11b. Harnad, S (2016) Animal sentience: The other-minds problem
Reading: Harnad, S (2016) Animal sentience: The other-minds problem. Animal Sentience 1(1)
Instructions for commenting: Quote the passage on which you are commenting (use italics, indent). Comments can also be on the comments of others. Make sure you first edit your comment in another text processor, because if you do it directly in the blogger window you may lose it and have to write it all over again.
***EVERYBODY PLEASE NOTE: I REDUCED THE MINIMUM NUMBER OF SKYWRITINGS. BUT THE READINGS ARE **ALL** RELEVANT TO AN OVERALL UNDERSTANDING OF THE COURSE. SO, EVEN IF YOU DO NOT DO A SKYWRITING ON ALL OF THEM, AT LEAST FEED EACH READING YOU DO NOT READ TO CHATGPT AND ASK IT FOR A SUMMARY, SO YOU KNOW WHAT THE READING SAID — OTHERWISE YOU WILL NOT HAVE A COMPLETE GRASP OF THE COURSE TO INTEGRATE AND INTERCONNECT FOR THE FINAL EXAM.***
ReplyDelete*Mental states are felt states, and to have a mind means to have the capacity to feel. In a word: sentience.*
ReplyDeleteFeelings, unlike behavior are not observable. We can only be certain of our own feelings/mind, how do we know that other organisms are sentient too? We can only infer the mind by the observable and that is the other-minds problem. However, this is not a problem for our species because we know other members of our species have a mind. We can use language as a vehicle of our thoughts and feelings. We can also observe and even predict how one will behave based on their saying. Our mind-reading skills are good, “mirror neurons” probably play a role when we observe and feel. Nevertheless, the other-minds problem applies to other species because it is not as clear to us what feeling or thinking is for them especially when they are not mammals. Our scepticism of other organisms’ sentience can change how we treat them. If they are in fact sentient, it would have detrimental effects on them like fear and suffering.
I agree with you that our confidence in other human minds comes from the way we can map their behavior and language onto our own felt experience. What your point made me think about is that this same strategy also shows why we should be more open to the possibility of feeling in other species. The very fact that we recognize our own feelings only through the way they appear in behavior means that behavior is our only real tool for anyone else. So if an organism shows patterns that in us come from feeling, that should count as meaningful evidence too. Our uncertainty does not erase their possible sentience, it just means we need to interpret their signals with more care.
DeleteWhen reading the paper, what struck me the most was Descartes' claim that animals don't feel anything. I kept wondering what brought him to making that claim, and the answer is nothing much? Descartes' biggest contradiction is the work he did to get to "the cogito" vs the work he did to claim that animals don't feel. In méditations métaphysiques Descartes goes through an entire experiment about blocking prior knowledge and any subjective experience (any sensorimotor input) to conclude "je pense donc je suis". It is the ONLY thing outside of mathematical proof by pain of contradiction that is certain. And yet, he needed nothing to claim that animals are not sentient and thus we can do whatever we want with them...
ReplyDeleteFollowing up on your point, I believe that his assumption is a product of the era Descartes lived in, during which a majority of people, including philosophers and scientists, believed that a divine being made humans the "special" species on planet earth. In my opinion, although he did go through a scientifically sound process to prove the Cogito, he did not attribute much importance to animals as they were not seen like creatures with a divine soul such as humans. Philosophers and scientists of the time, in my opinion, exhibited condescending behaviours toward animals and their religious mindset often led them to rushed conclusions by relying on deductive reasoning grounded in religious assumptions. To conclude, I would personally attribute Descartes’ views to the zeitgeist and would be inclined to believe that he would think differently if he were alive today.
Delete“With humans we can rely on language and neural similarity; with nonhuman animals we must rely on behavioural and anatomical evidence alone.”
ReplyDeleteThis passage raises an interesting point about human attitudes toward animals. Some individuals, like Descartes and Key, make strong claims about animals based on their perceived power or intellectual authority, asserting what is “true” about animal experience. Others, who recognize their own limitations and approach with empathy—trying to imagine what it is like for a fish or a dog on an operating table—rely on behavioural and anatomical evidence to make careful inferences. This distinction begs the question: are we morally commanded to care for animals and encourage others to do the same, integrating true concern with the ethical precautionary principle? Or are we justified in acting based solely on what we think is best? Humans are limited to our own subjective understanding; apart from that, we are constrained in truly knowing what it is like to be another being. This limitation may also point to a higher moral law, suggesting a Creator or good God who models moral uprightness for us to emulate. Recognizing our limits is crucial, not only for respecting animal sentience, but also for how we relate to other humans in our world.
Rachel, I really appreciate your point and how you discuss recognizing our own limitations. When reading this paper, I thought about how quickly confidence can become moral blindness. In fact, when someone insists that a species cannot feel, they automatically close the door to research and new evidence. This made me realize the importance of approaching animals with the assumption that our knowledge is incomplete (and might be incomplete forever) because we will never have full access to others' minds. Thus, even when in doubt, humans must choose the stance that minimizes harm to animals.
DeleteHopefully this makes sense:
ReplyDeleteIf we 1. know ourselves that individually we can feel and 2. can infer well how others are feeling based off of our own perception of how our feelings manifest into our own behaviors, then what does that mean in terms of both the easy and hard problems in the sense that yes, we don't know WHY we feel, but we are able to infer how others are feeling because we apply what we know about how we feel?
Attempting to answer my own question after yesterday's class: We know directly that we feel, and we use the way our own feelings manifest in behavior (grounding?) as the template for interpreting others’ behavior. This helps with the easy problem of explaining how we predict and understand others’ states. But it doesn’t touch the hard problem: even though these self-based inferences work (obviously to an extent), they give us no explanation for why feeling exists at all. Specifically talking about humans in this case to avoid the OMP
Hi Sannah!
DeleteThis makes sense, and I’d frame it like this:
We only know one case of feeling directly, our own. That’s the Cogito point Dr.Harnad stresses.
From there, we use the way our feelings show up in our behavior as the template for interpreting others’ behavior. That solves the “easy problem” in the social sense: it explains how we successfully predict and understand others’ states through mind-reading.
But none of this gets us any closer to the hard problem. Even perfect mind-reading doesn’t explain why feeling exists at all or what causal role feeling plays beyond the behaviours and neural mechanisms that could, in principle, be mindless. Our self-based inferences help us navigate others, but they don’t explain the existence of experience itself.
This comment has been removed by the author.
ReplyDeleteInstead of overly emphasizing how we can prove that animals have consciousness, this paper points out the very critical question: what if we are wrong? Denying sentience - and being wrong - is a cost that far exceeds the outcome where we grant them moral consideration when they do not need it. Furthermore, the idea of "asymmetric risk” reminded me of Pascal’s Wager;’ we can’t achieve certainty about other minds, much like we cannot be certain about God’s existence. In both cases the rational choice is a precautionary one. In the Wager, it is most rational to believe in God even, because the cost of being wrong about His existence (going to Hell) outweighs that of believing in His existence when He doesn’t (nothing). The safer bet for the other mind's problem is to believe they are sentient – because if they are and we do not treat them as such, we have committed ethical crimes.
ReplyDelete
DeleteElle, your way of thinking about this question of risk made me want to attempt to frame it within Kahneman and Tversky’s Prospect Theory, which is a formalized model of individual irrationality around risk. One way humans behave irrationally is that we are risk-preferring in losses and risk-averse in gains. That is, we would rather take a gamble on a loss than accept a certain loss.
If we think of the possibility of animal sentience (and the associated suffering) as a “loss”—because, as we discussed in class, most humans benefit from animal suffering in ways like eating meat or using animal-tested products—then it follows (according to Prospect Theory) that humans would rather take a gamble on that loss (i.e., a gamble on whether animals are sentient) than accept the loss for certain (e.g., giving up eating animal products).
On the other hand, if you frame this from a positive perspective—which I think Professor Harnad might—he would prefer the certain option: that animals are sentient and therefore we should not harm them in any unnecessary way. This is preferable to taking a gamble on whether they are sentient, which leads to the mixed system we have now, where animals have minimal rights and mass suffering continues.
The “other-minds problem” is really their problem.. the problem of the animals who might feel but can’t tell us. We only know our own feelings directly; everyone else’s we infer from behavior. That works well for humans (language) and somewhat for mammals and birds, but breaks down quickly for fish, reptiles, and invertebrates. Not because they’re mindless, but because we cant really read them. What stood out to me is the point about risk: If we wrongly assume an animal feels, no harm is done. But if we wrongly assume they don’t feel, the cost is entirely on them. That flips the usual mindset; it’s not about proving sentience first, it’s about avoiding catastrophic mistakes. If we can never be certain about another species’ feelings, how do we decide where to draw the line between sentient and non-sentient, and who gets the benefit of the doubt here?
ReplyDeleteHi Shireen, yes I agree that the paper makes an excellent points about the "other minds problem". Assuming an absence of sentience poses a large ethical risk of deploying harm and suffering. Essentially, uncertainty should not paralyze action, but rather motivate more caution, not less. This connects directly back to Descartes’ argument, where he draws a sharp line by claiming that the only thing we can know with certainty is our own mind. While that’s true, his conclusion that animals therefore lack minds entirely is both arbitrary and ethically dangerous. Instead of using uncertainty to justify excluding animals from moral consideration, we should treat uncertainty as a reason to extend the benefit of the doubt.
DeleteThanks for your comment Emily! I agree, especially about uncertainty demanding more caution rather than less. What this made me think about is how asymmetrical the risk really is. The “other-minds problem” isn’t our problem at all .. it’s the animals’ problem, because any mistake we make falls entirely on them, never on us. Descartes used uncertainty to withhold moral consideration, but the reading shows why that’s backwards: if we can never be certain about another species’ feelings, then certainty can’t be the criterion... So maybe the line shouldn’t be drawn at “prove they feel,” but at “is there any reasonable chance they do?”.. becuase once you shift the burden that way, the benefit of the doubt expands rather than contracts.
Delete“It certainly does not follow, however, from the undeniably true fact that we feel, that anything else that we may feel to be true really is true.”
ReplyDeleteI believe this is the most relevant quote from the text. The other-minds problem states that the only feelings we can be certain of are our own, and that we attribute sentience or feelings to species similar to human beings based on our subjective experience. Scholars like Descartes and Key argue that non-human animals, such as fish, are “feelingless machines,” and therefore we may treat them however we see fit.
My question follows from this assumption and is perhaps more philosophical in nature: Why is it that, even if we do not attribute human feelings to animals, we therefore have no responsibility to treat them correctly–or at least ethically? Many things lack feelings and are not sentient—water, air, the shoes on our feet—and yet we care for them because of the roles they play in our lives. We try not to pollute the water we drink or the air we breathe so we can live healthy lives, and we repair our shoes for both functional and aesthetic reasons. We interact with these things every day, and our own quality of life diminishes if they are mistreated.
Given this, how do thinkers who share the views of Key and Descartes reconcile the mistreatment of animals when their well-being so clearly affects our own?
Esosa,
DeleteYour question about why denying animal feeling seems to erase ethical responsibility gets right to the heart of Harnad’s point that the “other-minds problem” is not our problem but the animals’ problem, because any error we make falls entirely on them, not on us. As Harnad notes, even though we cannot be certain that other species feel, we also cannot be certain that they don’t, and the consequences of a false negative, assuming they are feelingless when they are actually sentient, are extreme for the animals we harm in that belief. Thinkers like Descartes and Key try to justify mistreatment by wrongfully treating animals as mindless machines, but Harnad shows that this stance ignores both the epistemic symmetry (we lack proof either way) and the ethical asymmetry (only they bear the cost of our doubt). So even if we cannot attribute “human feelings” to animals, the fact that the risk of being wrong is borne entirely by them is exactly why we do have responsibility to treat them ethically.
Harnad stresses that sentience means feeling, not just behaving, and although we are confident that mammals and birds feel, we become less certain as species become less like us. This uncertainty is dangerous because the cost of being wrong could be big and ethically troubling: if an animal can feel and we mistakenly assume it cannot, we risk inflicting suffering that we cannot detect. Harnad argues that skepticism about animal minds harms animals, but not us humans, and therefore we must give nonhuman species, especially those whose sentience is uncertain, the benefit of the doubt. Harnad argues that the cost of a false negative (assuming an animal doesn’t feel when it does) is far worse than a false positive, and I would agree.
ReplyDeleteWhat stood out to me is how Harnad reframes the other minds problem as their problem, not ours. We only know our own feelings directly, and every other case is an inference based on behavior. For humans this works because we have language and strong mind reading skills. But for nonhuman species, especially ones unlike us, the risk flips. If they feel and we deny it, the entire cost of that mistake falls on them. This makes sentience an ethical question more than a scientific one. When we are unsure, giving animals the benefit of the doubt is the only morally safe choice.
ReplyDeleteI agree with your arguement. Comparing this to Key's article "Why Fish do not Feel", he misses the morality of the benefit of the doubt. He actually does not doubt at all, and uses scientific terms to turn the audience away from it.
DeleteRena, I second that! Harnad (2016) extends the other-minds problem to nonhuman species, challenging Descartes’ theologically biased claim that animals cannot feel. He states that “human beings are the planet’s unrivalled champions in mind-reading one another”, suggesting that we can infer feeling through behaviour. However, the problem becomes more complex when we consider other species who we know (or at least think we know) feel but don’t know exactly what they feel. While we cannot be certain that we—or any other species—can feel, we still cannot doubt that we ourselves feel something. This ultimately leads to a larger ethical complication, highlighting the importance of recognizing that sentience exists, as failing to account for any feeling may result in further suffering or harm.
Delete"So it turns out that the other-minds problem is not our problem: It’s the problem of other species, if they do have minds "
ReplyDeleteWhat I found most convincing in Professor Harnad’s argument is the shift from epistemology to ethics. Even if uncertainty is unavoidable, the distribution of risk is not: any mistake we make falls entirely on the animals, not on us. This makes skepticism ethically irresponsible. I agree with Professor Harnad that our job is not to wait for perfect proof of sentience but to prevent avoidable harm. If scientific uncertainty is guaranteed, isn’t the precautionary principle the only rational baseline?
Hi Randala!
DeleteThat phrase stuck to me as well.
When he says the other-minds problem is not our problem but the problem of other species, he is also pointing to how uncertainty itself advantages feeling beings like us. Because we know with certainty that we feel, we are protected from skepticism in a way other species are not. Our own sentience gives us moral security, while theirs remains exposed to doubt.
This makes uncertainty an asymmetrical resource. It allows humans to set evidentiary standards from a position of safety, while animals pay the cost of falling short. In that sense, skepticism is not neutral but structurally biased toward those whose sentience is unquestioned. If uncertainty is unavoidable, then using it to delay protection is not rational caution but an ethical failure. The precautionary principle is the only position that does not exploit this imbalance.
When it comes to the sentience of other cognitive creatures we have to consider the other minds problem. We can observe the behavior of organisms but not their minds directly. So what their minds are feeling or thinking has to be inferred from their behavior. This gap is the other minds problem. With other humans we are more easily able to infer what others might be thinking and feeling because we have very advanced mind reading abilities from speech, behavior, and brain activity. But with other species, especially those like invertebrates or reptiles that have behavior very different than ours, we can struggle with inffering their mental states based on their behavior. This can cause us to disregard their sentience, leading to their suffering. So ethically speaking we should be giving them the benefit of the doubt and assuming that they are sentient.
ReplyDeleteIn this paper, Harnad says that the other minds problem is actually the other-species’-minds problem: we have accepted that other humans (most likely) do feel, but are still undecided on whether other species do. Animals are unfortunately without language, and thus cannot tell us that they feel, and so we are left with our other two forms of deduction: observing behaviours and mind-reading. This has led us to mostly believe that mammals and birds feel, but it grows harder still when looking at fish and insects. Harnad emphasizes that while it’s not possible at this point to confirm or deny whether invertebrates feel, we have to consider not just what is more probable, but also the consequences of being wrong - mainly for the animals we would judge to be unfeeling.
ReplyDelete11.b. Harnad’s whole point is that the “other-minds problem” stops being a philosophical puzzle and becomes a moral emergency once we talk about animals. He’s right that we can never prove another species feels, but the idea that we should withhold belief until we get certainty seems backwards. If the cost of a false negative is real suffering, then skepticism isn’t neutral; it’s harmful. What’s strong in the paper is how he flips the classic Cartesian doubt onto its head: the problem isn’t that we don’t know, it’s that they can’t tell us. Still, the precautionary argument, while compelling, leans heavily on human empathy as a guide, which is shaky across distant species. The piece is persuasive, but it leaves open the trickiest question: how far down the biological tree should this “benefit of the doubt” actually go?
ReplyDeleteThis Harnad paper on animal sentience makes a huge ethical point, emphasizing the other-minds problem and the consequences of error when we assume animals don't feel when they actually do. Prof Harnad argues that if feelings are being felt, they can be hurt, so when faced with uncertainty about sentience (like in fish), skepticism can lead to massive suffering. This perspective really clashes with last class’ paper on fish pain, where Key takes a strictly third-person, neuroanatomical approach, insisting that the feeling of pain is distinct from nociception (non-conscious response to painful stimuli). Key argues that since fish lack the necessary brain architecture, the most reasonable inference is that noxious stimuli “don’t feel like anything to a fish”. Basically, Harnad advocates giving the benefit of the doubt due to the ethical stakes, viewing Key's conclusion as a dangerous form of scepticism. Key, however, demands high-level anatomical proof based on a structure determining function principle before granting sentience.
ReplyDeleteI liked how Harnad reframed the other-minds problem as fundamentally an ethical problem especially after reading the paper about fish. Because we cannot directly feel another species’ pain, all we have are inferences based on behaviour. But Harnad stresses that uncertainty cuts both ways- we can’t be sure they do feel, but we equally can’t be sure they don’t. And since the moral cost of being wrong is so high, we should be on the side of sentience rather than skepticism. Key almost does the opposite- he relies on structural absence (no cortex, therefore no conscious state) to justify withholding the ethical significance. This kind of strict proof requirement is risky, as Harnad says, because withholding belief in sentience has consequences only for the animal. Harnad isn't saying ignore neuroscience but just that ethical reasoning has to account for the cost of our uncertainty, not just the evidence gap.
ReplyDelete“Thanks to our acute mind-reading abilities, we grant the benefit of the doubt — to some degree — to mammals and birds, because they and their young resemble us”
ReplyDeleteI think this point is important because it manifests itself in the world via policy and moral norms in addition to whether or not we privately empathize with other organisms. Animals that resemble humans more tend to be protected by society; for example, the average person would stop their car to avoid running over a squirrel but wouldn’t show as much empathy for a rat. We use the same intuition to decide what animals are acceptable to eat and establish rules on how ‘humane’ their slaughter needs to be. These tendencies are irrational because we know the other minds’ problem (that we cannot know what it feels like to be another organism) applies equally to every species. The principle that we can only infer an organism’s capacity to suffer through behavior and biology does not guarantee certainty. Therefore, our own bias should not grant certain species protection over others.
I keep noticing how inconsistently humans apply the other-minds problem. Technically, we can never be certain other humans feel, yet no one uses that doubt to justify harming people for trivial benefits. But we do use that doubt - often unconsciously - when it comes to animals. This suggests our behavior isn’t philosophical at all; it’s driven by neurobiological bias. Our empathy system is tuned to human faces, voices, and emotions, not to the bodies or expressions of other species. Culture then reinforces those biases by teaching us which animals “count.”
ReplyDeleteIn that sense, we fall into the same structure-function mistake as Key - assuming that only beings who look or act like us can feel. But this says more about our perceptual limits than about animals’ inner lives. The inconsistency makes me wonder whether our treatment of animals reflects moral reasoning or simply the boundaries of human-centered psychology.
That’s an interesting question. I think it comes down to what we discussed in class about our mind-reading abilities and when they are inhibited. Just as Professor Harnad (and multiple comments) have emphasized, the Other Minds problem isn’t as much of a concern for humans thanks to our ability to communicate feelings verbally and our mirror-neurons enabling our capacity to read other people by interpreting their behaviour. When non-human animals start to become increasingly distant from our own species, our mind-reading capabilities fail. Similarly, I can rationalize what it’s like to be a bat, but I will never understand what it feels like to be a bat, as felt states are uncomplemented categories (i.e., we know only what it feels like to experience our own human cognition and nothing beyond that). However, that should be even more reason to proceed with caution, paralleling our “innocent until proven guilty” mindset. Thus, relative to your comparison, I would argue your latter statement.
DeleteWhile we can never be certain of another’s sentience with mathematical proof, we can reach a level of practical certainty that makes skepticism “academic and otiose”. Our ability to "read" the minds of other species particularly mammals, is likely enhanced by a shared evolutionary history and biological mechanisms like mirror neurons, which may generate our "telepathic skills" by allowing us to vicariously experience the sensations of others. However, as you noted, sentience is not inferred from a single clue like learning; rather, it is a conclusion drawn from a confluence of anatomy, physiology, and behavioral capacity. Ultimately, while biological markers like mirror neurons and learning provide the "clues", the robust functional capacity of these animals makes the presence of feeling a scientific and ethical necessity to acknowledge. The real challenge for cross-species mind-reading is no longer proving if they feel, but accurately determining what they feel.
ReplyDeleteIn this article, Harnad shows how the other-minds problem isn't as much of a problem for humans, because of three factors that lead us to strongly believe that other humans have minds and, thus, can feel. First, language is a very important one, as it allows us to tell others how we are feeling, to describe to them what we are feeling. Second, we can also observe someone’s non-verbal behaviour and infer or confirm what they are feeling. Finally, humans have great mind-reading capacities; we can sometimes feel what someone else is feeling: this is empathy. While we can’t really use these factors on non-human species, that does not mean we should conclude they do not feel or do not have minds. Rather, Harnad warns us of the consequences of being wrong, that we risk assuming that animals do not feel when they do, and feel the hurt. The only thing we can be sure of is the Cogito: what we feel as an individual.
ReplyDelete