case of fish. The idea that it is fine to eat fish is pretty commonly held. Kurt Cobain of Nirvana fame wrote “It’s okay to eat fish ’cos fish don’t have feelings” in a well-known song. This view is also presumably held by Kevin Kline in the famous British movie, A Fish Called Wanda. Kline wants to discover the location of some stolen jewels and so, in an effort to get Michael Palin to talk, he takes Palin’s beloved tropical fish out of their tank one by one and slowly eats them, as Palin is forced to watch. It is obvious that Palin thinks of the fish as creatures with feelings. He desperately wants them not to experience pain or fear or anxiety. But Kline couldn’t care less. For him, they are zombies (or at least they should be treated as if they are). Who is right?

Some philosophers and scientists think that a fruitful way to proceed on the question of animal consciousness is via an investigation of their metacognition. If animals behave in ways that indicate that they have a cognitive grasp on how things appear to them, and not just on how they are, then the obvious conclusion is that things really do appear to them in various ways. And if things appear to them, then the animals must be conscious of those things – they must experience them. This is a strategy proposed by Shea and Heyes (2010) and also by Allen and Bekoff (1997). What would count as evidence that an animal has a cognitive grasp on how things appear to it? A complex form of behavior providing such evidence would be using appearances to deceive other animals. A simpler form of behavior would be recognizing how something visually appears color-wise (where that appearance is different from the customary and real color of the thing) and matching the appearance to the real color of something else in order to get a reward.

This seems to me a worthwhile and important field of research. However, we should be clear on what it shows. If a positive result is obtained, then that is evidence that the animal is indeed conscious. But if a negative result ensues, what follows is only that higher-order consciousness has not been found (that is, awareness of how things appear – a second-order mental state). And that is perfectly compatible with the existence of first-order consciousness – feelings such as pain, experiences such as the visual experience of red, etc. As already noted, no one wants to claim that a one-year-old child cannot feel pain because it is incapable of cognizing its state as painful.1 That sophistication does not come until later. I return to the topic of higher-order consciousness below.

So, how should we proceed? Given my earlier remarks in connection with Newton’s Rule, here is my suggestion. Take the case of pain. Humans, upon encountering a noxious stimulus S, feel pain. The feeling of pain in humans causes a certain pattern of behavior B very roughly as follows.

Suppose we find that in fish, there is the same pattern of behavior B in response to noxious stimuli. Then it is rational for us to prefer the hypothesis that the feeling of pain causes B in fish to the hypothesis that some other cause is operative, unless we have further evidence that defeats that preference. Additional confirmation that the feeling of pain causes B in fish is the cessation or reduction in B, given morphine or other opiates, as is the case with us – again, unless there is defeating evidence.

Let us then take a quick look at the behavior of fish. First, though, it is worth pointing out that on the input side, teleost fish (that is, fish with bony skeletons) have nociceptors just as we do. Under a microscope, they look just like our nociceptors. These receptors in their skin respond to the same noxious stimuli as ours. Interestingly, shark lack nociceptors, as do other elasmobranchs (fish such as sharks that have cartilaginous skeletons). One consequence of this is that shark are able to feed on prey that would otherwise be noxious. For example, hammerhead shark prey on stingrays. These sharks have been found with as many as 96 stingray barbs embedded in their mouths (Rose 2002)! Yet sharks react in the same way as telelost fish to being caught on a hook. They struggle and try to get away. What seems to trigger their escape response is interference with free movement. For elasmobranchs, so far as I am aware, there is no behavior, the best explanation of which is that they feel pain.2 Let us then put the case of sharks (and stingrays) to one side.

Fish exhibit trade-off behavior. In one experiment, trout were trained to feed in a part of the aquarium where they subsequently got a shock to the flank. It was found that the number of feeding attempts decreased with increased shock intensity. However, with increased food deprivation, the number and duration of feeding attempts increased, as did escape responses as the zone was entered. A plausible hypothesis is that fish balance their need for food against the avoidance of acute noxious stimuli. We do this too. Think about the case of picking up a very hot plate, in the one case when it is full of food and you are very hungry, and in the other when the plate is empty. You are much more likely to hold on to the plate in the former instance even though doing so is causing you pain. Fish, it seems, are like us. Similar behavior is found in hermit crabs, I might add.

In another experiment, Elizabeth Sneddon, a British scientist, injected bee venom and also acetic acid (the latter being the main ingredient in the vinegar used in the United Kingdom with fish and chips) into the lips of trout while they were under anesthetic. Sneddon chose the lips since trout lips have polymodal receptors very like those found in human lips. When the trout came to, they rubbed their lips against the sides of the tank and the gravel on the bottom. They also sat on the bottom and rocked from side to side. Primates in a poor welfare state display rocking behavior too – as a sign of their having been in acute discomfort.

This suggests that the trout have been through an aversive experience. It is also interesting to note that trout take about three hours to start feeding again after they have been injected with acetic acid – roughly the amount of time human beings whose lips have been injected with acid take to stop feeling pain.

Sneddon also found a greatly increased beat rate of the opercula (the bony flaps covering the gills) in the trout injected with bee venom or acetic acid as compared to the others. This is usually taken as an indicator of stress, and Sneddon takes it to add further support to her hypothesis that the trout injected with bee venom and acetic acid feel pain. In general, the overall pattern of behavior fish produce is indeed similar to that we produce in response to the feeling of pain, as is their reaction to opiates. So, we should prefer the hypothesis that they feel pain too.

But is there defeating evidence here? In human beings generally, the experience of pain is generated by activity in regions of the neocortex (specifically, the somatosensory cortex and the anterior cingulate cortex). Fish, however, lack a neocortex. This neurophysiological difference makes a difference, so some say: fish cannot feel pain. What are we to say about this?

The claim that in humans, pain and other experiences require a neocortex, is widely accepted. For example, the American Academy of Neurology (AAN) asserts (1989):

Neurologically, being awake but unaware is the result of a functioning brainstem and the total loss of cerebral cortical functioning … Pain and suffering are attributes of consciousness requiring cerebral cortical functioning.

This does not seem to sit very well with the facts. It certainly seems to be true that adult humans who, later in life, come to lack a functioning cerebral cortex, are then in a vegetative state. But this is not always true for children born without a cerebral cortex. Recently, Bjorn Merker (2007), who spent several weeks with decorticate children and their families, said the following:

These children are not only awake and often alert, but show responsiveness to their surroundings in the form of emotional or orienting reactions to environmental events …, most readily to sounds but also to salient visual stimuli…. They express pleasure by smiling and laughter, and aversion by “fussing”, arching of the back and crying (in many gradations), their faces being animated by these emotional states. A familiar adult can employ this responsiveness to build up play sequences predictably progressing from smiling, over giggling to laughter and great excitement on the part of the child.

(p. 79)

There can be no doubt that these children are very impaired behaviorally. But in addition to apparently showing pleasure, they also sometimes apparently feel pain, by rubbing an area that has been banged or pinched. This is shown by facial expressions such as wincing, grimacing and flinching in 14% of the children; vocally in ways such as crying, screaming and yelling in 78% of the children; and body use such as wriggling, pulling away and startling in 4% of the children. Certainly, their behavior is nothing like that of the few children who have congenital pain insensitivity. These children ignore noxious stimuli and feel no pain from them, with the result that they behave as if nothing bad has happened, sometimes with dire consequences. One such child, Gabby Gingrass, poked out an eye and bit her gums down to the bone when she was teething. She also dislocated her jaw with no one being any the wiser, and an infection resulted.

So, at a minimum, it is not even clearly true that in humans a neocortex is needed for consciousness and pain in particular. What about other species? Well, here it is worth noting the case of birds. Birds lack a neocortex. Yet they engage in some very complex behavior similar in various ways to ours. So, what is going on in birds at a neurophysiological level? It has recently been proposed that there are homologous cells in bird and human brains (cells, that is, that share a common evolutionary origin) that mediate the behavioral similarities. But how can this be, if humans have a neocortex and birds lack one? This question becomes even more puzzling if we think of the neocortex in the way it is often described as a unique component of mammalian brains, something without a prior history.

The solution to the puzzle lies with the realization that the neocortex did not suddenly appear by magic in mammals. What happened with mammals was that certain sorts of cells present in non-mammalian brains and around for hundreds of millions of years were grouped together into layers to form the laminar structure of the cortex. This is what is genuinely new. But the constituent neuron types and the microcircuitry aren’t new – or so at least it has recently been hypothesized, dating back to earlier speculation by Karten in the 1960s (Karten and Hodos 1967; also Karten 1997).

The relevant cells for birds are preserved in a structure of a vastly different shape from the neocortex, known as the dorsal ventricular ridge (DVR). The cells in the DVR share the same physiological properties as the cortical cells (Dugas-Ford et al. 2012). It has recently been hypothesized that a similar structure of cells is to be found in the forebrain of fish, too (Ito and Yamamoto 2009).

The upshot: the fact that fish lack a neocortex does not, in and of itself, defeat preference for the simplest hypothesis, namely that fish, like us, feel pain or something very like it. What defeats this potential defeater is the following: a) the case of birds shows that cells homologous to those in the neocortex can be present without a neocortex, and at least according to some scientists, fish have such cells and similar microcircuitry in their forebrains; and b) it isn’t clear that such a structure in the brain is needed anyway, even in the case of human beings, given the example of some decorticate children – indeed, it seems not to be the case.

So, I think that it is rational for us to prefer the view that fish have feelings to the view that fish do not. Of course, the feeling I have focused on, that of pain, is phylogenetically fixed. This is not true of so-called “secondary emotional experiences”, for example, feeling insulted or feeling remorseful. No claim is being made that fish are capable of undergoing these experiences.

There is one further line of resistance I want to mention to the claim that fish have feelings. It goes as follows.

So,

Were this argument sound, it would provide a defeater to the inference from fish behavior to the existence of fish feelings and pain in particular (in the case of B).

The argument is unpersuasive, however. It is not at all obvious that fish lack concepts, even if they lack our concepts. After all, they engage in some pretty intelligent behavior. And there is evidence that they can reason transitively (Grosenick et al. 2007).

The first premise is also highly dubious. As noted above, the feeling of pain (and many other feelings) are fixed in human beings by their biological makeup. They are naturally taken to be phylogenetic states, liability to which does not require learning. So, for many feelings, it doesn’t seem that concepts are required.

There is one well-known theory of the nature of the subjective character of experience that forges a direct link between feelings and concepts, namely the higher-order thought theory (Rosenthal 1986). On this view (the HOT view), for a creature to feel pain, its pain state must be accompanied by a higher-order thought to the effect that it is in pain, where this thought is arrived at non-inferentially. Some versions of the higher-order thought theory make more modest claims. But this version is wildly implausible. As noted earlier, small babies feel pain, yet they are not capable of thinking to themselves that they are in pain; for they lack psychological concepts generally, and the concept PAIN in particular.

Another problem is that the HOT view is committed to there being pains that do not feel any way, namely pains that do not have accompanying higher-order thoughts. On the face of it, this erroneously supposes that if a pain is unconscious – if, that is, it is a pain of which its subject is not conscious (and so not thinking about non-inferentially) – then it is a pain without any felt character. But surely pain is a feeling; and a feeling must have a felt character, even if it is a feeling of which its subject is not conscious (and in that sense, an unconscious feeling).

In conclusion, it is interesting to compare and contrast the case of fish with that of insects. Eisemann et al. (1984) comment in a review of biological evidence for pain in insects:

No example is known to us of an insect showing protective behavior towards injured parts, such as by limping after leg injury or declining to feed or mate because of general abdominal injuries. On the contrary, our experience has been that insects will continue with normal activities even after severe injury or removal of body parts. An insect walking with a crushed tarsus, for example, will continue applying it to the substrate with undiminished force. Among our other observations are those on a locust which continued to feed while itself being eaten by a mantis; aphids continuing to feed whilst being eaten by coccinellids; a tse-tse fly which flew in to feed although half-dissected; caterpillars which continue to feed whilst taccinid larvae bore into them; many insects which go about their normal life whilst being eaten by large internal parasitoids; and male mantids which continue to mate as they are eaten by their partners.

(p. 166)

Eisemann et al. (1984) also point out that insects do not respond to pain by ceasing to move or protecting injured parts in the way that mammals do. In general, they do not react to treatment that would undoubtedly cause severe pain in mammals. In these respects, insects are unlike fish.3

Notes

1 Well, almost no one. Carruthers (2000) says that consciousness goes with the capacity to make the appearance-reality distinction. On one reading of this claim (a cognitive one), his view has bad consequences for newborn babies!

2 There are reports from whalemen of sharks that they have split in two continuing to feed – likewise for sharks that have been disemboweled by other sharks attacking them. Apparently their fatal wounds do not cause them to feel pain.

3 The issue of insect pain is not quite as clear-cut as Eisemann et al. suggest. For more on this and on the question of animal consciousness generally, see Tye 2016.

References

Allen, C., and Bekoff, M. 1997 Species of Mind: The Philosophy and Biology of Cognitive Ethology. Cambridge, MA: MIT Press.

American Academy of Neurology. 1989 “Position of the American Academy of Neurology on certain aspects of the care and management of the persistent vegetative state patient,” Neurology 39: 125–126.

Bateson, M., Desire, S., Gartside, S., and Wright, G. 2011 “Agitated honeybees exhibit pessimistic cognitive biases,” Current Biology June 21; 21(12): 1070–1073.

Carruthers, P. 2000 Phenomenal Consciousness: A Naturalistic Theory. Cambridge: Cambridge University Press.

Dugas-Ford, J., Rowell, J., and Ragsdale, C. 2012 “Cell-type homologies and the origins of the neocortex,” Proceedings of the National Academy of Sciences 109(42): 16974–16979.

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

Fields, H. 1999 “Pain: An unpleasant topic,” Pain Supplement 6: S61–S69.

Grosenick, L., Clement, T., and Fernald, R. 2007 “Fish can infer social rank by observation alone,” Nature 445: 429–432.

Ito, H., and Yamamoto, N. 2009 “Non-laminar cerebral cortex in teleost fish?” Biological Letters 5: 117–121.

Karten, K. 1997 “Evolutionary developmental biology meets the brain: The origins of mammalian cortex,” Proceedings of the National Academy of Sciences 94, 2800–2804.

Karten, H. J., and Hodos, W. 1967 A Stereotaxic Atlas of the Brain of the Pigeon (Columba livia). Baltimore: Johns Hopkins University Press.

Merker, B. 2007 “Consciousness without a cerebral cortex: A challenge for neuroscience and medicine,” with commentaries, The Behavioral and Brain Sciences 30: 63–134

Newton, I. 1687 Principia Mathematica.

Rose, J. D. 2002 “The neurobehavioral nature of fishes and the question of awareness of pain,” Reviews in Fisheries Sciences 10(1): 1–38.

Rosenthal, D. 1986 “Two concepts of consciousness,” Philosophical Studies 49: 329–359.

Shea, N., and Heyes, C. 2010 “Metamemory as evidence of animal consciousness: The type that does the trick,” Biology and Philosophy 25: 95–110.

Shewmon, A., Holmes, G., and Byrne, P. 1999 “Consciousness in congenitally decorticate children: ‘Developmental Vegetative State’ as self-fulfilling prophecy,” Developmental Medicine and Child Neurology 41(6): 364–374.

Sneddon, L. U. 2003 “The evidence for pain in fish: The use of morphine as an analgesic,” Applied Animal Behaviour Science 83(2): 153–162.

Sneddon, L. 2012 “Pain perception in fish: Evidence and implications for the use of fish,” Journal of Consciousness Studies 18: 209–229.

Sneddon, L. U., Braithwaite, V. A., and Gentle, M. J. 2003 “Do fishes have nociceptors? Evidence for the evolution of a vertebrate sensory system,” Proceedings: Biological Sciences 270: 1115–1121.

Tye, M. 2016 Tense Bees and Shell-Shocked Crabs: Are Animals Conscious? Oxford: Oxford University Press.