Back in November 2016, so 5 years ago, I wrote a post in response to an academic paper by Donald Hoffman and Chetan Prakash called Objects of Consciousness, where I specifically critiqued their ideas on biological evolution. Despite co-authoring the paper, I believe this particular aspect of their paper is predominantly Hoffman’s, based on an article he wrote for New Scientist, where he expressed similar views. One of his key arguments was that natural selection favours ‘fitness’ over ‘truth’.
...we find that natural selection does not, in general, favor perceptions that are true reports of objective properties of the environment. Instead, it generally favors perceptual strategies that are tuned to fitness.
One way to use fewer calories is to see less truth, especially truth that is not informative about fitness. (My emphasis)
What made me revisit this was an interview in Philosophy Now (Issue 147, Dec 2021/Jan 2022) with Samuel Grove, who recently published Retrieving Darwin’s Revolutionary Idea: The Reluctant Radical. According to Grove, Darwin was reluctant to publish The Decent of Man, because applying natural selection to humans was controversial, despite the success of The Origin of Species by Means of Natural Selection (full title). The connection to Hoffman’s argument is that Darwin struggled with the idea that evolution could ‘select’ for ‘truth’. To quote Grove:
Natural selection is premised on three laws: the law of inheritance, the law of variation, and the law of superfecundity (where organisms produce more offspring than can possibly survive). Together, these laws produce selection, and over the course of time, evolution. Well, Darwin’s question was, how could evolution produce a subject capable of knowing these very laws? Or, why would evolution select for fidelity to truth or laws? Selection favours survival, not truth. (My emphasis again)
Darwin turned to arguments, that as Grove points out, were ‘the common garden variety racism of the time’ – specifically, ‘group selection’ that favoured Anglo Saxon groups. Apparently, Darwin was reluctant to consider ‘group selection’ (as opposed to ‘individual selection’), but did so because it led to a resolution that would have been politically acceptable in his day. I will return to this point later.
So, even according to Darwin, Hoffman may have a point, though I’m not sure that Darwin and Hoffman are even talking about the same idea of ‘truth’. More on that later.
For those unfamiliar with Hoffman, his entire argument centres on the fundamental idea that ‘nothing exists unperceived, including space and time’. For more details, read my previous post, or read his co-authored paper with Prakash. I need to say upfront that I find it hard to take Hoffman seriously. Every time I read or listen to him, I keep expecting him to say, ‘Ah, see, I fooled the lot of you.’ His ideas only make sense to me if he believes we live in a computer simulation, which he’s never claimed. In fact, that would be my first question to him, if I ever met him. It’s an idea that has some adherents. Just on that, I would like to point out that chaos is incomputable, and the Universe is chaotic on a number of levels, including evolution, as it turns out.
In a previous life, I sometimes became involved in contractual disputes on major engineering projects (in Australia and US), preparing evidence for lawyers, and having to address opponents’ arguments. What I found in a number of cases, was that people prepared simple arguments that were nevertheless compelling. In fact, they often delivered them as if they were a fait accompli. In most of these cases, I found that by digging a little deeper, they could be challenged successfully. I have to admit that I’m reminded of this when I examine Hoffman’s argument on natural selection favouring ‘fitness’ over ‘truth’.
Partly, this is because his arguments highlight contradictions in his own premise and partly because one of his key arguments is contradicted by evidence, which, I concede, he may not be aware of.
For a start, what does Hoffman mean by ‘fitness’?
He talks about fitness in terms of predators and prey:
But in the real world where predators are on the prowl and prey must be wary, the race is often to the swift. It is the slower gazelle that becomes lunch for the swifter cheetah.
This quote is out of context, where he’s arguing that ‘swiftness’ in response, be it the gazelle or the cheetah, favours less information, therefore less time; over more information, therefore lost time. Leaving aside the fact that survival of either animal is dependent on the accuracy of their ‘modelling’ of their environment, if the animal being chased or doing the chasing ‘doesn’t exist unperceived’, then they might as well be in a dream. In fact, we often find ourselves being chased in a dream, which has no consequences to our ‘survival’ in real life. The argument contradicts the premise.
Hoffman and Prakash quote Steven Palmer from a ‘graduate-level textbook’ (1999):
Evolutionarily speaking, visual perception is useful only if it is reasonably accurate . . . Indeed, vision is useful precisely because it is so accurate. By and large, what you see is what you get. When this is true, we have what is called veridical perception . . . perception that is consistent with the actual state of affairs in the environment. This is almost always the case with vision . . . (Authors’ emphasis)
Hoffman and Prakash then argue that ‘using Monte Carlo simulations of evolutionary games and genetic algorithms, we find that natural selection does not, in general, favor perceptions that are true reports of objective properties of the environment’. In other words, they effectively argue that Palmer’s emphasis on ‘veridical perception’ is wrong. I can’t argue with their Monte Carlo simulations, because they don’t provide the data. However, real world evidence would suggest that Palmer is correct.
I read a story on Quora by a wildlife ranger about eagles who have had one eye damaged, usually in intra-species mid-air fights. In nearly all cases (he described one exception), an eagle who is blind in one eye needs to be euthanised because they would invariably starve to death due to an inability to catch prey. So here you have ‘fitness’ dependent on vision being accurate.
Leaving aside all this nit-picking about natural selection favouring ‘fitness’ over ‘truth’, how does it support their fundamental thesis that reality only exists in the mind? According to them, their theory of evolution ‘proves’ that reality doesn’t exist unperceived. Can you even have evolution if reality doesn’t exist (except in the mind)?
And this brings me back to Darwin, because what he didn’t consider was that, in the case of humans, cultural evolution has overtaken biological evolution, and this is unique to humanity. I wrote another post where I argue that The search for ultimate truth is unattainable, but there are 'truths' we have found throughout the history of our cultural evolution and they are in mathematics. It’s true that evolution didn’t select for this; it’s an unexpected by-product, but it has led to the understanding of laws governing the very Universe that even Darwin would be amazed to know.
18 comments:
This article is a blatant strawman of Hoffman’s hypothesis, and all of the refutations here are stupid. In fact, there are only 2 points that your report even makes, and the rest is merely your expression of dismay, to the point that it overcrowds the actual refutation. Much like I’ve done here.
“Leaving aside the fact that survival of either animal is dependent on the accuracy of their ‘modelling’ of their environment, if the animal being chased or doing the chasing ‘doesn’t exist unperceived’, then they might as well be in a dream. In fact, we often find ourselves being chased in a dream, which has no consequences for our ‘survival’ in real life. The argument contradicts the premise.”
Firstly, https://doi.org/10.3390%2Fe22050514 proves that the fitness functions in the games don’t need much accuracy for fitness. Secondly, if the gazelle is a conscious agent, it is likely, that the cheetah is too, and conscious agents exist unperceived. This refutation results from Hoffman not saying which animals are conscious and which aren’t, but that is only because conscious realism is incomplete. Thirdly, Hoffman has said millions of times that the ICONS OF OUR INTERFACE ARE SUPPOSED TO BE TAKEN SERIOUSLY, NOT LITERALLY. The recycle bin icon on your computer isn’t real, but it can ruin months of good work cause it still represents something that exists. Even if the cheetah isn’t real, it still represents some sort of danger in objective reality, likely some conscious agent interaction. Seeing the cheetah allows the gazelle to better understand what to do next than if it were to process the enormously complicated objective reality.
“I would like to point out that chaos is incomputable, and the Universe is chaotic on several levels, including evolution, as it turns out.”
Monte-Carlo simulations have an element of randomness that allows you to see all possible outcomes of a particular fitness landscape. If seeing any degree of reality loses out a million to zero, which is the value that Hoffman gives, it will be untenable to say chaos theory saves the day.
“an eagle who is blind in one eye needs to be euthanised because they would invariably starve to death due to an inability to catch prey. So here you have ‘fitness’ dependent on vision being accurate.”
🤣😂😂😂🤣🤣. There is no way any sane person would make that argument. This is circular reasoning since it presupposes that normal vision is accurate. Secondly, this only shows that having an impairment in sight is a selective disadvantage, which is to say that having lesser vision capabilities than required for fitness is a disadvantage. That says nothing about the accuracy of normal vision. Also, just because the vast majority of non-veridical perception is counterintuitive to fitness doesn’t mean that the specific perception required for fitness is verdicial.
The argument you have made from that premise is similar to saying since oranges which are not knives are less effective than knives at killing, all non-knives are less effective than a knife at killing. Nuclear weapons, a non-knife, are way better at killing than knives, even if most non-knives are pretty bad.
“their theory of evolution ‘proves’ that reality doesn’t exist unperceived.”
NO! According to Hoffman, there is a reality when unperceived, but it only consists of conscious agents, and conscious agent interactions are sufficient for universal Darwinianism.
Hi Anonymous. Thanks for providing a serious response, even if you feel the need to hide your identity. I freely acknowledge that I struggle to take Hoffman seriously. His version of reality is indistinguishable from what we experience in dreams. In fact, it is exactly what one experiences in a dream.
On the issue of natural selection, it works by literally killing off those lifeforms that cannot survive in a specific environment before they can reproduce. This is an important point, because the core of my argument is built on this simple yet fundamental premise.
A dictionary definition of ‘veridical’ is truthful; coinciding with reality. I would say it’s subjective. We are all aware of optical illusions, but the litmus test for the ‘truthfulness’ of a perception is our interaction with it; it’s what every sentient creature does every waking moment of their lives. Contemporary models of consciousness contend that the brain is a ‘prediction-making’ organ, as opposed to a purely sensory one, which has been the traditional view. The only reason someone can catch a ball is because we have the ability to predict its trajectory, even allowing for the 120 millisecond delay from the sensory input to the brain’s cognisance of it. A cricket ball from a fast bowler can travel over 4 metres in that time. And a cricket ball is not a ‘conscious agent’ and will hit you if you take your eye off it; that is, if it’s ‘unperceived’.
I read your paper and perhaps your one of its authors, since you won’t give your name. I admit it’s very obtuse to my eye. My main point concerning natural selection is independent of any mathematical modelling – it works by killing off individual lifeforms, and in all cases, they die from the interaction with their environment, whether perceived or not.
So, in the case of the cheetah and the gazelle, they exist unperceived because they’re ‘conscious agents’. But what about the environment they are running through, including space and time; does it all cease to exist when they no longer ‘perceive’ it, which does happen in a dream.
In a YouTube interview with Hoffman, he once acknowledged that he wouldn’t step in front of a train because he conceded that it has ‘consequences’ (his word) even if it’s only a ‘desktop icon’ (my words).
Regarding your third-to-last paragraph, I give a specific example that demonstrates that ‘normal vision’ (for an eagle) has evolutionary advantage, whether I presuppose it or not. Confirming an assumption is correct by citing evidence is not circular reasoning.
I provide another discussion on the topic here.
“I read your paper; perhaps you are one of its authors since you won’t give your name.”
No, I am not, since actual scientists respond to serious objections in peer-reviewed papers. Legitimate scientists don’t go around commenting on blogs.
“what about the environment they are running through, including space and time; does it all cease to exist when they no longer ‘perceive’ it, which does happen in a dream.”
The environment doesn’t exist. That would be the reification fallacy, but there is something else that causes what we fallaciously reify to the environment. Hoffman proposes in conscious realism that conscious agent interactions cause everything, but we perceive that causation to be by inanimate objects instead. For example, let’s say lightning killed you, and you died. The lightning is only part of your illusion, it was rather conscious agent interactions that killed you in entirely imperceptible ways, but you and I would contend that it was the lightning’s fault.
“I give a specific example demonstrating that ‘normal vision’ (for an eagle) has an evolutionary advantage, whether I presuppose it or not.”
Yeah, everyone knows that normal vision is an evolutionary advantage. In fact, even Hoffman makes this assumption. That is not what you were arguing.
You concluded from that example: "‘fitness’ dependent on vision being accurate.” Which assumes that normal vision is accurate, to begin with. Not only have you not substantiated that claim, but that is also what you conclude. Get it?
“Confirming an assumption is correct by citing evidence is not circular reasoning.”
Your eagle example is not evidence that “normal vision is accurate”. It is only evidence that “normal vision’ (for an eagle) has an evolutionary advantage” unless you add the premise that accurate = evolutionary advantage, but that is precisely what you are concluding, hence circular reasoning.
I seriously don’t get why I am teaching a philosopher this.
“The litmus test for the ‘truthfulness’ of a perception is our interaction with it.”
The interface analogy exists precisely to counter this. It demonstrates that all perceived causation could just be the correlation-causation fallacy. There are entirely separate causal events in objective reality that cause what we reify to our illusion. This is the same possibility argued by solipsism and other versions of idealism. If I eat a banana, it might seem like the banana sated my hunger. Still, it could be that I only perceive that the banana is what did it, even if it was instead my decision to eat the banana that triggered an entirely different imperceptible causal events in objective reality that led to the fulfilment of my hunger. Basically, banana-eating never happened. Something else happened, but that something else manifests my illusion of banana eating.
Your eagle example is not evidence that “normal vision is accurate”.
So normal vision for an eagle is not accurate, yet it’s accurate enough to catch prey, obviously, which, as you attest, confers evolutionary advantage.
So lightning doesn’t physically kill you and ‘the environment doesn’t exist’. Therefore, there is nothing much we can agree on.
Like Hoffman, you don’t distinguish between reality and a dream, as per your last paragraph.
“it’s accurate enough to catch prey.”
Hoffman doesn’t even disagree with that. Hoffman only disagrees with what mainstream scientific realism calls “enough”. “accurate enough”, according to Hoffman, means extremely low. It can’t be 0 since our perception is based on reality but doesn’t resemble it.
“Like Hoffman, you don’t distinguish between reality and a dream, as per your last paragraph.”
Hoffman does make that distinction. It is just that your confusion stems from mixing up 3 different axioms:
1. Dreams
2. Perception of reality
3. Objective reality
Realism argues that there is a significant overlap between 2 and 3, with naive realism saying 2=3. Idealism and solipsism argue that 3 consists only of minds and 2 consists of things that disappear when unperceived, and 2 and 3 are vastly different. That is kind of Hoffman’s argument too.
Your problem arises cause of the way idealism and solipsism are marketed. People who want to spread the idea of anti-realism come up with the mystical notion that 3 is non-existent or that 2 replaces 3. This is patently absurd, where you get absurd conclusions like god is necessary for object permanence if idealism is true or your conclusion of 1=2=3. 3 can never be an empty set, as proven by “cogito, ergo sum”. 3 still consists of something, even if it is just disembodied minds, and this is not an insignificant detail. This 3 is distinct from 2 cause 2 consists of objects with properties and space-time.
In anti-realism, 1 and 2 are very similar, the only difference being that 2 is very precise. Your reductio ad absurdum is based on this premise and the premise that 2 replaces 3, which I’ve just disproven.
Anti-realism argues that minds cause everything in 1, 2 and 3. Mind is the noumenon that causes things in 3. 2 is the fake narrative, using non-existent objects, constructed by the mind, based on relevant causation in 3. 1 is a phoney narrative caused by the mind based on nothing.
According to Hoffman, the noumenon are conscious agent interactions, and 2 allows us to vastly simplify 3 so that we can understand what to do in 3 better than if 3 and 2 had more overlap.
If you read the link I gave you in my response to your first comment, you’d know that I distinguish between 1, 2 and 3, as you define them. I never used ‘the premise that 2 replaces 3’.
In fact, I contend that without 3 there can’t be 2, as I’ve already expounded upon. On the other hand, Hoffman argues that 3 is dependent on 2.
“I never used ‘the premise that 2 replaces 3’.”
No, you clearly used the premise: “Hoffman argues that 3 is dependent on 2.” to construct your Reductio ad absurdum of Hoffman’s argument. No one has ever made that argument; it is an elaborate straw-man of anti-realism.
According to Hoffman, 3 consists of conscious agent interactions and 2 consists of objects and properties of space-time. 3 gives rise to 2 but doesn’t resemble 2.
3 always gives rise to 2 regardless of your metaphysical position. The only difference is what is in 3 and the overlap between 3 and 2.
My point was that you say, "Hoffman argues that 3 is dependent on 2." even though he doesn't, and only mystics market it like that.
IN FACT, NEITHER SOLIPSISM NOR IDEALISM MAKES THAT ARGUMENT, BUT ALL YOUR ESSAYS SAY THEY DO. ALL METAPHYSICAL POSITIONS CONTEND THAT 2 COMES FROM 3.
Realism: 2 similar to 3 (which is what you argue) 3 = 2 = Objects of space-time
Idealism: 2 <> 3 (Hoffman) Where 3 = minds only and 2 = Objects of space-time
Solipsism: 2<>3, Where 3 = only my mind and 2 = Objects of space-time and other people
Dreams: 1
See how everyone agrees that there is a 3? Just disagrees what is in it?
"I distinguish between 1, 2 and 3, as you define them."
Yeah, but you argue that idealists and solipsists don't. My comment was pointing out that they do cause:
3 = mind
2 = Objects of space-time
And the mind has functions in 3 independent of creating 2. Get it?
You reframe my arguments so that you can claim I said something that I didn’t. I no longer take you seriously, and I don’t believe you understand Hoffman any better than I do.
“You reframe my arguments so you can claim I said something that I didn’t.”
No, that is what you are doing to Hoffman. My comments up until now are about how you are straw-manning Hoffman and all of anti-realism.
Re-read the comments. I AM TRYING TO CLARIFY YOUR MISCONCEPTIONS OF HOFFMAN’S ARGUMENTS. From comment 6 onwards, I am talking about how your misconceptions of Idealism and solipsism stem from how they are marketed and what Idealism and solipsism actually are.
Comment 6 onwards doesn’t refute your arguments. It clarifies your misconceptions about Idealism and Solipsism.
"On the other hand, Hoffman argues that 3 is dependent on 2."
The above is a blatant strawman of Hoffman's hypothesis, and that is what I've been arguing from comment 6 onwards.
I AM NOT SAYING THAT YOUR POINTS ARE WRONG. I AM SAYING THAT THEY ARE NOT A REFUTATION OF HOFFMAN CAUSE NEITHER HOFFMAN NOR IDEALISTS OR SOLIPSISTS MAKE THE POINTS THAT YOU ARE REFUTING.
“Like Hoffman, you don’t distinguish between reality and a dream, as per your last paragraph.”
Comment 6 talks explicitly about how Hoffman differentiates between reality and a dream.
You said that Hoffman doesn’t do this and the point of comment 6 was to tell you how he does so. Comment 6 was not a refutation of what you are saying.
I might be willing to argue with you, if we were in the same room. But, in this forum, it’s too easy to misconstrue what one says, whether deliberate or not.
Thank you for engaging.
"I might be willing to argue with you if we were in the same room."
Interestingly enough, I find it easier to express my thoughts online than in person, but many times it does feel like I am speaking English, whilst the other person thinks I am speaking greek and so responds in Spanish. If you get the metaphor.
Anyway, sorry if I came across as rude. Goodbye and have a good day.
Hello - I'm posting here as Anonymous-2. I find Hoffman's theory interesting, but have some reservations about it.
Firstly, I think it is too much vision-centered, and that may be a problem. It seems to me that in contrast to the myriads of incompatible or overlapping visual appearances of objects in the world, the sense of touch reveals the constant and therefore true size and shape of things. For instance, based on your understanding of what a cube is geometrically, you anticipate how it would feel if you took it in your hands, e.g. it would present sharp and straight edges that are felt to be equal in length etc.
Secondly, I wonder what Hoffman could say to the theory proposed by Steven Lehar, who seems to agree with him to a point, namely in acknowledging that our perception of reality is that of a virtual world, but, as far as I understand him, he accepts the mind-brain identity as the foundation of this virtual reality, and therefore thinks that we, our real selves in the real noumenal world are present there as totally embodied agents.
I confess I haven't heard of Steven Lehar.
You know there's a sense you have called proprioception, which allows you to know where every part of your body is in space, without sight or touch. In fact, we all take it for granted, but without it we'd struggle to do anything except sleep and breath. Even putting food in your mouth would suddenly become a trial.
Time and space is something that we 'sense', just to survive day to day. I wrote an essay about it 20 years ago, when I was a philosophy student.
Post a Comment