Overcoming Anthropocentric Bias

" "Life on farms may be worse than death"
by Robert Wiblin

COMMENTS
An important post: thanks Robert. If we were considering the ethics of human slavery, then we'd give little weight to the arguments of ethicists who were slaveowners - not because the slaveowners were necessarily insincere, but because we know how the human capacity for self-serving bias is deep-rooted - and also abundant evidence that humans have only weak introspective insight into our own motivations. So analogously, what about the ethics of nonhuman animal exploitation? Who should decide what's legally and ethically permissible? Can meat-eaters with a vested interest in the continuation of animal exploitation ("But I like the taste!") be trusted to offer a reliable source of unbiased judgement?

Of course, when considering the ethics of human and nonhuman slavery, just citing the potentially ethically catastrophic risk of self-serving bias doesn't disqualify, on its own, appeals either to the logic of the larder or any other potential rationalisation of human exploitation of members of other races or species. Lifelong vegetarians doubtless have biases of their own, e.g. maybe they're unusually prone to signaling their caring disposition to prospective mates or whatever. However, it is (almost) literally incredible to suppose that any ethicist accustomed to a meatless diet would conclude that we have an obligation to turn a vegetation-rich, animal-poor environment into today's gulag of factory farms and slaughterhouses - and an obligation to do so in the notional interests of the currently non-existing creatures who would be brought into being to suffer inside them. For a start, recall how factory-farmed animals are so distressed they have to be physically prevented (via tail-docking, debeaking, castration, etc) from mutilating themselves and each other. Humans have to be extraordinarily traumatized or psychotic to do anything of the kind. In short, if humans think we're behaving ethically, we're simply kidding ourselves.

* * *

Michael, as you know I've long advocated selective breeding of human and nonhuman animals for extreme happiness. This aspect of Carl's proposal strikes me as admirable. But what's lacking is any kind of acknowledgement that sentient beings should not be treated as property. By analogy, if we lived in in a slaveholding society, then a white slave-owner might argue for selectively breeding happier slaves rather than for emancipation. But you would (I trust) argue that the time for emancipation is now. In default of accurate neuroscanning, let us provisionally assume that any captive human or nonhuman animal that has physically to be prevented from mutilating himself or herself is profoundly distressed. If we are classical utilitarians, then it doesn't take a four-sigma level IQ to calculate that the fleeting pleasures of the dinner table do not ethically outweigh the horrors that went into its production. Carl's proposal, if implemented, would mitigate but not eliminate these horrors. No doubt life presents difficult ethical dilemmas that call for fine calculations and fancy maths; shutting down factory-farms isn't one of them.

* * *

Whether value can be naturalised is a distinct question from the nonmoral instrumental ought cited above, i.e. insofar as we want to understand the properties of the natural world, we ought to abandon preferred or privileged reference frames.

* * *

A unified science must ultimately provide an account of the origin of our normative concepts. Insofar as I know anything at all, I know my agony is bad for me. Natural science suggests that no here-and-nows are ontologically special. The agonies of other subjects of experience of comparable sentience elsewhere seem less significant than mine; but this is an epistemological limitation on my part, not some deep truth about the world.

* * *

Srdiamond, forgive me, but you missed my point, namely that the meta-ethical realist and anti-realist alike can use "ought" instrumentally. Meta-ethical antirealists who want to understand the world are not debarred from using normative concepts; it would be hard to do science without them.

A much stronger claim would be that a God-like superintellect with perfect knowledge of all possible first-person perspectives could not behave unethically. Perhaps imagine a generalisation of mirror touch-synaesthesia:
http://www.livescience.com/1628-study-people-literally-feel-pain.html
A convergence hypothesis would run counter to the orthogonality thesis assumed in most discussions of AGI.

* * *

"Sentience"? Just the standard sense of the term found in any dictionary, i.e. the ability to feel, perceive, or be conscious, or to have subjective experiences:
https://en.wikipedia.org/wiki/Sentience
The ganglia of insects may well experience phenomenal pain. But in the absence of a central nervous system, a unitary subject of experience would seem impossible. Thus the head of some species of locust can carry on feeding while the tail segment is being devoured. Antispeciesism is not the doctrine that "all species are equal", rather that it is irrational to discriminate against beings of comparable sentience merely on grounds of species membership.

* * *

There is nothing ad hoc about prioritising unitary subjects of experience because there is no ontological integrity to, say, a dozen ganglia each in its only body segment. Utopian technology may one day deliver the well-being of all ganglia and even individual nerve cells; but this happy day is a long way off.

Sadly, the claim that no species exists whose sentience is comparable in degree to humans is unsupported. Microelectrode studies confirm that the most intense forms of sentience are mediated, not by e.g. the brain structures supporting generative syntax, but by the evolutionarily ancient limbic system mediating our core emotions. What grounds have we for supposing that the larger limbic system of, say, a sperm whale supports less intense consciousness?

Less controversially, a pig is of comparable sentience - and for what it's worth, intelligence - to a two-year-old prelinguistic human toddler. Insofar as prelinguistic toddlers deserve love, care and respect, so do pigs. And to the counterargument that only toddlers have the "potential" to become adult humans, we may recall that toddlers with a progressive disorder (and thereby lacking in cognitive potential) are not thereby reckoned less worthy of love, care and respect. Likewise pigs - on pain of arbitrary anthropocentric bias.

* * *

When one rapidly withdraws one's hand from a hot stove, the hand withdrawal often precedes the pain. This doesn't entail that pain didn't play a causal role in my hand withdrawal: maybe peripheral nerve ganglia experience raw micro-pain. But if so, it's encapsulated. My CNS does not have direct access to peripheral nociceptors. So there is no "ontological integrity" in play here. Likewise with the locust I alluded to above whose head continues feeding while the tail is being devoured by a predator; there is no unitary subject of experience whose interests deserve consideration, just the experiences of its constituent nerve ganglia.

How 80 billion odd interconnected but discrete quasi-classical neurons in the CNS solve the phenomenal binding problem (cf. http://lafollejournee02.com/texts/body_and_health/Neurology/Binding.pdf) and generate bound objects and a fleeting synchronic unitary self is a very deep problem that IMO cuts to the heart of Moravec's paradox, explains why classical digital computers will never be nontrivially conscious, and rules out mind uploading. But this discussion would take a long way from the morally urgent question of factory farming.

Sentience? Well, the ganglia of a roundworm are minimally sentient; a pig or a toddler intensely so. Within the individual, any seasoned user of psychoactives will report interventions that either dim or amplify the intensity of experience. But the ethically relevant point is that the molecular substrates of our most intense experiences are far removed from e.g. the mutations of the FOXP2 gene that helped give rise to the generative syntax that "makes us human". Bentham put it well over 200 years ago: "The question is not, Can they reason? nor, Can they talk? but Can they suffer?"

* * *

"Ontological integrity"? Tim what is the difference between a mere structured aggregate of "mind dust" and a unitary subject of experience? What distinguishes a split-brain patient with a severed corpus callosum - or an invertebrate with multiple ganglia in its different segments - from a brain that supports a fleetingly unitary phenomenal self? Dreamless sleep aside, why doesn't
https://en.wikipedia.org/wiki/Mereological_nihilism
hold for the 80 billion odd seemingly discrete classical neurons of the vertebrate brain?

Intensity of experience? Establishing the neural correlates of consciousness is challenging for an agent like LSD which is undetectable three hours after ingestion. So let's focus on phenomenal pain. Nonsense mutations of the SCN9A gene completely abolish the capacity to experience phenomenal pain. Other alleles are associated with an unusually high or unusually low intensity of pain experience in response to a given noxious stimulus. On Carl's proposal we would presumably want to ensure nonhuman factory-farmed animals have ultra-low pain alleles of SCN9A - if not nonsense mutations. A strong ethical case can be made for preselecting benign low-pain alleles of SCN9A via PGD for future humans:
http://www.pnas.org/content/107/11/5148

How can we appraise the comparative severity of phenomenal pain in sperm whales [and pigs etc] and humans? Although we should be wary of facile "sizism", presumably the possession of a "pain centre" with 100,000 interconnected neurons might, other things being equal, be expected to yield a maximum intensity of pain greater than the agony ceiling of a pain centre of 25,000 structurally and functionally identical interconnected neurons with the same gene-expression profile, etc.

Yes, there are a host of complications here. I was simply leaving the question open - and I very much hope I'm wrong.

* * *

Srdiamond, apologies for any confusion. I was simply raising the possibility that nonhuman animals with larger pain centres than humans feel pain more intensely. I agree that generating pain of high intensity doesn't take a lot of neurons. But other things being equal - and depending on how the brain solves the binding problem - then 10,000 interconnected pain-processing neurons can presumably generate a greater phenomenal intensity of suffering than 1000 interconnected pain-processing neurons. More generally, the greater diversity of phenomena that mature humans can feel distressed or miserable about compared to nonhuman animals seems to be a function of the projections of neurons from the limbic system to the neocortex rather than any greater diversity of neuronal cell types in the limbic system itself.

Ant pain? I certainly favour erring on the side of caution. See e.g.
http://www.utilitarian-essays.com/insect-pain.html
But whereas phasing out suffering in higher vertebrates is feasible with existing technologies, doing the same for insect pain will require the utopian technology of our successors next century(?) and beyond - and a momentous ethical revolution to match.

* * *

srdiamond, I certainly wouldn't dismiss Moore's Open Question argument. Indeed, on the face of it, the argument is decisive. Whatever apparently dreadful or wonderful phenomenon-exists in the natural world, one can still ask if it's (dis)valuable. However, what is not an open question, at least for me, is whether my unbearable distress is disvaluable for me. And I'm arguing that it's only an epistemological limitation on my part, not some deep ontological truth about the world, that leads to any failure in my recognition that ( it is objectively the case that) your unbearable distress is disvaluable too. The badness of your agony is not an open question to a mirror-touch synaesthete - or a God-like superintelligence who could apprehend all possible first-person perspectives.

A counterargument might be that the existence of (dis)value in the world is inconsistent with the naturalistic third-person ontology of physical science. If eliminativist materialism were the case, this would be so. However, we're not zombies: first-person facts, not least the existence of (dis)valuable experiences, don't possess some sort of second-rate ontological status. If we assume Strawsonian physicalism, they are as much a part of the natural world as the rest mass of the electron. I can't define the normative aspect of disvaluable experience in terms of anything more semantically primitive. But if, for example, you try and hold your hand in ice-cold water for as long as you can, the experience is not motivationally inert. What property of the experience causes you to withdraw your hand?

Anyhow, this philosophising takes us a long way from the morally urgent question of nonhuman animal suffering. The worst source of severe and readily avoidable misery that exists in the world, today, i.e. factory-farming, is wholly manmade. Unless one is a complete moral nihilist, we have an obligation to stop it.

* * *

Talk of the case for moral clarity can make one sound like a religious conservative - and probably reads jarringly out of place on a blog like overcomingbias. However...

If we were discussing human child abuse, we'd all give short shift to a self-professed utilitarian who urged breeding happy toddlers who didn't mind being abused, and in the meantime he'll carry on as before because he likes the taste etc. A convergence of different indices suggest that a pig is at least as sentient as a human toddler. Yet human do things to pigs that would get us locked up for life if our victims were human. Insofar as we aren't complete moral nihilists, there are situations when one simply says: This is morally indefensible and should stop. The greatest source of severe, chronic and readily avoidable suffering in the world today is factory farming. Is it really ethically acceptable to wait several decades until selective breeding / in vitro meat or whatever brings our systematic abuse of sentient beings to a close?

* * *

Cyan, I wasn't making the claim that peripheral nerve ganglia suffer ["David Pearce: Nociceptors don't suffer - they send signals"] merely leaving open the possibility that they - and the multiple ganglia of segmented organisms without a central nervous system -undergo micropains. Noxious stimuli trigger the same opioid/dopamine neurotransmitter responses in roundworms and in people. And recall too the writhing tail shed by a lizard that has shed this appendage to distract a predator. This point arose in the context of srdiamond's question of whether it is arbitrary to prioritise unitary subjects of experience.

The insular cortex and suffering? Alas suffering is probably more evolutionarily ancient. Although the anterior insular cortex seems to be necessary for empathetic pain perception (cf. http://www.ncbi.nlm.nih.gov/pubmed/22961548) intensely unpleasant experiences are possible in the absence of any functioning insular cortex at all (cf. http://www.jneurosci.org/content/29/9/2684.full).

* * *

Drewfus, the paradox - to put it euphemistically - in arguing that "we ought not to use value judgements" is that this plea itself expresses a value-judgement. We can't have it both ways. Your account of ethics deserves a fuller treatment than I can offer here. So for now just a couple of observations. First, advocates of "ahimsa", and a cruelty-free vegan lifestyle, aren't predominantly the rich. Rather they are mainly some of the poorest people on the planet. Sophisticated Westerners write dismissively of "sacred cows"; but many Indians are less prone to arbitrary anthropocentric bias than Western species supremacists. Secondly, most if not all of the practices we now find morally offensive (e.g. human sacrifice, genocide, persecution of homosexuals as an abomination in the eyes of the Lord, etc) presuppose beliefs - and indeed entire conceptual frameworks - that are false. Much more controversially, I'd argue that we may cheat Hume's Guillotine with the following argument. In a nutshell, my agony has a primitive, irreducible normative aspect. This irreducible normative aspect might seem to have no bearing on the equivalent agony of other sentient beings: why should I care? But any such indifference on my part is a mere delusion of perspective, akin to the genetic fitness-enhancing perception that I'm the centre of the universe. Science teaches us that there are no ontologically privileged here-and-nows. So inasmuch as my agony is bad for me, then equivalent agony is bad for anyone, anywhere. However, a lot more needs to be said here to make any kind of value-realism work...

* * *

A note on language. By referring to other sentient beings as "livestock", we endorse the property status of nonhuman animals. Use of the term "animals" as distinct from humans is pre-Darwinian: "Human and nonhuman animals" is more cumbersome but less loaded. Even my restrictive use of the term "we" to refer to humans rather than all sentient beings expresses an insidious anthropocentric bias.

More substantively, I agree with Carl: breeding happier, less pain-ridden nonhuman [and human] animals is feasible now with existing technologies. Nonhuman animals on factory farms are so distressed they have to be prevented physically from mutilating themselves via debeaking, tail docking, (unanaesthetised) castration, etc. Any technology that reduces the burden of suffering in enslaved nonhuman [and human] animals is to be welcomed.

More fundamentally, in what circumstances is one ethically entitled to harm another sentient being? What ethical weight should we give the plea of "But I like the taste" to the killing, harming or abuse of human and nonhuman animals? Natural science teaches us to aspire to a "view from nowhere", an impartial God's-eye-view. Presumably, a notional full-spectrum superintelligence could impartially access and weigh all possible first-person perspectives, e.g. factory-farmed pigs and a meat-eating human diners, and act accordingly. Posthuman superintelligence is allegedly imminent; transhumanists sometimes worry about [human-]Friendly AGI. Yet is the notion of distinctively Human-Friendly AGI even intellectually coherent? (compare "Aryan-friendly AGI" or "Cannibal-friendly AGI"). Rather than aiming at humane exploitation, I think we should be aiming for high-tech Jainism, i.e. overcoming anthropocentric bias. In short, should humans devote our efforts to finding more humane ways to exploit other sentient beings, or instead finding ways to help them?

* * *

Srdiamond, I certainly wouldn't dismiss Moore's Open Question argument. Indeed, on the face of it, the argument is decisive. Whatever apparently dreadful or wonderful phenomenon-exists in the natural world, one can still ask if it's (dis)valuable. However, what is not an open question, at least for me, is whether my unbearable distress is disvaluable for me. And I'm arguing that it's only an epistemological limitation on my part, not some deep ontological truth about the world, that leads to any failure in my recognition that ( it is objectively the case that) your unbearable distress is disvaluable too. The badness of your agony is not an open question to a mirror-touch synaesthete - or a God-like superintelligence who could apprehend all possible first-person perspectives.

A counterargument might be that the existence of (dis)value in the world is inconsistent with the naturalistic third-person ontology of physical science. If eliminativist materialism were the case, this would be so. However, we're not zombies: first-person facts, not least the existence of (dis)valuable experiences, don't possess some sort of second-rate ontological status. If we assume Strawsonian physicalism, they are as much a part of the natural world as the rest mass of the electron. I can't define the normative aspect of disvaluable experience in terms of anything more semantically primitive. But if, for example, you try and hold your hand in ice-cold water for as long as you can, the experience is not motivationally inert. What property of the experience causes you to withdraw your hand?

Anyhow, this philosophising takes us a long way from the morally urgent question of nonhuman animal suffering. The worst source of severe and readily avoidable misery that exists in the world, today, i.e. factory-farming, is wholly manmade. Unless one is a complete moral nihilist, we have an obligation to stop it.

* * *

Srdiamond, Nonhuman animals do not have "experiences". They have experiences. Or are you proposing that suffering arose with the advent of Homo sapiens? This pre-Darwinian idea seems hard to reconcile with the genetic, neurological and behavioural evidence.

Humans clearly vary in their perspective-taking capacities, ranging from hyper-empathetic mirror-touch synaesthetes to victims of the more severe forms of autism spectrum disorder who lack any theory of mind at all. The mind-reading prowess of even low-AQ folk is still biased and selective; very few of us can aspire to a Jain-like concern for all sentient beings. For evolutionary reasons, ethnocentric and anthropocentric bias is endemic. Some people try harder than others to overcome it; other folk don't even try at all.

I'm not quite sure why you doubt that some people, at least, empathise with the victims of factory farming. For example, what do you think it feels like to be a pig castrated without an anaesthetic? The subjective agony of the castrated pig is as much a fact about the natural world as the rest mass of the electron or the second law of thermodynamics. If we aspire to full-spectrum superintelligence, then we'll need a richer understanding of third-person and first-person facts alike.

Signalling? Well, in common with discussions of psychological hedonism and the claim that everyone is "really" selfish, where do they ultimately leave us? For example, does devising an entirely self-interested account of the motivations of campaigners against (human) child abuse detract from the case against abusing children? Likewise with the case against abusing nonhuman animals - our fellow subjects of experience.

* * *

Srdiamond, Robinson Crusoe can be just as ethically concerned - or amorally unconcerned - with the interests of other sentient beings on his island as are contributors here about factory-farming. Empathetic intelligence is not purely about "signalling". Conceiving of other subjects of experience as "livestock", i.e. mere objects to be treated as property, reveals our profound cognitive deficits in perspective-taking capacity - the distinctively human intelligence that helped make one species the most cognitively successful on the planet. We need to enrich and de-bias this perspective-taking capacity, not stunt it.

* * *

Tim, imagine if someone claimed, "Surely the fates of today's Third World infants and toddlers are an insignificant side issue. So, please forget about signalling to each other how caring you are, and focus on the real problem at hand". You would give such an argument short shrift. But the plight of nonhumans animals of comparable sentience to human infants and toddlers is precisely what's at stake here.

Tim, we recognise that a human toddler with a progressive disorder who will never live to celebrate his third birthday deserves just as much love, care, and respect as his normally developing peers. In the case of members of our own species, we respect the interests of infants and toddlers for who they are, not for who they may - or may not - one day become. To deny that beings of equivalent sentience deserve equal consideration reflects mere speciesist bias on our part. No doubt such bias has been genetically fitness-enhancing; but it's also arbitrary, irrational, and ethically catastrophic.

* * *

Hedonic Treader, yes, the term "speciesism" is often misinterpreted. The antispeciesist does not claim that all species are equal, or that the well-being of a mosquito as valuable as a pig or a human toddler. Rather beings of equivalent sentience deserve equal consideration. Intuitively, animal advocates care disproportionately about the interests of non-humans. But few would argue that if human infants and prelinguistic toddlers were treated in the way we treat factory-farmed animals, then bringing the infant and toddler holocaust to an end would be our overriding ethical priority. If we grant what the scientific evidence suggests, namely that a pig is of comparable sentience to a human toddler, then the same considerations apply.

* * *

Tim, in common with humans infants and toddlers, nonhuman animals lack our sophisticated (and false) belief in an enduing metaphysical ego. So they don't set out to commit suicide. But in conditions of extreme distress, human and nonhuman animals alike will self-mutilate; harm themselves and their companions; and display all the signs of severe psychosis. This is the reason why factory-farmed animals are subjected to tail-docking, debeaking, [unanaesthetised] castration and so forth. For sure, self-serving humans can devise sophisticated reasons to rationalise animal abuse and child abuse if they have a taste for either; but on some fairly modest assumptions, they are both ethically indefensible.

Barnley, if you belive that "true self-awareness" is a precondition for suffering, do you also believe that "true self-awareness" is a precondition for pleasure? If so, does a loss of meta-cognitive capacity during sexual orgasm entail that orgasm isn't really pleasurable?

I'm sceptical. Either way, this perspective is ethically catastrophic.

Barnley, blind panic or raw agony are intensely unpleasant. Such experiences are characterised by an absence of meta-cognition. To claim that picnic-ridden and agony-racked humans don't suffer because they are too distressed to be capable of reflective self-awareness would be perverse. So what entitles humans to make such a claim of nonhuman animals? A pig for example, is of comparable sentience (and for what it's worth intelligence: http://www.nytimes.com/2009/11/10/science/10angier.html: "Pigs Prove to Be Smart, if Not Vain") to a human prelinguistic toddler. Only irrational speciesist bias prevents us from treating pigs and toddlers with comparable care and respect.

Game, the notion of humans versus "animals" is pre-scientific. There are human and nonhuman animals. Insofar as we want to be scientific rationalists, such pre-Darwinian dichotomies are best avoided.

VW, I worry you may be attacking a straw man. Drug-free life underpinned by gradients of bliss is feasible for human and nonhuman animals alike. Not least, the very happiest of hyperthymic people alive today serve as an "existence proof" that life animated by gradients of intelligent well-being is feasible. Conversely, some depressive people - and billions of factory-farmed nonhuman animals - spend almost their whole lives below Sidgwick's "hedonic zero".
Just how rich a quality of life we want our grandchildren to enjoy is an open question. But choosing and/or designing benign alleles of e.g. the COMT gene (cf. http://www.ncbi.nlm.nih.gov/pubmed/17687265) promises to enrich our quality of life without resorting to drugs or wireheading - and without loss of cognitive function or critical insight. By the same token, there is no technical reason why we can't extend our benevolence to sentient beings of other species. Of course, such benevolence would stand in stark contrast to the cruelties of factory-farming.

1 : 2 : 3 : 4 : 5 : 6 : 7 : 8 : 9 : 10 : 11 : 12 : 13 : 14 : 15 : 16

David Pearce (2013)
dave@hedweb.com

HOME
2024
2023
2022
2021
2020
2019
Talks 2015
Pre-2014 (FB)
Quora Answers
Facebook (2018)
Facebook (2017)
Facebook (2016)
Facebook (2015)
Facebook (2014)
LessWrong 2013
Some Interviews
The Abolitionist Project
Hedonistic Imperative FB group Postings