Key, Brian (2016) Why fish do not feel pain. Animal Sentience 3(1) (after reading the article, read the abstracts of some of the commentaries too, for contrary vuews).
Only humans can report feeling pain. In contrast, pain in animals is typically inferred on the basis of nonverbal behaviour. Unfortunately, these behavioural data can be problematic when the reliability and validity of the behavioural tests are questionable. The thesis proposed here is based on the bioengineering principle that structure determines function. Basic functional homologies can be mapped to structural homologies across a broad spectrum of vertebrate species. For example, olfaction depends on olfactory glomeruli in the olfactory bulbs of the forebrain, visual orientation responses depend on the laminated optic tectum in the midbrain, and locomotion depends on pattern generators in the spinal cord throughout vertebrate phylogeny, from fish to humans. Here I delineate the region of the human brain that is directly responsible for feeling painful stimuli. The principal structural features of this region are identified and then used as biomarkers to infer whether fish are, at least, anatomically capable of feeling pain. Using this strategy, I conclude that fish lack the necessary neurocytoarchitecture, microcircuitry, and structural connectivity for the neural processing required for feeling pain.
Only humans can report feeling pain. In contrast, pain in animals is typically inferred on the basis of nonverbal behaviour. Unfortunately, these behavioural data can be problematic when the reliability and validity of the behavioural tests are questionable. The thesis proposed here is based on the bioengineering principle that structure determines function. Basic functional homologies can be mapped to structural homologies across a broad spectrum of vertebrate species. For example, olfaction depends on olfactory glomeruli in the olfactory bulbs of the forebrain, visual orientation responses depend on the laminated optic tectum in the midbrain, and locomotion depends on pattern generators in the spinal cord throughout vertebrate phylogeny, from fish to humans. Here I delineate the region of the human brain that is directly responsible for feeling painful stimuli. The principal structural features of this region are identified and then used as biomarkers to infer whether fish are, at least, anatomically capable of feeling pain. Using this strategy, I conclude that fish lack the necessary neurocytoarchitecture, microcircuitry, and structural connectivity for the neural processing required for feeling pain.
ReplyDelete“The strategy adopted here involves three steps: (1) identify which brain region is directly responsible for the feeling of pain in humans; (2) define the neuroanatomical features of this region that allow the neural processing that generates felt pain; and (3) assess whether fish have similar structural features that can then be used to make inferences about whether fish can, or cannot, feel pain.”
“It is not yet known how these self regulatory oscillations may generate the pain sensation; however, strong functional interconnectivity between discrete cortical regions is essential for pain.”
In the article, Key creates a 3 pronged framework detailing a litmus test for whether some organism can feel pain, the target of this article being fish. Essentially, this test rests on determining neural structures in human brains that appear to be correlated with pain (Key asserts that such areas are “directly responsible for the feeling of pain in humans”), and asking if the target organism(here, fish) has analogous structures. In the absence of such structures, Key is confident in declaring that fish don’t feel any sort of pain. There seem to be some serious problems here, the most glaring being the assumption that the brain regions he designates in humans represent the only path towards feeling, and therefore in their absence, feeling couldn’t arise. It doesn’t seem like much of a stretch to say that the issue of what causal mechanisms lead to feeling pain is underdetermined; even if there existed an unbelievably detailed and airtight formalism of how these structures causally generate felt pain in humans, that wouldn’t exclude an entirely different (or even somewhat different) pathway to feeling elsewhere. Key is mapping human structures onto fish and using their not lining up to justify his bold claim -- an irony considering Key’s insistence that the problem he’s addressing comes from a tendency to over-anthropomorphize non-human animals.
I think the argument of the author is build around not only the brain regions necessary for “feeling pain”, but also about specific features that the neural structures must have. I think that the author could argue back and say that the neural systems of fish and human are quite similar in their functioning (it’s a bit tautological to say that they are made of neurons, but one could imagine a very different adaptative biological system that function like neurons, but without neurons and glia), and that given that they both have neurons, the neural system need the features described (according to the empirical evidence Key provided) to “feel pain” with a neural system.
DeleteI am not sure if Key is right about his claim, but I think there is food for thought. To take another example, the feeling of vision is probably very different from species to species. Nevertheless, it could be argued that to see colors the system must have a way or another to process the wave length of light. But could bats “see” colors by echolocation, in their own bat-way? It could be argued that felt-states analogous to human colors could be used to describe the way bats perceive their environment: a “color red-like” for sharp angles, a “color blue-like” for flat homogenous surfaces, etc.
There is no way to prove that the felt state associated with echolocation is not “like” the colors that humans see. I think that at the end, it relies on how we want to categorize feelings: is color vision restricted to light-related perception? And echolocation would be restricted to textures and speed?
Key seems to be confident that pain must involve the physiological features he describes. Maybe he is wrong, and other ways of “feeling pain” for fishes are possible and analogous enough to human pain to stop hurt unnecessarily these (probably sensitive) beings.
The point you bring up about the possibility of bats seeing color with echolocation as being similar to humans "seeing" color reminds me of humans assigning the label of language to the sign language monkeys have been taught. In the same way, saying bats could see color because we do not understand echolocation is a more simplistic way for us to approach the issue that I do not believe is necessarily the correct way to do this.
DeleteI do agree however that Key's questions of neural circuitry in humans used to look at pain in fish is food for thought while I am not entirely sure he answers every concern about his premises particularly the points Kevin brings up.
I do agree however that Key's questions of neural circuitry in humans used to look at pain in fish
Yes, Key is being anthropocentric in insisting that pain can only be generated by the brain in the human way.
DeleteLazy evolution does not "optimize," coding the "best possible solution" to an adaptive problem. It just codes enough to survive and reproduce successfully across generations ("satisficing"). Sometimes evolution will chance upon a different solution to the same adaptive problem in different species (because the solutions are underdetermined). A famous example is the two forms of vision -- retinal vision, as it evolved in vertebrates, and ommatidial vision, as it involved in invertebrates. And sometimes the same ancestral structure will evolve somewhat differently in later descendent species. (Our memory system is partly descended from parts of the brain that had been devoted to smell.)
For species as close to us and as similar to us in their appearance and behavior as fish, their behavior alone, when they are hurt, should be enough to dispel OMP doubts (except for those with interests vested in the fishing industry -- or the taste of fish, or the fun of angling -- who tend to use the OMP as evidence that their target species is insentient). By Key's criteria not only do fish not feel pain (they just "process damage input"), but they also do not see (they just "process optical input"), all of it "robotically."
“Although we will never be fish, I contend that we can strongly defend the thesis that fish do not feel pain based on inferences derived, in this case, from experimental evidence from neuroscience and evolutionary biology investigations on humans and nonhuman animals.”
ReplyDeleteReading this article reminded me a lot of what I learned about functionalism, which (although mostly disproved) posits that mental states are functional states of a Turing machine. A Turing machine can be pretty much anything as long as it does all that a Turing machine can do. Therefore, if two organisms have different nervous systems but are in the same functional state (despite the different “hardware” that is implementing this state), they could be said to have the same mental state. For example, say we are looking at pain in both an octopus and a human. Even though the nervous system of an octopus is much different than that of a human, they are both feeling pain if they are in the same functional state indicating pain.
While it is true that you can never be a fish, you can conclude that a fish DOES feel pain if you see it through a functionalist lens. Although functionalism has many drawbacks, I think it allows you to look at felt states with less of a biological view. Key’s argument is based on the physiological features and neural structures that underlie human nociception, but since we are mostly ignorant when it comes to the hard problem of feeling/felt states, his argument is not convincing.
I like that you recognize the limitations of the functionalist view, and I agree with you that functional states could be good enough information for us to draw inferences from until we have better scientific methods of proving or disproving the theory that fish feel pain. I think one of the main drawbacks of this method is something that the author addresses in the article, which relates to anthropomorphism. I am on the side of Key in the sense that I think it’s important to be critical and cautious when it comes to anthropomorphism. If we hold such a view too strongly, we could come to say something like “Because plants turn in direction of the sun, we know they /feel/ and /like/ the sun.” But this would be an error and it would be to attribute too much of a human-like consciousness to plants. In the case of fish, though, I am a little confused, because there were moments when I saw traces of anthropomorphism in Key’s statements, and I was, like you, not convinced by his argument. His counterargument to “fish pain” and anthropomorphism was especially weak in my opinion.
DeleteI think I disagree with you and Key here – I think the fear of ‘anthropomorphizing’ in the science community (or of refusing to ‘mind read’) has been a pretty big handicap in terms of understanding animal cognition.
DeleteI read a book last year (can’t remember the name) specifically about fish cognition, and how the scientific community has been seriously held back and biased in that it generally is more hesitant to accepts the presence of pain in species whose facial expressions differ from ours (such as fish) – I think it argued that scientists more commonly fear “anthropomorphizing” species whose facial expressions aren’t recognizable to us – which is guilty of the very thing they fear (that is bias due to anthropomorphizing)! The book I read talked about overwhelming evidence for fish pain (and pain and consciousness in other species), that is often denied by the scientific community, as the default assumption is that other animals are zombies, until otherwise proven. (But as discussed in class – there’s no way to measure what another being feels without being that other being). To me this seems like a ridiculous anthropocentric biased stance on the zombie problem. Mind-reading is a useful tool. Also, it seems that lazy evolution would select for feeling and pain, as this is a quick way to program for beneficial behaviours.
Emmanuelle, a Turing machine cannot be anything in the universe, it can simulate just about anything in the universe, if you find an an algorithm (squiggle/squoggle) that formally encodes enough of the relevant properties of the thing you are simulating (e.g., a vacuum-cleaner, which is not a Turing Machine, which in turn is only computationally equivalent to a vacuum cleaner if implementing the right software). Then the TM's formal properties can be interpreted as the real thing's properties and the TM can be used to predict and explain further properties of the real thing, and can even be piped into a virtual reality that fools our senses to appear as if it is the real thing.
DeleteComputationalism (which is functionalism about felt states and TM states, i.e., computational states) is what Searle's argument and the symbol grounding problem refute.
Distinguish computation (formal symbol manipulation) from physiology (which is what Key is talking about). Computationally simulated physiology would not feel pain any more than computationally simulated water would be wet.
We don't need functionalism to tell us that the brain is almost certainly what generates feeling, somehow. (The hard problem is explaining how and why it does so, not whether it does so).
Marianne, see above: yes, Key's human-brain-centred aguments against pain in fish is anthropomorphic (actually mammaliomorphic).
I think our everyday solution to the OMP with people (which is based on their behavior, not their physiology) is adequate also for fish, indeed for all vertebrates and most invertebrates. The OMP really only starts with the very simplest invertebrates, and then microbes and plants (and includes also in vitro organs, single cells and robots). (It would be Trumpian trollery to invoke the OMP only selectively, as needed for protecting one's oen commercial interests or taste/sport preferences, declaring the obvious evidence for sentience to be "fake news" because of OMP.)
Erin, what you read was probably Balcombe's book, "What a fish knows,". He was also here at the 2018 Summer School on the other-minds problem in other species.
But if you really know how and why "lazy evolution would select for feeling and pain, as this is a quick way to program for beneficial behaviours" then you have a solution to the Hard Problem and you should share it with us. For, on the face of it, all lazy evolution would need would be damage-detection and learning...
(That said, just as it's almost certain that the brain generates feeling, it's almost certain that feeling evolved because of adaptive advantages: the unanswered question in both cases remains: how and why?)
“Given the proposition that the above organizational principles are necessary for conscious neural processes of pain”
ReplyDeleteThe author keeps repeating the notion of “conscious pain processing”. However, they do not seem to make a distinction between conscious neural processes of pain and unconscious neural processes of pain. Perceiving pain is not the same thing as feeling of pain. So, I think what they should say here is “neural processing of pain”.
“These observations suggest that the identification of a painful stimulus is not dependent upon the normally unpleasant feeling arising from the stimulus. It indicates that pain can be distinguished by the intensity and the nature”
This clearly shows that we are not looking for the origin of the “feeling” (which is a part the hard problem). But we are looking for where specifically the pain originates from. Even though we can conclude that fish doesn’t have the circuitry (that is similar to the human circuitry) to have nociception, we don’t know if they have feelings (since we haven’t solved the hard problem) and the OMP still stops us from understanding what is going on in their brain.
When we are just abstractly looking at certain type of function, (like pain) we leave “feeling” out of the question. However, even though their evidence might be showing that fish does not have the similar nociceptor structure as humans, the fact that they completely leave out feeling, I think, invalidates their conclusion that fish do not feel pain. I think they need to be able to explain pain in relation to feeling, then they would have to solve the hard problem to explain the absence of pain.
From another perspective, we wouldn’t want to kill a human with no pain reception when there is no reason to do so. But, I would like to ask this question from a moral perspective. Imagine 5 people on an island without any food, we know that one of them has no pain perception whatsoever. However, other than pain, that person looks normal; it seems to have feelings as all the others. Is it any better to choose to kill this particular person over the other 4 people on the island? I think the answer to the question is no, because it still feels something; even though it doesn’t perceive pain, we cannot conclude anything about the feelings related to pain. I know that inferring the state of a fish is more difficult, since humans infer each other’s states easier, but this question still applies to fish because fish doesn’t seem to be just zombies.
Lastly, I don’t think that looking at humanlike pain brain anatomy is enough. If some stimuli leads to a change (temporal or permanent) in the subject, that stimuli should be in the category of things that can affect the subject. It is known that after painful stimuli, fish loses normal behaviour. (Braithwaite, Victoria A. and Droege, Paula (2016)) In the end, a fish is not totally indifferent to painful stimuli even if it has pain perception or not.
De-Weaseling Exercise I
DeleteOnur, be careful about weasel words (WWs)!
You made "a distinction between conscious neural processes of pain and unconscious neural processes of pain"
A pain is just a feeling, so let's call it a feeling. And then get rid of the WW "conscious":
What is "a distinction between felt neural processes of feeling and unfelt neural processes of feeling"?
What we feel is the feeling. Presumably neural processes cause the feeling.
So now we have "a distinction between felt neural processes that cause feeling and unfelt neural processes that cause feeling."
But, thus de-weaseled, this no longer makes sense: We don't feel neural processes, we feel feelings.
So one way to make sense of the first part, "felt neural processes that cause feeling," is that it is redundant. What you meant by "felt neural processes" was "neural processes that cause feeling." Fine. Then let's drop the redundancy and just use "neural processes that cause feeling." Done. Now what about the second part?
"unfelt neural processes that cause feeling": "unfelt neural processes" would normally be alright, because all neural processes are unfelt. But you introduced a (reasonable) shorthand for neural processes that cause feeling: We can call them "felt neural processes" (keeping in mind what it's shorthand for). And then "unfelt neural processes" would mean "neural processes that do not cause feeling."
But then what about the second part of "unfelt neural processes that cause feeling"? If we unpack the shorthand, it becomes "neural processes that do not cause feeling that cause feeling."
And that becomes self-contradictory, and hence meaningless.
But this is what happens when we let weasel-words slip in!
De-Weaseling Exercise II
DeleteThe same is true with "perceiving pain is not the same thing as feeling pain": what on earth is the difference? "perceiving" is merely another weasel word for "feeling" here.
So you say that Key, instead of saying his own weasely "conscious processing of pain" (which, once de-weaseled, and substituting generic "feeling" for "pain" to show that this applies to any feeling, would become the redundant "felt processing of feeling," again ambiguous about whether it means the processing itself is felt, rather than just that the processing causes the feeling, as above) for which you propose "neural processing of pain," which again just means the neural processes that cause the feeling of pain (or any feeling).
I spelled all this out for you (and everyone) to show that no answers or clarification of anything comes from using WWs or accepting others' WWs. All they do is cloud an already hard (but simple) question, which, completely de-weaseled is and remains: how and why do neural processes cause feeling?.
Same is true for the OMP "How can we determine whether or what another feels?" "Do fish feel (pain)?"
In your comment you are mixing up HP and OMP. Nociception is the Easy Problem. It just means damage-sensing. It is meant to be neutral about whether the organism feels the damage (or anything) or just detects and responds to it. Nociception can be insentient (unfelt); pain cannot be.
In most cases the nociception also causes a feeling of pain, so the distinction is usually spurious; just scientists scientistically making a display of hard-headed objectivity when our usually trustworthy mind-reading, our OMPscope, tells us otherwise. But there is no such thing as "unfelt pain"; only feelingless damage-detection. There is no "explaining pain in relation to feeling"? That would be explaining feeling in relation to feeling.
Your island with the person who feels other things but does not feel pain is again incoherent. What else does he feel? fear, anguish, sorrow? Or nothing negative at all? Only neutral and pleasurable feelings? These island examples are always otiose, if not tendentious, but what is the point? Finding the one it would hurt the least to die? What about their ages, their characters, their states? And what light does any of this cast on either the sentience of fish or the needless hurting and killing of nonhuman species?
I guess I wasn’t clear with my writing. I totally agree with you. Some of what you are telling is actually what I tried to tell about this paper. I totally reject their weasel words. That is why I tried to deconstruct Key's idea about "conscious neural processes of pain" because in the article, they speak as though there is a distinction between conscious and unconscious neural processing, which I reject. Hence, I suggested that they should not use “unconscious/conscious”, they should just say neural processing of pain. And this neural process, as you also mentioned, cause feeling (either pain or any feeling).
Delete“The same is true with "perceiving pain is not the same thing as feeling pain": what on earth is the difference? "perceiving" is merely another weasel word for "feeling" here. “
Here I realize that I meant to say “perception of pain” not perceiving pain because clearly perceiving is a weasel word. And, what I mean by perception of pain is the neural processing of pain (ie. damage sensing). So, I tried to mean that the neural processing of pain (in other words perception of pain), is not the same thing as the feeling of pain (or any other feeling for instance). Neural processing is the easy problem, whereas the feeling of pain (or any other problem) is the hard problem. Even though, we see a causality between the doing and the feeling of doing, this doesn’t explain how to solve the hard problem.
So, Key’s weaseling around about the “conscious neural processing of pain” is just non-sense, it doesn’t teach us anything. And I want to repeat, I think they totally leave out the question of feeling in the paper, and that by itself invalidates their idea, because hard problem is still not solved.
“There is no "explaining pain in relation to feeling"? That would be explaining feeling in relation to feeling.”
I agree, again I thought I could use the short word “pain” to describe “neural processing of pain”. Hence, rest of my writing became a snowball of unsound arguments for you. Sorry about that.
For the island scenario, we know that he doesn’t have the damage sensing. However, we do not know if he has feelings or not. Same applies to all other individuals on this island, we cannot know if others have feelings or not (OMP). So even though that individual doesn’t have damage sensing, we cannot say anything about his feelings. Hence, I think in terms of picking anyone on the basis of damage-sensing abilities would not be the best choice since they probably all have feelings.
"And what light does any of this cast on either the sentience of fish or the needless hurting and killing of nonhuman species?"
So, it, I think, shows that sentience is about the OMP, not about damage-sensing and what not. And since we cannot solve the OMP either with fish, or other nonhuman species, killing a fish/a non human-animal is killing a most-probably sentient animal.
Onur, pain is not an input, it's a response. The nervous system does not process pain, it processes sensory input, and when the input is damaging (nocive), the nociceptors may produce a feeling of pain. The damage can also be unfelt, or felt as other than pain.
DeleteSame for "perceiving pain" vs. "perception of pain": either way it means feeling pain. And if pain is not felt then it is either feeling (or not feeling) sensory effects of damage.
But I agree that on the substantive rather the terminological points, we agree.
“It has been argued here that it is possible to logically and confidently propose the strong inference, based on principles of evolutionary biology and neuroscience, that fish do not feel noxious stimuli.”
ReplyDeleteI believe Key brings up some interesting points that need to be considered in the explanation of pain and consciousness and in the addressing of the hard problem for nonhuman organisms. However, I am not entirely sure that he answers all of the concerns his premises bring about and I was not entirely satisfied with his response to counter-arguments as well. Especially the argument that fish feel “fish pain” and not “human pain” I don’t believe was very well addressed. I am also curious concerning the above quote, what Key would say is the evolutionary reason that fish evolved in such a way as to not feel pain.
Whether "fish pain" or "pheasant pain," pain hurts (and that's what this is about).
DeleteIf anyone could explain how and why some species (that can do similar things) feel pain and others are insentient zombies, then they would have solved the HP. But the OMP stands in the way.
(Plants do almost nothing but grow; and microbes can do much less than species with nervous systems can do. So they already differ on the Easy Problem.)
"Since we can't prove other minds problem either way, do we just accept that they do because more evidence in that direction, like physics with quarks?" - from Animal sentience: the other minds problem. Note: I read Key's work, then yours, then the counterpoints to Key's.
ReplyDeleteTo connect the two readings, this is what Key was saying about giving animals the benefit of the doubt, despite legit neurological differences in humans and, because it's the topic, fish. Monkeys at least have similar firing in homologous regions as us, but fish just don't. So, at the very least, fish don't feel in the same sense that we do. But if they don't feel like we do, aren't we kind of changing the definition of feeling? Either an organism feels in the same way humans do or it doesn't, and it (as an organism) just has reflexive reactions to stimuli.
However, as I read Balcomb's "Cognitive evidence of fish sentience" and Braithwaithe and Droege's "Why human pain can't tell us whether fish feel pain" I had no choice but to accept that fish do feel pain. I'm not sure how Key let these examples slip through the cracks as he was devising his arguments, but I'm glad I read the counterpoints. One question I do have is specific to the trout, in Braithwaithe and Droege's paper, that exhibited impaired avoidance of a novel object when treated with a noxious stimulus, but able to avoid novel objects when treated with an analgesic. This is just my simple understanding, but the example in my head is someone avoiding bending over due to back pain–repression of an action–but not being afraid to beind over without that same pain. The trout example seems like they do something when in pain and don't when not in pain. Is it that the pain is stifling their ability to realize they should avoid a novel object? Or the action repression is avoidance of novel objects? If the latter is the case, the human equivalent, to me, would be not limping when you are healthy but limping when injured. Is this equivalent to the trout as well?
Good questions, but they are really about what trout feel, not whether they feel. (Both are OMP questions, but I don't think it's justified to keep hurting trout to get a better idea of exactly what their pain feels like. Do you?)
DeleteHi Gabriel!
Delete“So, at the very least, fish don't feel in the same sense that we do. But if they don't feel like we do, aren't we kind of changing the definition of feeling? Either an organism feels in the same way humans do or it doesn't, and it (as an organism) just has reflexive reactions to stimuli.”
I understand the point being made about the physiology of a fish not being similar to that of humans. This is evident. I don’t quite agree with the statement you make about how we are changing the definition of feeling. As professor Harnad pointed out in another Skywriting, whether it is “fish pain”, “human pain” or any other type of pain, pain hurts. Not knowing whether pain is or is not felt exactly the same way falls under the Other Minds Problem because we are unable to become a fish and find out ourselves whether fish are feeling. I do not believe we need to alter our definition of feeling just because we decide that fish are able to feel, and fish are not similar to we are physiologically.
I also agree with professor Harnad. I also do not believe that it is justified to hurt trout just to get a better idea of exactly what their pain feels like. This is an ethical and moral issue that we humans have because we are able to feel. As someone who knows what it is like to feel pain and to hurt, I would never want to inflict pain on other feeling beings. If we replaced the trout with a human, it would be obvious that this is wrong.
Brian Key’s arguments supporting the claim that fish do not feel pain are based on a pervasive dichotomy present in the pain field. In pain research, conscious processing of noxious stimuli is referred to as “pain” while unconscious processing of noxious information is referred to as “nociception”. Consciousness in that sense does not represent the ability to feel and have felt states as we have seen in this course. Consciousness in the empirical sense used here represents a synonym for a patient’s arousal, awareness and wakefulness. My experience as a cognitive science student in a clinical neuroscience lab has made me realized that the distinction between these terms is not really relevant as far as the medical community is concerned. However, the distinction between pain and nociception, i.e. conscious vs unconscious processing, constitutes the chore of Key’s argumentation.
ReplyDeletePain management is strongly dependent on a patient’s ability to report their pain. Self-report is in turn strongly dependent on linguistic capacities, a recurrent theme throughout this course. The frontal, insular and cingulate regions highlighted by neuroimaging studies as neural correlates of pain all come from experiments in which patients were asked to self-report their pain. This can be done verbally or with the use of a visual analog scales displaying happy to sad faces or numbers from 0 to 10. Indeed, the only way to test the validity and relevance of these neural correlates is to obtain pain ratings directly from the subjects. This constitutes the gold standard of pain research. From these kinds of experimental designs are derived affirmations of the type: “when X level of pain is felt in subjects, an average activation of Y Hz is observed in the Z region”. All of what pain neuroimaging really does is looking at brain regions that are active when patients or participants report pain. Key proposes that, by looking at the presence of these brain regions in other species, we can conclude which feel pain and which do not. That’s a ridiculous approach, especially considering that NO brain region is strictly specialized in pain processing.
It’s not hard to see how this reliance on self-report in pain management disadvantages non-verbal and physically limited patients. Indeed, any non-verbal organism is disadvantage by such definition of “conscious” pain processing. Key claims that “If “fish pain” existed, it would need to be consciously processed using at least the minimal neural architecture described [in humans]”. This implies that the brain regions mentioned above are essential for consciousness to emerge. Since when has consciousness been localized in the brain? This is again confusing the when and where questions of mental activity with the how and why questions of the emergence of feelings. I think that the clinical definition of consciousness, heavily dependant on linguistic capacities and widely used in sedated or anesthetized patients, misleads some neuroscientists into believing that pain is a human experience.
"Consciousness" means feeling. Verbal reportability (unique to humans) means verbal reportability. They're not the same, although there's only something to report if you feel something.
DeleteThe distinction between nociception (damage-detection) and pain is almost certainly empty. Where there's nociception, there's pain. And pain is felt ("consciously" -- to toss in another WW).
The point of assigning the Key article to show you where the real OMP is not. Of course fish feel, and feel pain. The doubt begins much closer to the animal/plant boundary and the boundary is probably the presence of a nervous system.
“Moreover, some vertebrates, such as fish, may lack the neural machinery or architecture to consciously experience (i.e., to feel) noxious stimuli as painful (Key, 2015a).”
ReplyDeleteI disagree with Key. We can’t conclude that fish lack the neural machinery or architecture to consciously experience pain, we can only conclude that fish lack the same neural machinery or architecture that we have to consciously experience pain. It is kind of like underdetermination; it may be possible for them to achieve the experience of pain through different neural mechanisms than us, in fact, it wouldn’t even be shocking, considering the huge gap between a human body and a fish’s body. It’s no surprise that fish, who are so different from us morphologically and live in different environments, have different brain structures from us.
Also, it ignores the other-mind problem. We will never know whether fish experience pain or not, so it would be best to operate under the assumption that they do. The cost of us being wrong would not be very significant. If we were to only act as if fish felt pain in the presence of evidence that would be unreasonable. There is no way to ever have scientific evidence, because that would be solving the hard problem of consciousness which is unsolvable.
I agree with you, Johnny.
DeleteJust because they don't have the same neural structures we do, doesn't mean there aren't "different models that lead to the same output" - ie his argument ignores the possibility of underdetermination.
Also, the cause of pain he discusses in human brains - is that truly the cause of pain, or the causal mechanism of our reactions to painful stimuli? Because all that shows is that we have structures that make us react a certain way - (ie we could still be zombies). Fish react to 'painful' stimuli with distress. So either the distress demonstrates that both humans and fish feel pain, or it demonstrates nothing about feeling at all, and only that we have neural structures that cause us to react in a certain way when presented with ‘painful stimuli’.
Hi Johnny,
DeleteI also agree with your comment. Additionally, according to Key, "An alternative hypothesis as to why fish do not need the architecture of a human
cortex to feel pain is because fish do not feel ‘human pain’, but instead they feel ‘fish pain’. Consequently, a cortex-like architecture is not considered necessary for fish to feel pain."
Although Key does point this out, I question the argument that a cortex-like architecture to feel pain is even necessary for humans. As we have discussed in the class before, where (and when) does not tell us why and how.
Moreover, I agree with your point that we won't know if a fish can feel pain or feel at all and "it would be best to operate under the assumption that they do." But unlike what Key has proposed, I don't think that when people assume that other humans or non-human animals feel pain, it has anything to do with the "cortex-like architectures" they may or may not share with them.
Key explains that noxious stimuli are "stimuli that cause or have the potential to cause body tissue damage" (Key, 2016:2). He claims "fish may lack the neural machinery or architecture to consciously experience noxious stimuli as painful" (Key, 2016:2). He explains that pain in humans in generated in the brain, specifically by neural activity in the cerebral cortex. Key claims that because fish do not have similar architecture and structures that are thought to be the cause of our pain production, they do not feel pain.
ReplyDeleteMy question is why would fish have to feel pain in the same way that we do? In Key's conclusion he states, "What then do noxious stimuli feel like to a fish? The evidence best supports the idea that they don’t feel like anything to a fish." (Key, 2016:17). But just because noxious stimuli don't result in the same neural things for fish as they do for us doesn't mean that they don't result in some sort of *feeling* or *pain*. Pain does not need to come about in the same way as it does for us.
I would like to look at this through a concept of pain that David Lewis suggested. Lewis said that the causes and effects of pain may differ (but nonetheless still be pain). In Lewis' paper "Mad Pain and Martian Pain" he argues that the concept of pain is a concept of a state that occupies a certain causal role - a state with certain typical (but not fixed) causes and effects (Lewis, 1983). Lewis describes a Martian who experiences pain but her pain differs greatly from ours in its physical realization. Lewis states that pain is a state which occupies a causal role for a population (Lewis, 1983). Therefore, human pain is the state that occupies the causal role of pain for humans, Martian pain is the state that occupies the causal role of pain for Martians, and perhaps fish pain is the state which occupies the causal role of pain for fish. Based on Lewis' conception of pain we can include the pain that fish may feel (which may not manifest in the same way as ours do). The bottom line is that fish may feel pain but perhaps the pain of fish differs from ours in its physical realization (brain systems and structures at play which result in the *feeling* of pain). Fish differ in many many ways from us so why would may they not differ in how they feel pain?
“Unfortunately, these erroneous inferences are strengthened by beliefs that the motivation to perform certain behaviours is predominantly driven by rewards (either positive or negative).”
ReplyDeleteI am very puzzled by this statement in Key’s article, and this might be due to incompetence in biology, so please correct me if I am wrong. Here, I understand that Key says that it would be wrong to assume that a fish’s motivation to escape an undesirable situation should not be mistaken for the fish feeling any negative reward from this situation. But he also claims that “fish clearly respond to noxious stimuli, and […] these behaviours are subject to operant leaning using noxious stimuli”. To me, operant learning is all about (consciously) perceiving positive and negative feedback from your action. I do not understand how a fish could learn from noxious stimuli without feeling which type of reward (positive or negative) entails from the contact. Something must exist, if not pain, that influences (that’s interpreted by) the fish to avoid certain situations, and I do not see anything like that happening if the fish doesn’t feel at all.
"Accepting at “face value” that fish feel pain may seem like a harmless alternative, but it has led to inappropriate approaches to fish welfare. It could also lead to legislative restrictions on fish-related activities with potentially serious negative implications for native subsistence fishing, human nutrition and food supply, and economic development. "
ReplyDeleteAs stated in the previous skywritings Key’s argument relies on the assumption that pain must be generated by the same structures and pathways in fish that pain arises from in humans. Even if fish don’t feel the same exact pain that we feel, that does not mean it does not exist. And, as noted in Harnad’s piece on animal sentience, there are many other sensations outside of physical pain that contribute to animal suffering (i.e. isolation, restriction of freedom, depletion of normal amount of resources) that should be taken into consideration regarding legislative restrictions.
However, Key refutes the idea of an alternative “fish pain”, stating:
“If “fish pain” existed, it would need to be consciously processed using at least the minimal neural architecture described above. Thus, in the absence of this architecture, fish cannot feel any sort of pain.”
He centers this rebuttal on the lack of conscious processing in fish. It is interesting how this is used to justify harm or lack of protective legislation. I thought of people with Congenital Insensitivity to Pain (CIP), where one cannot consciously feel pain. No one would use this lack of normal pain processing to justify inflicting harm to these individuals because they would still be experiencing physical harm and damage. Clearly this consideration falters as we move further from our own species to biologically dissimilar species.
Re: Emily Bouton November 17, 2019 at 5:20 PM
DeleteThis specific part left me doubtful and honestly quite unconvinced too, but for slightly different reasons. His argument that we have not cared for fish adequately because we consider that they feel pain sounds ridiculous to me. If anything, we have behaved as if fish did not feel any form of pain.
If you take the example of salmon farms, where fish are locked in small cages where they barely have any room to swim in order to meet demand, it seems quite obvious that we have not taken fish pain into account. This example also refutes Key’s argument that accepting the existence of “fish pain” can have “potentially serious negative implications for native subsistence fishing, human nutrition and food supply, and economic development.” We have continued to develop more economically-efficient ways of fishing, and we have take in to a scale that is so immensely superior to native subsistence fishing that the comparison is absurd.
So, to come back to your original comment, not only would we not harm humans that can’t feel pain, but we have continued to harm fish despite recognizing that they do feel pain. Key’s arguments do not hold up.
Key takes a functionalist approach to the question of whether or not fish feel pain. "Pain" is a conscious feeling that happens in the brain, and is a result of the organism experiencing some noxious stimuli. He argues that only organisms with the correct neuroanatomical hardware are capable of feeling pain. Organisms lacking the hardware may react to noxious stimuli (nocifensive behaviours), but they do this unconsciously as that is an unconscious reaction, Key states "(nocifensive behavior) has evolved over hundreds of millions of years to facilitate survival of the organism, and not because it has feelings or possesses any inner motivation to act".
ReplyDeleteRegarding the avoidance of noxious stimuli in organisms that lack the "right" hardware to feel pain (in this example, a rat with spinal cord lesions avoiding a shock by moving their lower limbs which now lack sensation as result of the lesion), Key argues "this behaviour is executed clearly without feeling, and the shock is merely a noxious stimulus that activates neural circuitry that is both beneficial for survival and indicate of the capacity to learn."
I'm personally at odds with this stance. Is what Key defines as pain not also a result of evolution, beneficial for survival? An animal obviously cannot report what it is feeling because it does not have verbal capacity like humans, in that way we can't fully know if it does feel pain I suppose. If an animal is reacting in a nociodefensive way to something we inflict on it, why would the "benefit of the doubt" approach be harmful? In my opinion, the overall sentiment of this argument seems to be a case for us humans to rationalize being apathetic to hurting other animals, which I would say is undoubtedly more "harmful" (for us, other living things, the planet) than giving fish the benefit of the doubt.
“I'm personally at odds with this stance. Is what Key defines as pain not also a result of evolution, beneficial for survival?”
DeleteThis made me think a lot about the evolutionary advantage to feeling pain, and why giving a fish the benefit of the doubt is not only the right thing from an empathetic point of view, but from an evolutionary point of view as well.
The fact that pain is painful (obviously) makes it so that an organism will avoid whatever it is that caused that pain. In the wild, this may be a predator’s bite, a poisonous berry, a fall from a cliff, etc. Avoiding these things increases your chances for survival, and if you follow the obvious assumption that we are wired to want to survive then it’s clear why organisms would feel pain.
I feel like this is all incredibly intuitive and basic, but assuming that a fish feels pain because it would make the most evolutionary sense for it to do so solidifies my stance that fish most likely do feel pain.
Even though evolution is lazy, it still has allowed for organisms to flourish in their environments and continue their bloodlines for thousands of years. Therefore, we can conclude that an organism *most likely* has basic nociception “hardware”, in the same way that we can conclude that other humans *most likely* also feel. This kind of reasoning doesn’t bypass the OMP, but it allows a just-so approximation that goes beyond just empathy and “benefit of the doubt”.
(1) The comparison between not wanting to cause fish pain and not getting vaccinated is outrageous. The outcome of believing fish may feel pain is to simply not eat fish, and vegan diets are not equivalent to the threat of epidemics which could result from people not getting vaccines. The validity of adopting precautious measures depends in part on the possible consequences, of which not eating fish is not at all comparable to not getting vaccinated. Not eating fish does not put anyone in harm's way - it does just the opposite. It was hard to not read the rest of the article following this ostentatious remark. The references to tormenting animals in experiments did not make it appealing either.
ReplyDeleteI also love how the listed negative outcomes of not wanting to hurt fish are in fact non-issues. Legislation to control fishing activities could in fact be highly beneficial - help prevent overfishing which threatens marine ecosystems, or better regulation of fish farms which pose great risks for the spread of diseases and contribute to the climate crisis. Human nutrition and food supply can only benefit from less animal consumption and seeking better alternatives to the polluting and cruel practices of animal agriculture could greatly benefit the economy. I realize I'm preaching to the choir... But good grief. The arguments being made are short-sighted.
(2) What Key appears to be doing is pointing to specific brain areas and activation patterns correlated with pain perception and demonstrating their causal role by showing the difference it makes in animals (including people) who have damage to those areas. He’s trying to provide a causal explanation by showing the role of certain mechanisms in producing pain perception. But there is no explanation of HOW these mechanisms cause pain. There are indications of what is going on, but not how these functions operate... Or is there? I think it is hard reading through essays like this because we’ve been taught in school that, although correlation does not equal causation, the goals of much research is to localize brain areas/activity associated with various outcomes. So, it has seemingly been normalized to settle for this kind of evidence as being valuable conclusive data for explaining mental phenomena. It is admittedly hard for me to feel confident in pointing to how and why exactly these arguments are wrong and what they are missing.
DeleteOf course, this essay doe does not address or resolve the Other Minds problem. This problem does challenge the conclusions being drawn from pain research when the researchers are not feeling the pain themselves and thus cannot be certain of their reports... and yet I would argue, for the same reason that I believe fish and all animals do feel pain, that we CAN know that they experience pain... which again puts me in a difficult position. I want to use the Other Minds problem to poke holes in this logic that I disagree with, and yet I want to dismiss the Other Minds problem to argue for empathizing with animals.
(3) The arguments based on supposedly conscious versus non-conscious visual perception is also questionable. By what measures do we consider something to be conscious? If we respond to something, is it not conscious? Would some forms of perception being vegetative (i.e. not explicitly intentional, ex: not actively thinking about having to breathe and yet breathing automatically) somehow be less noteworthy? What does it mean for patients with insular cortex lesions to have been able to “recognize painful stimulus as painful” and yet “not experience the unpleasant sensations associated with it” (p.10)? How could we detect something without feeling? How can unpleasantness be detached from pain perception?
DeleteI don’t understand any argument for why nociceptive defensive behaviours would NOT be sufficiently indicative of pain. I think there is ample evidence suggestive of animals being able to report on their feelings. Expressions need not be linguistic for them to be significant. Shedding tears, cowering, fleeing, yelping - there is plenty of evidence in animal responses that serve as evidence of their feelings. People have to be wilfully ignorant of obvious cues to believe that only human animals have complex feeling states.
Besides, fish going about their regular behaviours following craniotomies does not mean that they don’t feel pain. A simple comparison would be that it is common for menstruation to be quite painful and yet half of the population does not cease to function while on their periods. As much as possible, we go about our regular behaviours because we do not have a choice, and neither do fish.
(4) I understand that we do not wish to over-extend our anthropomorphic assumptions onto our interpretations of other animals if we are to recognize and understand their unique circumstances and behaviours. There are many misunderstandings and mistakes based on such assumptions. Take for instance the difficulty of knowing whether or not your pet cat or dog is in pain. This is best understood when taking into account that it is typically adaptive for prey to hide any disabilities or wounds so as not to become targets of predators. With this in mind, it makes perfect sense that your cat doesn’t come right up to you and show you that they’ve stepped on a nail and need help. This is not evidence that they don’t feel pain the same way that children do. Their responses are simply different due to circumstances.
DeleteNevertheless, given how practical and reliable our mind-reading abilities are when extended to other humans, it would be foolish to not utilize these empathic abilities to try to understand other species. It hurts me to hear a dog crying for its owner while they’re tied to a fence post outside of a grocery store. It hurts me because I know that those sounds are expressing sadness and fear. I don’t need to see inside of that dog’s head to know that. I know it because I feel it. Key should understand this given his reference the power of humans to intuitively understand the feelings of our friends and family members (p.2). There is no good reason not to extend that empathy further.
Finally, it is laughable that in the conclusion he so quickly jumps from dismissing anthropomorphism to then suggesting alternative measures for pain-related restrictive behaviours in animals EXPLICITLY BASED ON ASSUMPTIONS THAT ANIMAL RESPONSES SHOULD BE SIMILAR TO HUMAN RESPONSES. How on earth is it reasonable to base analyses of pain behaviours in animals on the observations of humans and yet not extend the same logic more generally?
Rachel, bravo for your incisive critiques of Key -- all merited by him. Some of your points were not even made by the commentators in the journal.
DeleteRachel, the example you gave in part 3 really helped me understand more clearly my thoughts about the paper. I think it was a really excellent example that someone such as Brian Key would not have come up with.
DeleteI have to say I agree with everything you said, and I would like to add to the point that you make that "Key should understand this given his reference the power of humans to intuitively understand the feelings of our friends and family members (p.2). There is no good reason not to extend that empathy further." I think that if the word empathy is replaced with feeling, then we're simply talking about the hard problem and the OMP. I think your points serve to indirectly argue against Dennett, Quine, Rorty, Hofstadter, the Churchlands, Andy Clark, Lycan, Rosenthal, Harman, etc. about why the "zombic hunch", and the hard problem is necessary in any discussion in cognition science.
I want to make it clear that I agree with you, but I also found your argument kind of circular in a way. You say that we should extend our empathy and use it to interpret animal's feelings, but you also say that it's absurd to explicitly base assumptions on the idea that animal responses should be similar to human responses. I agree with you, so I can't really answer my own question, but I can't help but notice the circularity here that it's okay to relate animals to ourselves in one way but not in another. (I get why, I just think your argument could be reworded)
As other comments have pointed out, Key’s strategy to explain that fish do not feel pain is flawed on many fronts.
ReplyDelete“(1) identify which brain region is directly responsible for the feeling of pain in humans.”
The only question I have for this is why? Before reading on, I thought that it shouldn’t matter which region processes pain in the human brain, since knowing the region won’t tell us why and how we feel pain (or feel at all). However, Key explains why he wants to identify these brain regions in his third point.
“(3) assess whether fish have similar structural features that can then be used to make inferences about whether fish or cannot feel pain.”
I already see many problems with this. Let’s say we can find the brain region responsible for feeling pain; that doesn’t tell us anything beyond a correlation between structure and function. But even with the knowledge of this correlation, we still will not know how or why we feel pain. And since this is correlational, we cannot gather a causal explanation from it.
Also, whether or not humans and fish share similar structural features related to pain is meaningless, and just another correlation at most. How does knowing the structures responsible for pain in humans help infer anything about whether or not fish feel pain? Additionally, so what if fish lack the structures that are related to feeling pain in humans? That doesn’t mean that they don’t feel pain or that they don’t feel at all.
Trying to infer whether other species can feel or have the same feeling as human sounds really ungrounded to me. First of all, all species have different anatomy and structures, be it for the easy problem -- functioning, or the hard problem -- feeling. It can in no way be justified to deny the existence of fish's feeling (of pain) simply based on their lack of the same neural structures as humans, especially when the two species are in fact so distinct from each other in almost every way they live and behave -- they can hardly be analogous in most scenarios.
ReplyDeleteIndeed, Key has already thought of and even responded to this kind of objection in the paper: "A rebuttal to the argument that fish lack the necessary neuroanatomy for feeling pain is that fish have evolved other brain structures to perform pain-related neural processing. ... However, as noted earlier, most of the fish nervous system circuitry is constructed using feedforward pathways that allow rapid execution of unfelt motor programs subserving survival behaviours."
But his rebuttal does not sound convincing either. "No evidence for the emergence of novel pain regions in the fish nervous system" does not necessarily lead to the conclusion that there must be "a major restructure of existing neural regions in fish to accommodate their additional roles in pain". Chances are that there's something left for us to discover. Likewise, "most of the fish nervous system circuitry is constructed using feedforward pathways that allow rapid execution of unfelt motor programs subserving survival behaviours" cannot warrants the conclusion that fish cannot feel pain. He cannot not prove that this unfelt-motor-behavior-supporting nervous system simply cannot allow the transmission of felt states at the same time.
As a commentary for Key's paper, Balcombe's "Cognitive evidence of fish sentience" is also interesting. He attempted to refute Key's point by a special behavior that Tigerfish demonstrates -- the bird-hunting behavior, which, as he described, is "an ambush planned" by this "intelligent, aware, flexible animal". Interesting as it may sound, I have a little doubt that a survival-necessary hunting behavior can be used as an evidence of the delicate feeling/intelligence. Although I totally agree with him on the basis that "lack of a cortex is flimsy grounds for denying pain to fish, for on that criterion we must also then deny it to all non-mammals, including birds, which goes against scientific consensus."
ReplyDeleteI find it highly doubtful that there could not be another way to get at pain than merely how humans do. Humans pain signature comes from the brain, but there is not a barrier for evolution to make pain evolve differently in other species. It makes sense that fish evolved in a much different path due the drastically different environment they have to grow up in. In addition, we see this across species with other traits such as vision, sound, and smell. Furthermore, we can have different pain signatures within species as well. That is why pain research on humans today, focuses on looking at individuals pain signatures and finding a solution to alleviate that pain in a more customized way. So knowing that evolution is lazy, and that at the very least, it would need to create a system to detect pain and use it as a means to learn and detect harm for even fish. With that said, I wonder if there is a MINSET type thing for the hard problem. Is there a hierarchy in terms of feeling and are there feelings more crucial than others. For example, would pain detecting and feeling pain more important than the feeling you get when you see the color red. Therefore, because feeling pain is more important, it is seen in more species.
ReplyDeleteAlthough we can not say for certain whether fish feel pain (the other minds problem), I see more benefit than harm in doing so. To say that fish feel no pain, puts you at risk of potentially harming fish that can in fact feel pain. To say that a fish feels pain, when they don’t actually feel pain puts you at place where you just provide a fish with a pain free life.
"In conclusion, fish lack the strong reciprocal and networked circuitry required for conscious neural processing" (14)
DeleteKey makes a strong claim in the above sentence. This paper is in reference to feeling pain, however my stating that fish lack the neural circuits for conscious neural circuitry spans all feelings. These include the feelings of hunger, sight, temperature, and of course pain. Although as you said Amy, there is no way for us to know 100% that fish feel pain, the other minds problem, it seems like they would have to for evolutionary reasons. Pain and all of the other feelings listed above are vital for survival. For example, those with Congenital Insensitivity to Pain, humans who lack pain awareness (ie. cann't feel pain) often have to deal with severe consequences such as wounds, broken bones, burns etc. because they don't have the innate "supervised learning mechanism" of pain. From this standpoint, it doesn't seem like fish would be extinct if they did not feel pain.
Additionally, I disagree with Key's point about those with spinal cord injuries not having any sensation below the lesion because there have been many reports on phantom limb experiences. These phantom experiences vary drastically from patient to patient, and can encompass tingling sensations as well as extreme pain. I wonder how Key would address this. Would he justify it by saying that the people with these sensations still have cortices and therefore still feel pain even in regions that aren't connected anymore?
Key argues that because fish do not possess the same cortical architecture that is responsible for pain in humans, they do not feel pain. A central part of his thesis is based on the principle that structure determines function. That may be true for hearts and bladders, but I doubt that this line of argumentation can be applied to the brain. As we have already seen in this course, we cannot infer how and why from where and what. Key only identifies cortical regions whose activity is correlated with pain, which is not a causal explanation for pain. Therefore, his claim that fish do not feel pain because they lack such brain regions seems baseless to me.
ReplyDelete“Interestingly, after such a stab wound injury, fish continue to exhibit qualitatively normal behaviours; there is no observable suppression of their feeding, swimming and mating behaviours (Schmidt et al., 2014), suggesting that they are not experiencing post-craniotomy pain.”
Key makes a big jump in concluding that because fish do not display changes in behaviors that we would typically expect to see in humans (or other mammals) in response to painful stimuli, they do not experience pain. I think Key is falling prey to an anthropomorphic perspective here (despite decrying it so vehemently in other sections of his article!). I suppose this can be extended to Key’s overall thesis, which treats the human brain as the gold standard against which very different brain systems are compared.
In relation to your comment above (which I wholeheartedly believe) I think there is yet one more aspect that needs to be discussed. While Key does drawl several comparisons against the human brain explaining that the regions responsible for pain in humans are not present in fish, I see nothing to stop the fish's brain from having developed an alternate function for a region that is present.
DeleteAdditionally, I find your comment on the behaviors to be spot on. I would also hypothesize that even though they are feeling pain, perhaps via evolutionary trial and errors it was determined that the most beneficial behavior after pain is the standard behavior of feeding, swimming, reproducing, etc.... I'm not a fish anatomist, but I can't imagine they have much of a means of accelerating natural healing or caring for a wound so the behavior of eating and living as if the fish were in a normal state seems very logical to me.
I feel that it might be redundant to discuss the other minds problem (OMP), because I think most of us were probably thinking of this response. Key’s wrong, etc.
ReplyDeleteI do want to comment on some of the inadvertent consequences of his remarks, however, and what it means for humanity.
I think Key’s entire line of thinking (which, again, lacks an understanding of OMP) feeds into a very hierarchy-driven view of nature. He quite literally compares rodents to fish in their brain structure, ability to learn operantly, etc. His (faulty) conclusion is that fish don’t feel pain while rodents do, and the wording of it all is very suggestive that we should have more compassion for rats than fish.
The question that stayed in my mind as I finished this reading was just: “Who cares?”. Key spends 17 pages desperately trying to prove that fish don’t feel pain (who knows, maybe he really really likes sushi), but fails to impact ANYTHING. Nothing he says matters, if we’re relying on his argumentation. Notice how he doesn’t seem bothered by rodent research designs, regardless of this hierarchy he’s built for himself, regardless of the fact he’s said himself they feel pain. He continues to cite that research without hesitation. So, the value of this hierarchy (and subsequently this entire article) is quite null, and is either extremely hypocritical, or just representative of his lack of comprehension on the topic.
And even then, why does it even matter if fish DO feel pain? We’ll never know (again, OMP), and it still doesn’t really make sense to needlessly sacrifice a creature, regardless of whether they feel at all, or whether they feel the same way you feel. Tofu sushi is delicious…
“The possibility that fish lack the necessary neural “hardware” to feel pain is typically overlooked because of anthropomorphic tendencies that bias interpretations of behavioural observations (Horowitz and Bekoff, 2007). Given this controversy, an alternate view is emerging. The proposition is that it is impossible to ever know what a fish feels, and as a consequence fish should be given the “benefit of the doubt” and unconditionally bestowed with the ability to feel pain.”
ReplyDeleteI feel as though it is perfectly reasonable for us to give fish and other organisms “the benefit of the doubt”. Despite Key’s claims about fish not being able to actually feel any pain at all, we should not treat fish as lesser organisms because we will never be able to confirm his claims! This is the other mind’s problem! We will never be able to know with certainty whether another human or another organism is feeling! If anything, humans should be the ones giving all other organisms the benefit of the doubt because we know what it is like to feel pain! We shouldn’t deem it acceptable to hurt other organisms because there is a chance that they don’t actually feel anything at all.
Key argues that the other minds problem is not a valid one to defend our instinct to give fish “the benefit of the doubt”. I understand Key’s point about the other mind’s problem being unsolvable, but I do not agree that it should be dismissed overall because it is impossible to become a fish and experience what it experiences.
“Accepting at “face value” that fish feel pain may seem like a harmless alternative, but it has led to inappropriate approaches to fish welfare (Arlinghaus et al., 2007; Diggles et al., 2011). It could also lead to legislative restrictions on fish-related activities with potentially serious negative implications for native subsistence fishing (Capistrano and Charles, 2012), human nutrition and food supply (Tacon and Metian, 2013), and economic development (Tveteras et al., 2012).”
Here it seems that Key is arguing against humans believing the possibility that fish can feel because it would be disadvantageous for humans to do so. This statement made me feel uneasy. Key argues that we should keep viewing other organisms that aren’t humans, like fish, as non-feeling because it benefits us a great deal. In my opinion, arguing that viewing fish as non-feeling helps us maintain aspects of our economy, culture and living habits is a poor and shallow argument. I don’t believe introducing this argument is a way to scientifically approach or address the other minds problem. In fact, I was quite shocked with his decision to present these economic and cultural issues so early in the paper.
"Lead to unsupported anthropomorphic conclusions - believing fish either feel happy being in company of other fish or they feel disappointed when they fail to capture prey"
ReplyDeleteAs many of my peers have already pointed out, Keys fails to acknowledge the issue of the other minds problem when defending his assertion that fish do not feel pain. The above quote made clear to me how short-sighted he was in this respect. So what if we attribute the behavior of a single or school of fish with human emotions? Just like we cannot be sure that there is more than one sentient being on the world (apart from oneself), the outcome of our intuitive mind-reading abilities would just make us feel empathetic toward them. Even if my feelings aren't valid (such as with the toothache example), I cannot deny that I do feel something. Worse case (as Keys suggests), the widespread adoption of these anthropomorphic conclusions lead to legislative restrictions and the need to restructure our global food industries and consumption. Nonetheless, this change would not mark the be all end all of human extinction. We can never prove that a fish can or cannot feel pain, why then assume that they don't just because it fits within the anthropocentric framework at which our socioeconomic and political systems currently function? This reality does not sit well from a moral standpoint.
ReplyDeleteIt is telling that Keys tends to contradicts himself throughout this paper. While he argues that fish could not possibly feel pain because they lack the human-like neural structures to do so, he ends the paper with an acknowledgement that "we do not have the tools at present to definitively "prove" with a single experimental approach that fish do not feel pain". While this may be mean that it is necessary to converge a multitude of data from diverse methodologies to answer whether a fish can feel pain, it is ironic nonetheless that he himself is applying a single approach (structure to function) as the basis of his argument.
Additionally I find it confusing that Keys asserts that "it may be appropriate to retain the nomenclature of positive and negative rewards for fish". However unlike with humans (and other species with more similar neural architecture; e.g., rats), it should be judged as purely instinctual and for survival rather than for pleasure or to escape painful feelings. I do not understand how our human circuitry also did not evolve to respond to positive and negative stimuli for the purpose to ensure survival? Just because we know (via self-awareness and assume with our mind-reading abilities) that humans have these 'additional' feelings that we experience in response to how we navigate situations in our environment, there is no way via the other minds problem that we could ever judge the same of other species. Keys's take on fish sentience is short-sighted. As he contradicts himself again, "absence of evidence for pain does not mean evidence of absence of pain".
Thoughts on the overall paper: The article parses the various brain regions involved in processing pain in humans. It addressed the impact of a genetic/disease induced impairment/lesion in the brain on processing pain and reacting to the pain. It highlighted the importance of the cortex, in particular the insula, and global integration, in processing and feeling pain in humans. In the paper, Key is answering the questions “which region in the nervous system analyses pain, the anatomy of this region, and do fish have similar brain regions for pain as humans”. However, in doing so Key makes many assumptions.
ReplyDeleteFirstly, as mentioned in some comments above, the basic assumption of the paper is anthropomorphic, and based on the idea that only similar anatomy can lead to similar feelings. The authors argue that reflexive behaviours are generated by the brainstem and felt emotions by the cortex, in humans. Then they assume that fish must display the same correlation between structure and feeling. These assumptions are not scientific.
Secondly, the author assumes that signal amplification and integration (across brain regions) is required for felt pain, solely by arguing with the example of how humans process painful stimuli. The pain processing of no other sentient being in considered in the analysis.
Thirdly, Key claims that “for pain to have any survival value, it must be mapped to specific body locations”. This again is argued by using only the example of humans. In the case of smaller animals is the specification of the type of pain and body part important for survival? Or is it sufficient to be able to reflexively move away from the painful stimuli? This is not addressed by the author.
Fourth, Key argues that fish only respond to noxious stimuli through operant learning, and do not feel pain. They do not offer an explanation on how fish learns this association with a noxious stimulus in the absence of any feeling. Even if “fish have evolved other brain regions to execute the function of pain”, how do they recognize that a particular stimulus could be pain-inducing? There needs to be some sensation associated with the stimulus.
The authors assumptions also point us to a possible limitation of human understanding. Even if we assume that fish feel pain if they behave in a manner similar to us (or me, which is all I can be certain of) based on their reaction to certain stimuli, we are still very much centered in a human-outlook of what pain is. In essence, we might only be able to say “we feel that fish feel pain”? And not that we know that fish feel pain. (Similar sentiment has been expressed in other comments).
So, at best the article offers a review of brain regions involved in pain processing humans, and their functioning in healthy and diseased individuals. Then it goes on to argue that to feel pain all vertebrates must have the same features in their nervous system as humans (a non-scientific approach). It does seem to address at all the question of whether fish feel pain.
The points mentioned above, and in other comments align with those of the authors of “Why human pain can’t tell us whether fish feel pain” and “Fish brains and behaviour indicate capacity for feeling pain”.
------------------------------------------------
Specific thought: On page 10, while discussing patients with cortical lesions the author says “These patients could all recognize a painful stimulus as painful but did not experience the unpleasant sensations associated with it”. How do patients feel/know something is painful without feeling it? There seems to be an inherent contradiction in this argument.
“Fish may lack the neural machinery or architecture to consciously experience noxious stimuli as painful”
ReplyDeleteFirstly, the paper is riddled with weasel words, which is highly confusing for the reader. In the above quote, “to consciously experience noxious stimuli as painful” can be reduced “to feel”. Key’s claim rests on the assumption that the ‘where and when’ pain (to feel) occurs in humans is the ‘where and when’ pain (to feel) occurs in fish. And we are still completely in the dark about how pain occurs, yet Key is quick to back his claim with mere correlative fMRI evidence (which shows us NOTHING causally).
“Accepting at ‘face value’ that fish feel pain may seem like a harmless alternative, but it has led to inappropriate approaches to fish welfare…. Restrictions on fish-related activities... human nutrition and food supply… economic development”
I am totally enraged by this comment. This false positive (we assume fish feel pain when they don’t) seems like a way safer route than the false negative (we assume fish don’t feel pain when they do) provided by Key. He forgets (whoops) to mention what would happen (is happening) if we accept that fish do NOT experience pain. Upwards of 120 billion fish are killed on commercial farms alone annually.
This paper reminds me of our class discussion on the T4 passing robot, Gabe. I have not seen Gabe display any sign of pain in one of our classes this semester. If we were to cut Gabe open, we would see that he does not look like us, humans, internally. But we aren’t going to cut him open, unless we are saving this for our last class? He just has the same inputs and the same outputs; he looks and acts like anyone of us. Now, would you feel comfortable kicking Gabe? If you can’t tell the difference between Gabe and another bystander, don’t make further conclusions. Don’t kick him. Even if Gabe isn’t made up of cells, tissue, organs, etc., are we to assume that he doesn’t feel pain? The feeling of pain seems to be underdetermined in both Gabe and in fish. While fish may not have the same ‘neural machinery or architecture’ as us humans, who are we to say they don’t feel (pain) – especially when we rely on purely correlative evidence.
As the comments above, I disagree with Key’s method and his argument in that how he derives the conclusion of “Fish does not feel pain” from a very anthropocentric view of how the human mechanism of pain work. I also agree that the OMP is something important that Key should have mentioned when considering this question. However, if I am allowed to use my intuition, I do feel that there is this inner conflict inside of me. While I disagree with the rationale Key uses, I find myself more certain that cats and dogs feel pain than birds, and then than insects and fish. My intuition is that I feel more sympathetic when I say dogs suffering than I see a fish being killed. But why? If the feeling of these species all involves the OMP, and if how the neural structural resemble that of human should not be a valid point to judge whether a species has pain. Is it just my intuition fooling me? Does this intuition arise just because we are so prone and used to think in an anthropocentric manner? Nevertheless, I think I am still ready to say, although with reluctance, that, rationally speaking, I have more certainty to say that dogs have pain than I can say about the fish. I wonder why I still feel like this way after separating my anthropocentric intuition here. Why would there be such a conflict? Does that suggest that I should not trust this belief at all, or that there is still something that should be given some weight about the methodology from the point of view of how organism structure is similar to ours?
ReplyDeleteWhile I respect the underlying research of the thesis presented by this paper, the conclusions drawn should be questioned I believe. While fish may not seem to possess the same biological/physiological characterteristics that humans seem to possess specifically fot the experience of pain, that does not in turn mean that they are uncapable of pain or an equivalent unpleasant experience.
ReplyDeleteThere are also numerous dangers for conclusions of this type for the well-being of animals. If conclusions are drawn that fish (in this case) cannot experience pain, does this justify cruel treatment or life in captivity being raised for meat? Some would argue yes. I’d like to argue that even if this report truly meant that fish do not feel any pain, this should not justify bad treatment of animals (which we incidentally are too). I think the comparison of assuming that animals feel pain at “face-value” to the false report of vaccines causing autism is completely misplaced and unnecessary. The two are not comparable.
ReplyDeleteI had written to professor Harnad about this week's content and thought I should post it here as well.
ReplyDeleteWhen reading "Why fish do not feel pain" by Key, I was repulsed at the thought of us humans deciding whether another living being has the capacity to feel pain or not. Not even two pages into the paper and I found myself really bothered and disturbed by the fact that Key brought up some of the economic and cultural disadvantages of granting fish the right to feel pain. I could not believe that he would bring these points, that I do not accept, as important reasons not to accept the "face value" that fish feel pain. I was also quite surprised that he would mention these arguments before trying to make any scientific arguments as to why he believes so. It just shows us how entitled human beings can be and often are... After completing the reading, I thought about the way humans view and treat other animals for the rest of the week. The things we do to other living, breathing and feeling beings is so unfair and unacceptable. I have a dog, which I love as much as any family member, and I thought about his difficult past. I adopted my dog when he was 4 years old from a shelter. He was abused and abandoned by his previous owners. My dog never deserved to be treated so badly. Upon thinking about my dog, I thought about professor Harnad's laptop sticker comparing dogs and pigs... and it is true... why would we treat these two beautiful animals so drastically differently?
I found the questions on pain and animal sentience really interesting and it gave me a new perspective on why we do certain things. Like Harnard said, whether or not they feel the same pain as us doesn't matter, what matters is that they feel pain at all and that pain hurts. This I agree with, and the next question will then be, so why do you want to inflict unnecessary pain to other beings? I don't think that any human, at the core of their being, wants to inflict pain or suffering to others, since it doesn't want to suffer himself. But a lot of conditions, environment, conditioning, push humans to do certain non-ethical things with other humans, the planet, and other beings. Other than eating animals for survival, for people like us who have the choice, I think it comes down to satisfying our desires, and pleasures. Why would I stop fish or meat if I like it? A lot of people just don’t care whether animals are suffering or not, like a lot of people don’t care if they hurt someone else , even if with other human beings , even with the OMP, we can be sure that they feel like we do since they have the same biology as us. These questions are ethical and moral ones and the way we treat others says a lot about the dominance of the ego in our lives.
ReplyDelete“The possibility that fish lack the necessary neural “hardware” to feel pain is typically overlooked because of anthropomorphic tendencies that bias interpretations of behavioural observations” Given this controversy, an alternate view is emerging. The proposition is that it is impossible to ever know what a fish feels, and as a consequence fish should be given the “benefit of the doubt” and unconditionally bestowed with the ability to feel pain.” In his extensive paper, Key (2016) asks whether fish can feel pain and in providing his answer, unwittingly treads on the issue of the other minds problem – that we as humans cannot know what it is to feel (if there is feeling) for a fish. In so doing, Key reveals an anthropomorphic tendency, as he highlights that the lack of structural homology between the neural circuitry of a fish vis a vis a human reveals a lack of capacity to feel pain. The fact that his focus lies purely with the organic means of processing external damaging stimuli, but not the felt state of the fish is quite problematic in my view.
ReplyDeleteKey seems to believe there there is one neuroanatomical presentation that feels pain: ours. He believes that since fish do not possess human hardware (specifically, since they do not possess human “neural machinery”), fish are unable “to consciously experience noxious stimuli as painful” (Key, 2016, pg. 2).
ReplyDeleteWhen Key says “fish do not feel pain”, what he really means is “the brain regions that are active in humans who are experiencing painful stimuli are not active in fish”. Here, Key is appealing to the belief (dispelled by Fodor) that knowing where or when something happens in the brain tells us how or why it does so.
As many have said already, fish do respond to painful stimuli. Functionally, fish operate as though they experience pain. It simply does not matter what kind of pain they are feeling in the presence of a painful stimulus: fish have nocifensive behaviours.
Hi Paul,
DeleteI was thinking exactly the same thing! From what I understood, when Key talks about whether or not fish feel pain, he is not referencing the feeling involved in the “Hard Problem”. It is more what is the mechanism behind pain. Also, we do not know what if feels like to be a fish and he is way too clumsy in the way he talks about the “benefit of the doubt”.