Katelyn Jetelina has been providing some useful information on this. Her conclusion at this point seems to be ‘more data needed’.
wilkox
Karma: 354
The epistemology was not bad behind the scenes, it was just not presented to the readers. That is unfortunate but it is hard to write a NYT article (there are limits on how many receipts you can put in an article and some of the sources may have been off the record).
I’d have more trust in the writing of a journalist who presents what they believe to be the actual facts in support of a claim, than one who publishes vague insinuations because writing articles is hard.
Cade correctly informed the readers that Scott is aligned with Murray on race and IQ.
He really didn’t. Firstly, in the literal sense that Metz carefully avoided making this claim (he stated that Scott aligned himself with Murray, and that Murray holds views on race and IQ, but not that Scott aligns himself with Murray on these views). Secondly, and more importantly, even if I accept the implied claim I still don’t know what Scott supposedly believes about race and IQ. I don’t know what ‘is aligned with Murray on race and IQ’ actually means beyond connotatively ‘is racist’. If this paragraph of Metz’s article was intended to be informative (it was not), I am not informed.
It seems like you think what Metz wrote was acceptable because it all adds up to presenting the truth in the end, even if the way it was presented was ‘unconvincing’ and the evidence ‘embarassing[ly]’ weak. I don’t buy the principle that ‘bad epistemology is fine if the outcome is true knowledge’, and I also don’t buy that this happened in this particular case, nor that this is what Metz intended.
If Metz’s goal was to inform his readers about Scott’s position, he failed. He didn’t give any facts other than that Scott ‘aligned himself with’ and quoted somebody who holds a politically unacceptable view. The majority of readers will glean from this nothing but a vague association between Scott and racism, as the author intended. More sophisticated readers will notice what Metz is doing, and assume that if there was substantial evidence that Scott held an unpalatable view Metz would have gladly published that instead of resorting to an oblique smear by association. Nobody ends up better informed about what Scott actually believes.
I think trevor is right to invoke the quokka analogy. Rationalists are tying ourselves in knots in a long comment thread debating if actually, technically, strictly, Metz was misleading. Meanwhile, Metz never cared about this in the first place, and is continuing to enjoy a successful career employing tabloid rhetorical tricks.
The section on ‘How do you do it?’ looks like a generalised version of John Platt’s Strong Inference, a method of doing science that he believed ‘makes for rapid and powerful progress’. The essence of Strong Inference is to think carefully about a scientific question (the goal) to identify the main competing hypotheses that have yet to be discriminated between (the blockers), and devise and perform experiment(s) that rapidly discriminate between them (taking responsibility to remove the blockers and actually perform the next step).
Strong inference consists of applying the following steps to every problem in science, formally and explicitly and regularly:
1) Devising alternative hypotheses;
2) Devising a crucial experiment (or several of them), with alternative possible outcomes, each of which will, as nearly as possible, exclude one or more of the hypotheses;
3) Carrying out the experiment so as to get a clean result;
1′) Recycling the procedure, making subhypotheses or sequential hypotheses to refine the possibilities that remain; and so on.He also favoured Actually Thinking, as opposed to scientific busywork:
We speak piously of taking measurements and making small studies that will “add another brick to the temple of science.” Most such bricks just lie around the brickyard. Tables of constants have their place and value, but the study of one spectrum after another, if not frequently re-evaluated, may become a substitute for thinking, a sad waste of intelligence in a research laboratory, and a mistraining whose crippling effects may last a lifetime.
And, part of the point here is “it is very hard to talk about this kind of thing”. And I think that if the response to this post is a bunch of “gotcha! You said this comment was bad in one particular way, but it’s actually bad in an interestingly different way”, that kinda feels like it proves Elizabeth right?
This seems like a self-fulfilling prophecy. If I wrote a post that said:
It’s common for people on LessWrong to accuse others of misquoting them. For example, just the other day, Elizabeth said:
wilkox is always misquoting me! He claimed that I said the moon is made of rubber, when of course I actually believe it is made of cheese.
and philh said:
I wish wilkox would stop attributing made-up positions to me. He quoted me as saying that the sky is blue. I’m a very well-documented theskyisgreenist.
The responses to that post would quite likely provide evidence in favour of my central claim. But this doesn’t mean that the evidence I provided was sound, or that it shouldn’t be open to criticism.
For several of the examples you give, including my own comments, your description of what was said seems to misrepresent the source text.
Active suppression of inconvenient questions: Martín Soto
The charitable explanation here is that my post focuses on naive veganism, and Soto thinks that’s a made-up problem.
This is not a charitable or even plausible description of what Martín wrote, and Martín has described this as a ‘hyperbolic’ misrepresentation of their position. There is nowhere in the source comment thread that Martín claims or implies anything resembling the position that naïve veganism is ‘made-up’. The closest they come is to express that naïve transitions to veganism are not common in their personal experience (‘I was very surprised to hear those anecdotal stories of naive transitions, because in my anecdotal experience across many different vegan and animalist spaces, serious talk about nutrition, and a constant reminder to put health first, has been an ever-present norm.‘) Otherwise, they seem to take the idea of naïve transitions seriously while considering the effects of ‘signaling [sic] out veganism’ for discussion: ‘To the extent the naive transition accounts are representative of what’s going on...’, ‘seems to me to be one of the central causes of these naive transitions...‘, ‘...the single symptom of naive vegan transitions...’.
(Martín also objected to other ways in which they believe you misrepresented their position, and Slapstick agreed. I found it harder to evaluate whether they were misrepresented in these other ways, because like Stephen Bennett I found it hard to understand Martín’s position in detail.)
Active suppression of inconvenient questions: ‘midlist EA’
But what they actually did was name-check the idea that X is fine before focusing on the harm to animals caused by repeating the claim- which is exactly what you’d expect if the health claims were true but inconvenient. I don’t know what this author actually believes, but I do know focusing on the consequences when the facts are in question is not truthseeking.
The author is clear about what they actually believe. They say that the claims that plant foods are poisonous or responsible for Western diseases are ‘based on dubious evidence’ and are ‘dubious health claims’. They then make an argument proceeding from this: that these dubious claims increase the consumption of animal-based foods, which they believe to be unethical, with the evidence for animal suffering being much stronger than the evidence for the ‘dubious health claims’.
You may disagree with their assessment of the health claims about plant foods, and indeed they didn’t examine any evidence for or against these claims in the quoted post. This doesn’t change the fact that the source quotation doesn’t fit the pattern you describe. The author clearly does not believe that the health claims about plant foods are ‘true but inconvenient’, but that they are ‘dubious’. Their focus on the consequences of these health claims is not an attempt to ‘actively suppress inconvenient questions’, but to express what they believe to be true.
Active suppression of inconvenient questions: Rockwell
This comment more strongly emphasizes the claim that my beliefs are wrong, not just inconvenient.
Rockwell expresses, in passing, a broad concern with your posts (‘...why I find Elizabeth’s posts so troubling...‘), although as they don’t go into any further detail it’s not clear if they think your ‘beliefs are wrong’ or that they find your posts troubling for some other reason. It’s reasonable to criticise this as vague negativity without any argument or details to support it. However, it cannot serve as an example of ‘active suppression of an inconvenient question’ because it does not seem to engage with any question at all, and there’s certainly nowhere in the few words Rockwell wrote on ‘Elizabeth’s posts’ where they express or emphasise ‘the claim that [your] beliefs are wrong, not just inconvenient’. (This source could work as an example of ‘strong implications not defended’).
Active suppression of inconvenient questions: wilkox[1]
...the top comment says that vegan advocacy is fine because it’s no worse than fast food or breakfast cereal ads...If I heard an ally described our shared movement as no worse than McDonalds, I would injure myself in my haste to repudiate them.
My comment does not claim ‘that vegan advocacy is fine because it’s no worse than fast food or breakfast cereal ads’, and does not describe veganism or vegan advocacy as ‘no worse than McDonalds’. It sets up a hypothetical scenario (‘Let’s suppose...‘) in which vegan advocates do the extreme opposite of what you recommend in the conclusions of the ‘My cruxes’ section of the Change My Mind post, then claims that even this hypothetical, extreme version of vegan advocacy would be no worse than the current discourse around diet and health in general. This was to illustrate my claim that health harms from misinformation are ‘not a problem specific to veganism’, nor one where ‘veganism in particular is likely to be causing significant health harms’.
Had I actually compared McDonalds to real-world vegan advocacy rather than this hypothetical worst-case vegan advocacy, I would have said McDonalds is much worse. You know this, because you asked me and I told you. (This also doesn’t seem to be an example of ‘active suppression of inconvenient questions’.)
Frame control, etc.: wilkox
Over a very long exchange I attempt to nail down his position:
Does he think micronutrient deficiencies don’t exist? No, he agrees they do.
Does he think that they can’t cause health issues? No, he agrees they do.
This did not happen. You did not ask or otherwise attempt to nail down whether I believe micronutrient deficiencies exist, and I gave my position on that in the opening comment (‘Veganism is a known risk factor for some nutrient deficiencies...’). Likewise, you did not ask or attempt to nail down whether I believe micronutrient deficiencies can cause health issues, and I gave my position on that in the opening comment (‘Nutrient deficiencies are common and can cause anything ranging from no symptoms to vague symptoms to life-threatening diseases’).
Does he think this just doesn’t happen very often, or is always caught? No, if anything he thinks the Faunalytics underestimates the veg*n attrition due to medical issues.
You did ask me what I thought about the Faunalytics data (‘Do you disagree with their data...or not consider that important...?’).
So what exactly does he disagree with me on?
This is answered by the opening sentences of my first comment: ‘I feel like I disagree with this post, despite broadly agreeing with your cruxes’, because I interpreted your post as making ‘an implicit claim’ that there are ‘significant health harms’ of veganism beyond the well-known nutritional deficiencies. I went on to ask whether you actually were making this claim: ‘Beyond these well-known issues, is there any reason to expect veganism in particular to cause any health harms worth spending time worrying about?’ Over two exchanges on the ‘importance’ of nutrient deficiencies in veganism, I asked again and then again whether you believe that there are health harms of veganism that are more serious and/or less well-known than nutrient deficiencies, and you clarified that you do not, and provided some useful context that helped me to understand why you wrote the post the way you did.
My account of the conversation is that I misread an implicit claim into your post, and you clarified what you were actually claiming and provided context that helped me to understand why the post had been written in the way it was. We did identify a disagreement over the ‘importance’ of nutrient deficiencies in veganism, but this also seemed explicit and legible. It’s hard to construe this as an example where the nature of the disagreement was unclear, or otherwise of ‘nailing jello to the wall’.
Wilkox acknowledges that B12 and iron deficiencies can cause fatigue, and veganism can cause these deficiencies, but it’s fine because if people get tired they can go to a doctor
I did not claim that fatigue due to B12 or iron deficiencies, or any other health issue secondary to veganism, is ‘fine because if people get tired they can go to a doctor’. I claimed that to the extent that people don’t see a doctor because of these symptoms, the health harms of veganism are unlikely to be their most important medical problem, because the symptoms are ‘minor enough that they can’t be bothered’, they ‘generally don’t seek medical help when they are seriously unwell, in which case the risk from something like B12 deficiency is negligible compared to e.g. the risk of an untreated heart attack’, or they ‘don’t have good access to medical care...[in which case] veganism is unlikely to be their most important health concern’. I did not say that every vegan who has symptoms due to nutritional deficiencies can or will go to a doctor (I explicitly said the opposite), nor that this situation is ‘fine’.
But it’s irrelevant when the conversation is “can we count on veganism-induced fatigue being caught?”
‘Can we count on veganism-induced fatigue being caught?’ is not a question raised in my original comment, nor in Lukas Finnveden’s reply. I claimed that it would not always be caught, and gave some reasons why it might not be caught (symptoms too minor to bother seeing a doctor, generally avoid seeking medical care for major issues, poor access to medical care). Lukas Finnveden’s comment added reasons that people with significant symptoms may not seek medical care: they might not notice issues that are nonetheless significant to them, or they might have executive function problems that create a barrier to accessing medical care. There’s nowhere in our brief discussion where ‘can we count on veganism-induced fatigue being caught?’ is under debate.
Bad sources, badly handled: wilkox
Wilkox’s comment on the LW version of the post, where he eventually agrees that veganism requires testing and supplementation for many people (although most of that exchange hadn’t happened at the time of linking).
I did not ‘eventually agree’ to these points, and we did not discuss them at all in the exchange. In my first comment, I said ‘Many vegans, including myself, will routinely get blood tests to monitor for these deficiencies. If detected, they can be treated with diet changes, fortified foods, oral supplementation, or intramuscular/intravenous supplementation.’
This sounds like you’re saying “I won’t prescribe B12 until my patient gives up oreos” or even “I won’t prescribe B12 until everyone gives up oreos”, which would be an awful way to treat people.[1]
[...]
You probably mean “I don’t think Elizabeth/anyone should spend time on veganism’s problems, when metabolic issues are doing so much more aggregate harm.”
I wouldn’t say either of these things. A quick and easy treatment like B12 replacement is not mutually exclusive with a long-term and difficult treatment like diet modification. (This is not an abstract question for me; prescribing a statin and counselling on lifestyle changes are both things I do several times a week, and of the two, the script is orders of magnitude easier for both me and the patient, but we’ll usually do both in parallel when treating dyslipidaemia.)
As I said earlier in the thread, I’m all in favour of you or anybody else spending time on making people aware of the risk of nutrient deficiencies associated with veganism and what to do about them. (Again, this is not an abstract issue to me; I routinely discuss, screen for, and treat nutrient deficiencies with vegan and vegetarian patients.) I do recognise that you’ve had some bad experiences doing this, which is unfair.
- ^
I’m not sure if you chose this example intentionally, but for what it’s worth: Oreos are vegan.
- ^
I absolutely agree. McDonalds and the other demons of the Western Diet cause much more harm, both in absolute terms and per capita. That was really my point; within the class of ‘health misinformation and disinformation that causes harm’, furphies about vegan nutrition are a comparatively minor problem.
You don’t just have a level of access, you have a type of access. Your access to your own mind isn’t like looking at a brain scan.
From my Camp 1 perspective, this just seems like a restatement of what I wrote. My direct access to my own mind isn’t like my indirect access to other people’s minds; to understand another person’s mind, I can at best gather scraps of sensory data like ‘what that person is saying’ and try to piece them together into a model. My direct access to my own mind isn’t like looking at a brain scan of my own mind; to understand a brain scan, I need to gather sensory data like ‘what the monitor attached to the brain scanner shows’ and try to piece them into a model. This seems to be completely explained by the fact that my brain can only gather data about the external world though a handful of imperfect sensory channels, while it can gather data about its own internal processes through direct introspection. To make things worse, my brain is woefully underpowered for the task of modelling complex things like brains, so it’s almost inevitable that any model I construct will be imperfect. Even a scan of my own brain would give me far less insight into my mind than direct introspection, because brains are hideously complicated and I’m not well-equipped to model them.
Whether you call that a ‘level’ or ‘type’ of access, I’m still no closer to understanding how Nagel relates the (to me mundane) fact that these types of access exist to the ‘conceptual mystery’ of qualia or consciousness.
The Mary’s Room thought experiment brings it out. Mary has complete access to someone elses mental state, form the outside, but still doesn’t experience it from the inside.
Imagine a one-in-a-million genetic mutation that causes a human brain to develop a Simulation Centre. The Simulation Centre might be thought of as a massively overdeveloped form of whatever circuitry gives people mental imagery. It is able to simulate real-world physics with the fidelity of state-of-the-art computer physics simulations, video game 3D engines, etc. The Simulation Centre has direct neural connections to the brain’s visual pathways that, under voluntary control, can override the sensory stream from the eyes. So, while a person with strong mental imagery might be able to fuzzily visualise something like a red square, a person with the Simulation Centre mutation could examine sufficiently detailed blueprints for a building and have a vivid photorealistic visual experience of looking at it, indistinguishable from reality.
Poor Mary, locked in her black-and-white room, doesn’t have a Simulation Centre. No matter how much information she is given about what wavelengths correspond to the colour blue, she will never have the visual experience of looking at something blue. Lucky Sue, Mary’s sister, was born with the Simulation Centre mutation. Even locked in a neighbouring black-and-white room, when she learns about the existence of materials that don’t reflect all wavelengths of light but only some wavelengths, Sue decides to model such a material in her Simulation Centre, and so is able to experience looking at the colour blue.
In other words: the Mary’s Room thought experiment seems to me (again, from a Camp 1 perspective) to illustrate that our brains lack the machinery to turn a conceptual understanding of a complex physical system into subjective experience.[1] This seems like a mundane fact about our brains (‘we don’t have Simulation Centres’) rather than pointing to any fundamental conceptual mystery.
- ^
This might just be a matter of degree. Some people apparently can do things like visualise a red square, and it seems reasonable that a person who had seen shapes of almost every colour before but had never happened to see a red square could nevertheless visualise one if given the concept.
- ^
Apologies for the repetition, but I’m going to start by restating a slightly updated model of what I think is going on, because it provides the context for the rest of my comment. Basically I still think there are two elements to our disagreement:
The Camp 1 vs Camp 2 disagreement. Camp 1 thinks that a description of the physical system would completely and satisfactorily explain the nature of consciousness and subjective experience; Camp 2 thinks that there is a conceptual element of subjective experience that we don’t currently know how to explain in physical terms, even in principle. Camp 2 thinks there is a capital-H Hard Problem of consciousness, the ‘conceptual mystery’ in Rafael’s post; Camp1 does not. I am in Camp 1, and as best I can tell you are in Camp 2.
You think that all(?) ‘mental states’ pose this conceptual Hard Problem, including intentional phenomena like thoughts and beliefs as well as more ‘purely subjective’ phenomena like experiences. My impression is that this is a mildly unorthodox position within Camp 2, although as I mentioned in my original comment I’ve never really understood e.g. what Nagel was trying to say about the relationship between mental phenomena being only directly accessible to a single mind and them being Hard to explain, so I might be entirely wrong about this. In any case, because I don’t believe that there is a conceptual mystery in the first place, the question of (e.g.) whether the explanandum is an utterance vs a belief means something very different to me than it does to you. When I talk about locating the explanandum at utterances vs beliefs, I’m talking about the scope of the physical system to be explained. When you talk about it, you’re talking about the location(s) of the conceptual mystery.
What you call “model” here would presumably correspond only to the the externally observable neural correlate of a belief, not to the belief. The case would be the same for a neural correlate of an experience, so this doesn’t provide a difference between the two. Explaining the neural correlate is of course just as “easy” as explaining an utterance. The hard problem is to explain actual mental states with their correlates. So the case doesn’t explain the belief/experience in question in terms of this correlate.
As a Camp 1 person, I don’t think that there is any (non-semantic) difference between the observable neurological correlates of a belief or any other mental phenomenon and the phenomenon itself. Once we have a complete physical description of the system, we Camp 1-ites might bicker over exactly which bits of it correspond to ‘experience’ and ‘consciousness’, or perhaps claim that we have reductively dissolved such questions entirely; but we would agree that these are just arguments over definitions rather than pointing to anything actually left unexplained. I don’t think there is a Hard Problem.
It would be unable to distinguish between a p-zombie, which does have the neural correlate but not the belief/experience, and a normal person.
I take Dennett’s view on p-zombies, i.e. they are not conceivable.
So it seems that either the explanation is unsuccessful as given, since it stops at the neural correlate and doesn’t explain the belief or experience, or you assume the explanandum is actually just the neural correlate, not the belief.
In the Camp 1 view, once you’ve explained the neural correlates, there is nothing left to explain; whether or not you have ‘explained the belief’ becomes an argument over definitions.
I would not count “psychotic” here, since one is not necessarily directly acquainted with it (one doesn’t necessarily know one has it).
Would it be fair to say then that by ‘mental states’ you mean ‘everything that the brain does that the brain can itself be aware of’?
I thought you saw the fact that beliefs are about something as evidence that they are easier to explain than experiences
I don’t think there is any connection between whether a thought/belief/experience is about something and whether it is explainable. I’m not sure about ‘easier to explain’, but it doesn’t seem like the degree of easiness is a key issue here. I hold the vanilla Camp 1 view that everything the brain is doing is ultimately and completely explainable in physical terms.
or that they at least more similar to utterances than to experiences
I do think beliefs are more similar to utterances than experiences. If we were to draw an ontology of ‘things brains do’, utterances would probably be a closer sibling to thoughts than to beliefs, and perhaps a distant cousin to experiences. A thought can be propositional (‘the sky is blue’) or non-propositional (‘oh no!’), as can an utterance, but a belief is only propositional, while an experience is never propositional. I think an utterance could be reasonably characterised as a thought that is not content to stay swimming around in the brain but for whatever reason escapes out through the mouth. To be clear though, I don’t think any of this maps on to the question of whether these phenomena are explicable in terms of the physical implementation details of the brain.
So accepting beliefs as explanandum would not be in principle different from accepting fears or experiences as explanandum, which would seem to put you more in Camp #2 rather than #1.
I think there is an in-principle difference between Camp 1 ‘accepting beliefs [or utterances] as explanandum’ and Camp 2 ‘accepting experiences as explanandum’. When you ask ‘What counts as a simple explanandum such that we would not run into hard explanatory problems?’, I think the disagreement between Camp 1 and Camp 2 in answering this question is not over ‘where the explanandum is’ so much as ‘what it would mean to explain it’.
It might help here to unpack the phrase ‘accepting beliefs as explanandum’ from the Camp 1 viewpoint. In a way this is a shorthand for ‘requiring a complete explanation of how the brain as a physical system goes from some starting state to the state of having the belief’. The belief or utterance as explanandum works as a shorthand for this for the reasons I mentioned above, i.e. that any explanation that does not account for how the brain ended up having this belief or generating this utterance is not a complete and satisfactory explanation. This doesn’t privilege either beliefs or utterances as special categories of things to be explained; they just happen to be end states that capture everything we think is worth explaining about something like ‘having a headache’ in particular circumstances like ‘forming a belief that I have a headache’ or ‘uttering the sentence “I have a headache”’.
By analogy, suppose that I was an air safety investigator investigating an incident in which the rudder of a passenger jet went into a sudden hardover. The most appropriate explanandum in this case is ‘the rudder going into a sudden hardover’, because any explanation that doesn’t end with ‘...and this causes the rudder to go into a sudden hardover’ is clearly unsatisfactory for my purposes. Suppose I then conduct a test flight in which the aircraft’s autopilot is disconnected from the rudder, and discover a set of conditions that reliably causes the autopilot to form an incorrect model of the state of the aircraft, such that if the autopilot was connected to the rudder it would command a sudden hardover to ‘correct’ the situation. It seems quite reasonable in this case for the explanandum to be ‘the autopilot forming an incorrect model of the state of the aircraft’. There is no conceptual difference in the type of explanation required in the two cases. They can both in principle be explained in terms of a physical chain of events, which in both cases would almost certainly include some sequence of computations inside the autopilot. The fact that the explanandum in the second case is a propositional representation internal to the autopilot rather than a physical movement of a rudder doesn’t pose any new conceptual mysteries. We’re just using the explanandum to define the scope of what we’re interested in explaining.
This is distinct from the Camp 2 view, in which even if you had a complete description of the physical steps involved in forming the belief or utterance ‘I have a headache’, there would still be something left to explain, that is the subjective character of the experience of having a headache. When the Camp 2 view says that the experience itself is the explanandum, it does privilege subjective experience as a special category of things to be explained. This view asserts that experience has a property of subjectiveness that in our current understanding cannot be explained in terms of the physical steps, and it is this property of subjectiveness itself that demands a satisfactory explanation. When Camp 2 point to experience as explanandum, they’re not saying ‘it would be useful and satisfying to have an explanation of the physical sequence of events that lead up to this state’; they’re saying ‘there is something going on here that we don’t even know how to explain in terms of a physical sequence of events’. Quoting the original post, in this view ‘even if consciousness is compatible with the laws of physics, it still poses a conceptual mystery relative to our current understanding.’
Mental states do not need to be “about” something, but it is pretty clear they can be.
I’m still a bit confused by what you mean by ‘mental states’. My best guess is that you are using it as a catch-all term for everything that is or might be going on in the brain, which includes experiences, beliefs, thoughts, and more general states like ‘psychotic’ or ‘relaxed’.
I agree that mental states do not need to be about something, but I think beliefs do need to be about something and thoughts can be about something (propositional in the way you describe). I don’t think an experience can be propositional. I don’t understand this relates to whether these particular mental states are able to be explained.
It is clear that utterances are caused by mouth movements which are caused by neurons firing, but it is not clear how neurons could “cause” a belief, or how to otherwise (e.g. reductively) explain a belief.
My best account for what is going on here is that we have two interacting intuitive disagreements:
The ‘ordinary’ Camp 1 vs 2 disagreement, as outlined in Rafael’s post, where we disagree where the explanandum lies in the case of subjective experience.
A disagreement over whether whatever special properties subjective experience has also extend to other mental phenomena like beliefs, such that in the Camp 2 view there would be a Hard Problem of why and how we have beliefs analogous to or identical with the Hard Problem of why and how we have subjective experience.
Does this account seem accurate to you?
Huh, this is interesting. I wouldn’t have suspected this to be the crux. I’m not sure how well this maps to the Camp 1 vs 2 difference as opposed to idiosyncratic differences in our own views.
In your view however, if I’m not misunderstanding you, beliefs are more similar to utterances than to experiences. So while I think beliefs are equally hard to explain as experiences, in your view beliefs are about as easy to explain as utterances. Is this a fair characterization?
This is a fair characterisation, though I don’t think ease of explanation is a crucial point. I would certainly say that beliefs are more similar to utterances than to experiences. To illustrate this, sitting here now on the surface of Earth I think it’s possible for me to produce an utterance that is about conditions at the centre of Jupiter, and I think it’s possible for me to have a belief or a thought that is about conditions at the centre of Jupiter, and all of these could stand in a truth relation to what conditions are actually like at the centre of Jupiter. I don’t think I can have an experience that is about conditions at the centre of Jupiter. Strictly, I don’t think I can have an experience that is ‘about’ anything. I don’t think experiences are models of the world, in the way that utterances, beliefs, and thoughts can be. This is why I would agree that it is not possible to be mistaken about an experience, though in everyday language we often elide experiences with claims about the world that do have truth values (‘it looks red’ almost always means ‘I believe it is actually red’, not ‘when I look at it I experience seeing red but maybe that’s just a hallucination’).
I find the following obvious: thoughts or beliefs are on the same subjective level as experiences,
What do you see as the important difference between ‘subjective’ and ‘objective’? Is subjectivity about who has access to a phenomenon, or is it a quality of the phenomenon itself?
The reason I think utterances are “easy” to explain is that they are physical events and therefore obviously allow for a mechanistic third-person explanation. The explanation would not in principle be different from explaining a simple spinal reflex. Nerve inputs somehow cause nerve outputs, except that for an utterance there are orders of magnitude more neurons involved, which makes the explanation much harder in practice. But the principle is the same.
I agree with this.
It is unclear how to even grasp subjective beliefs and experiences in a mechanical language of cause and effect.
If for the sake of argument we strike out ‘beliefs’ here and make it just about experiences, this seems to be a restatement of the Camp 1 vs 2 distinction. As a Camp 1 person, a mechanical explanation of whatever chain of events leads me to think or say that I have a headache would fully dissolve the question. I wouldn’t feel that there is anything left to explain. From what I understand of Camp 2, even given such an explanation they would still feel there is something left to explain, namely how these objective facts come together to produce subjective experience.
Okay, so you are saying that in the first-person case, the evidence for having a headache is not itself the experience of having a headache, but the belief that you have the experience of having a headache.
Not quite. I would say that in the first-person case, the explanandum – the thing that needs to be explained – is the belief (or thought, or utterance) that you have the experience of having a headache. Once you have explained how some particular set of inputs to the brain led to that particular output, you have explained everything that is going on, in the Camp #1 view. Quoting the original post, in the Camp #1 view ‘if we can explain exactly why you, as a physical system, uttered the words “I experienced X”, then there’s nothing else to explain.’
So according to you, one could be wrong about currently having a headache, namely when the aforementioned belief is false, when you have the belief but not the experience. Is this right?
I would actually agree that ’you can’t be mistaken about your own current experiences’, but I think the problem Rafael’s post points out is that Camp #1 and Camp #2 would understand that to mean different things.
Intuitively it doesn’t seem possible to be wrong about one’s own current mental states.
I’m a bit confused about what you mean by ‘mental states’. It’s certainly possible to be wrong about one’s own current mental state, as I understand the term; people experiencing psychosis usually firmly believe they are not psychotic. I don’t think the two Camps would disagree on this.
The three examples you mention, of having a headache, being depressed (by which I assume you mean feeling down rather than the psychiatric condition specifically), and feeling pain, all seem like examples of subjective experiences. Insofar as this paragraph is saying ‘it’s not possible to be wrong about your own subjective experience’, I would agree, with the caveat as above that what I think this means might be different to what a Camp #2 person thinks this means.
So the explanandum, the evidence, would in both cases be something mental. But it seems you require the explanandum to be something “objective”, like an utterance.
I don’t require the explanandum to be an utterance, and I don’t think there’s any important sense in which an utterance is more objective than a thought or belief. My original comment was intended only to point out that in the first-person case you have privileged access to certain data, namely the contents of your own mind, that you don’t have in the third-person case. The reasons for this are completely mundane and conditional on the current state of affairs, namely that we currently have no practical way of accessing the semantic content inside each others’ skulls other than via speech. It’s possible to imagine technology that might change this state of affairs, like a highly accurate thought-reading device for example.
I do think the explanandum is required to be an output, because being able to explain or predict the output is the test of your model of what is going on. If you predict ‘this person is going to say they don’t have a headache’, and the person says ‘I have a headache’, then there’s something wrong with your model.
You’re absolutely right that this is the more interesting case. I intentionally chose the past tense to make it easier to focus on the details of the example rather than the Camp #1/Camp #2 distinction per se. For completeness, I’ll try to recapitulate my understanding of Rafael’s account for the present-tense case ‘I have a headache right now’.
From my Camp #1 perspective, any mechanistic description of the brain that explained why it generated the thought/belief/utterance ‘I have a headache right now’ instead of ‘I don’t have a headache right now’ in response to a given set of inputs would be a fully satisfying explanation. Perhaps it really is impossible for a human brain to generate the output ‘I have a headache right now’ without meeting some objective definition of a headache (some collection of facts about sensory inputs and brain state that distinguishes a headache from e.g. a stubbed toe), but there doesn’t seem to be any reason why this impossibility could not be a mundane fact conditional on the physical details of human brains. The brain is taking some combination of inputs, which might include external sensory data as well as introspective data about its own state, and generating a thought/belief/utterance output. It doesn’t seem impossible in principle that, by tweaking certain connections or using TMS or whatever, the mapping between these inputs and outputs could be altered such that the brain reliably generates the output ‘I don’t have a headache right now’ in situations where the chosen objective definition of ‘having a headache’ holds true. So, for Camp #1 the explanandum really is the output ‘I have a headache right now’. (The purpose of my comment was to expand the definition of ‘output’ to explicitly include thoughts and beliefs as well as utterances, and to acknowledge that the inputs in the case ‘I have a headache’ really are different to those in the case ‘John says he has a headache’.)
Camp #2 would say that it is impossible even in principle to be mistaken about the experience of having a headache. They might say it is impossible to meaningfully define ‘having a headache’ only in terms of sensory and/or introspective inputs to the brain. In their view, there is a sort of hard, irreducible kernel of experiencing-a-headache-subjective-qualia-stuff which is closely entangled with the objective inputs and outputs (they would agree that you are more likely to experience a headache if you were hit on the head with a hammer, and more likely to say ‘I have a headache’ if you were experiencing a headache), but nevertheless exists independent from and in addition to these objective facts and is not reducible to an account of only the inputs, outputs, and mapping between them. The explanandum, in their view, is the subjective-qualia-stuff. Camp #2 would fully admit that it’s really difficult to pin down the nature of the subjective-qualia-stuff; that’s why it’s a Hard Problem.
I’ve done my best here to represent Camp #2 accurately, but it’s difficult because their perspective is very alien to me. Apologies in advance to any Camp #2 people and happy to hear your corrections.
This is a clear and convincing account of the intuitions that lead to people either accepting or denying the existence of the Hard Problem. I’m squarely in Camp #1, and while I think the broad strokes are correct there are two places where I think this account gets Camp #1 a little wrong on the details.
According to Camp #1, the correct explanandum is still “I claim to have experienced X” (where X is the apparent experience). After all, if we can explain exactly why you, as a physical system, uttered the words “I experienced X”, then there’s nothing else to explain. […] In other words, the two camps disagree about the epistemic status of apparently perceived experiences: for Camp #2, they’re epistemic bedrock, whereas for Camp #1, they’re model outputs of your brain, and like all model outputs of your brain, they can be wrong.
I think this is conflating two difference senses of ‘claim’. The first sense is the interpersonal or speech sense: John makes a claim to you about his internal experience, in the form of speech. In this sense, ‘John claims to have a headache’ is the correct explanandum, in the Camp #1 view, of John telling you he has a headache, because it’s the closest thing to John’s actual experience that you have access to.
However, there is something different going on in the case where you yourself seem to have had an experience. You can believe you have had a certain experience without telling anybody about it, or without even uttering the words ‘I experienced X’ into an empty room, so the interpersonal or speech sense of ‘claim’ doesn’t really seem to apply. This only leaves us with the sense of ‘making a claim to yourself’, which might more precisely be called ‘thinking’ or ‘believing’.
Even in the Camp #1 view, there really is something different about a claim you make to yourself. You have privileged access to the contents of your own mind that you don’t have to contents of other people’s minds, by virtue of the mundane physical fact that the neurones in your brain are connected to the other neurones in your brain but not to the neurones in other people’s brains. Even if you don’t utter the words ‘I experienced X’, there is still something to be explained that lies between ‘actually experiencing X’ and ‘claiming in speech to have experienced X: why did you have the thought or belief ‘I experienced X’, instead of ‘I didn’t experience X but it would be useful for me to lie about it’? The explanandum in the case of your own experience is located a little deeper than it is in the case of the experiences of others. You can still be wrong about the underlying reality of your experiences – perhaps the memory of having a headache was falsely implanted with nefarious technology – but you have access to a type of evidence about it that John does not.
(I’ve never been able to figure out if Thomas Nagel, in ‘What is it like to be a bat?’, believes that the mere existence of this sort of privileged evidence about one’s own experiences tells us something about the nature of qualia/subjectivity. He says ‘The point of view in question is not one accessible only to a single individual. Rather it is a type.’ But, from my Camp #1 perspective, he never seems to explain what the difference is.)
So consciousness will be a densely connected part of this network – no more, no less – and it will have fuzzy boundaries because there is, ultimately, no ground truth as to what does or doesn’t constitute consciousness.
Perhaps this overly nit-picky, but I don’t believe Camp #1 intuitions imply that consciousness is or arises from a particular ‘part’ of the brain, in the sense that you could say ‘it comes from the neurones in this region’ or ‘it comes from the subset of neurones lighting up on this fMRI’, even allowing fuzzy boundaries. There’s no reason to expect the physical substrate of the brain, or even the network topology of the its connections, to always map straightforwardly to some feature or property of the mind, and particularly not for more abstract and higher-level properties. Sometimes there is such an obvious mapping (e.g. visual pathways), but there’s no more reason to expect that there is a ‘consciousness part of the brain’ than a ‘reasoning part of the brain’ or an ‘optimising part of the brain’; it might just be a thing that the whole brain is or does. By analogy, you might be able to point to a particular bit of circuitry in a computer that processes raw data from a camera sensor, but you can’t point to any one part and say ‘this is where the operating system comes from’.
The upshot is the same: Camp #1 will view consciousness as an ‘inherently fuzzy phenomenon’. We might just find it to be even fuzzier than you suggest here.
Surely you don’t think that’s the right moral category for ethical veganism?
I don’t really understand what you’re asking here. How would you describe the moral category you’re referring to, and why do you think it doesn’t or shouldn’t apply to veganism?
Thoughts on why the post gave (me) the impression it did, in no particular order:
‘Trade-offs’ is broad and vague, and the post didn’t make a lot of detailed claims about vegan nutrition. This makes sense in the context of you trying to communicate the detailed facts previously, but coming to the post without that context made it hard to tell if you were just making an unobjectionable claim or trying to imply something broader.
Some statements struck me as technically true but hyperbolic. Examples:
You can get a bit of all known nutrients from plants and fortified products, and you can find a vegan food that’s at least pretty good for every nutrient, but getting enough of all of them is a serious logic puzzle unless you have good genes.
[...]
Some people are already struggling to feed themselves on an omnivore diet, and have nothing to replace meat if you take it away.
[...]
If vegans are equally healthy but are spending twice as much time and money on food, that’s important to know.
[...]
If there are three vegan sources and you’re allergic to all of them, you need animal products.
I was very confused by what you were hoping to learn from an RCT or ‘good study’, and my impression of your in-office nutritional testing was that you were trying to gather new primary data about vegan nutrition. Because the basic facts about the risk of nutrient deficiencies in veganism seemed uncontroversial, my interpretation of that was that you thought there were other and potentially more significant ‘trade-offs’ that might exist at the level-of-evidence gap between an RCT or in-office testing and e.g. the introduction to the Wikipedia article on veganism.
The implied model for the relationship between diet and health felt...off, or at least different from my own. I tend to think of diet as an input to a set of homeostatic processes, which are generally robust but can be slowly pushed off-balance by sustained problems and usually fixed with gentle correction. This post seemed to model diet as an all-or-nothing ‘logic puzzle’ to be either solved or, more likely, failed.
...what you think could be done to convey the important, true points with as little animosity-due-to-misreading as possible
Beyond the points above, I genuinely don’t know. I’m probably not the intended audience anyway, since that seems to be vegans or potential vegans who don’t already know about the risks of nutrient deficiencies. The only thing I can think to contribute here is: if you’ve tried presenting the basic facts of the matter, and experienced pushback for it, does that necessarily mean that just presenting the facts is the wrong strategy? It’s a charged topic with a polluted local memespace, so some level of malicious or confused pushback is guaranteed; it’s what you would expect to see even when you are communicating well.
That is a frustrating situation. As you note in the introduction this is a charged topic that tends to lead to poor discussions, so you deserve credit for wading in anyway.
Given the context I laid out, is there anything I could have done to create a more productive discussion with you, personally?
I’m not sure. The discussion in this comment thread (and others) has been productive in the sense that I now have a much better understanding of your position and the context. In terms of the original post, I don’t know if a one-sentence summary would have changed much; given my impression of the post it might have looked like an attempt to motte-and-bailey. I could try to break down why the post gave me the impression it did, if you think that would be useful.
I’m as contemptuous as anyone of most ‘bioethics’, but this does gloss over the main objection. From the Nature News article (non-paywalled version):
In other words, the ethically fraught part is not the self-experimentation per se but promulgating what she did. There is a real risk that publication would, at the margin, increase the chances of somebody trying to do the same thing (or much more likely, a misunderstood and mangled version) and ending up harming themselves, either directly or indirectly by missing out on conventional treatment. I don’t think what she did was wrong, but I also don’t think it can be denied that there are genuine ethical considerations here. Calling for additional commentary seems like a pretty level-headed response.