For this review, I’d probably give it a +4, mostly for nicely summarizing the book well, but also because Daniel Dennett made some very useful productive mistakes, and identified a very important property that has to be explained, no matter what theory you choose, and I’ll describe it here.
The important property is that the brain is distributed, and this matters.
For the most likely theory by far on how consciousness actually works in the general case, see my review of Anil Seth’s theory, and the summary is that Anil Seth’s theory broadly solves the hard problem by reducing consciousness into specific components, and focusing on the specific properties instead of the general cloud of properties that consciousness invokes, and the cases where it’s wrong aren’t central, and are generalizable/patchable for more situations.
I basically agree with Global Workspace Theory as an explanation of why consciousness has the weird properties it does in humans, beyond the fact that we are conscious.
But Anil Seth’s model isn’t the focus, Daniel Dennett’s model is, so I’ll do that here.
On section 1, Rafael Harth argues that Dennett is making a distinction between 2 models of consciousness, and this difference comes down towards whether consciousness is more like a Cartesian theater, where there’s a single unitary actor, or the Multiple Drafts Model, where consciousness is essentially a draft we always rewrite throughout our lives, and there’s no ground truth to the matter, and the inputs go through a distributed processing system, and he favors the latter model of consciousness.
On section 2, the review shifts to a discussion of Heterophenomenology, where the premise is that consciousness debates should be treated like fictional works, where things are always in principle epistemically questionable (except for stuff that is impossible by pure logic), and Rafael Harth notes that this seemingly closes the door to Camp #2 style reasoning, though TAG has a counter argument here:
Also, it talks about how a distributed model is better than a unitary model, and how the Multiple Drafts model can solve quite a number of philosophical problems.
On section 3, I agree with Rafael Harth that Dennett doesn’t really justify his ideas well, and I found this quote below especially alarming in how badly he failed theory of mind for the intellectual views he disagrees with, and while I myself believe the Cartesian theater doesn’t exist, and tend to agree way more with Camp 1 over Camp 2, the quote is still indefensibly bad, and I’ll reproduce it:
It is time to recognize the idea of the possibility of zombies for what it is: not a serious philosophical idea but a preposterous and ignoble relic of ancient prejudices. Maybe women aren’t really conscious! Maybe Jews! What pernicious nonsense.
I have to agree with this explanation for why he doesn’t do well justifying consciousness:
At various points in the book, Dennett has a back-and-forth with a fictional character named Otto, who is introduced to represent the opposing view. These dialogues may have been the time to address Camp #2, but Otto is stuck in a strange limbo between both camps that offers little fuel for resistance. He’s condemned to keep opposing Dennett but never disputes heterophenomenology, so his resistance is bound to fail. He will usually reference some internal state, which Dennett will inevitably explain away, leaving Otto bruised and defeated.
For Section 4, we get treated to a few different topics.
One of them is about the distributed nature of the brain, but also about how a unified consciousness can emerge from this approach, if at all.
My own take is that I think 1 is partially true, in the sense that there is a way to synchronize consciousness into an approximately unified state, but that 2 is false, in that the unitary consciousness is not conceptually distinct from the parts that give rise to consciousness.
The binding problem is essentially solvable, but a lot of the specialness of the problem gets removed, IMO, because it’s at best approximately unified, and never perfectly unified, and the reason we perceive our consciousness as unitary is because the human body is small enough, and the latency is optimized such that you can almost immediately get conscious awareness of something, because there are hard latency constraints from nature, and if you fail, you get eaten (BTW this also explains why humans use so much energy, since latency needs to be optimized hard, and thus you have to go deep into unfavorable territory to use a brain in an external environment):
On Blindsight, Harth realizes that Dennett’s arguments are unfortunately not very falsifiable, and the natural strengthening of the claim is basically false as an empirical matter, which is a problem here.
On the Color Phi phenomenon, Harth is correct to question Dennett’s certainty, and this is a problem that the book as a whole shares, where it’s too certain of it’s own conclusions, and while the models it’s arguing against are pretty cartoonishly wrong, it’s too certain of itself.
Overall, I view Dennett as someone who made productive mistakes on the way to a better theory of consciousness, but also was too certain of itself in retrospect, combined with the bad theory of mind makes me agree with Harth that the books should not be read, and instead the reviews should be read.
However, the productive mistakes were productive enough, combined with the review being good at it’s job that I do have to give it a +4 here.
For this review, I’d probably give it a +4, mostly for nicely summarizing the book well, but also because Daniel Dennett made some very useful productive mistakes, and identified a very important property that has to be explained, no matter what theory you choose, and I’ll describe it here.
The important property is that the brain is distributed, and this matters.
For the most likely theory by far on how consciousness actually works in the general case, see my review of Anil Seth’s theory, and the summary is that Anil Seth’s theory broadly solves the hard problem by reducing consciousness into specific components, and focusing on the specific properties instead of the general cloud of properties that consciousness invokes, and the cases where it’s wrong aren’t central, and are generalizable/patchable for more situations.
More here:
https://www.lesswrong.com/posts/FQhtpHFiPacG3KrvD/seth-explains-consciousness#7ncCBPLcCwpRYdXuG
I basically agree with Global Workspace Theory as an explanation of why consciousness has the weird properties it does in humans, beyond the fact that we are conscious.
But Anil Seth’s model isn’t the focus, Daniel Dennett’s model is, so I’ll do that here.
On section 1, Rafael Harth argues that Dennett is making a distinction between 2 models of consciousness, and this difference comes down towards whether consciousness is more like a Cartesian theater, where there’s a single unitary actor, or the Multiple Drafts Model, where consciousness is essentially a draft we always rewrite throughout our lives, and there’s no ground truth to the matter, and the inputs go through a distributed processing system, and he favors the latter model of consciousness.
On section 2, the review shifts to a discussion of Heterophenomenology, where the premise is that consciousness debates should be treated like fictional works, where things are always in principle epistemically questionable (except for stuff that is impossible by pure logic), and Rafael Harth notes that this seemingly closes the door to Camp #2 style reasoning, though TAG has a counter argument here:
https://www.lesswrong.com/posts/naEDcbicBBykiTFwi/book-review-consciousness-explained-as-the-great-catalyst#N4JhNNqLupbX8maRx
Also, it talks about how a distributed model is better than a unitary model, and how the Multiple Drafts model can solve quite a number of philosophical problems.
For more on distributed decisions, read here:
https://www.lesswrong.com/posts/32sm7diYTky5KhF6w/distributed-decisions
On section 3, I agree with Rafael Harth that Dennett doesn’t really justify his ideas well, and I found this quote below especially alarming in how badly he failed theory of mind for the intellectual views he disagrees with, and while I myself believe the Cartesian theater doesn’t exist, and tend to agree way more with Camp 1 over Camp 2, the quote is still indefensibly bad, and I’ll reproduce it:
I have to agree with this explanation for why he doesn’t do well justifying consciousness:
For Section 4, we get treated to a few different topics.
One of them is about the distributed nature of the brain, but also about how a unified consciousness can emerge from this approach, if at all.
My own take is that I think 1 is partially true, in the sense that there is a way to synchronize consciousness into an approximately unified state, but that 2 is false, in that the unitary consciousness is not conceptually distinct from the parts that give rise to consciousness.
The binding problem is essentially solvable, but a lot of the specialness of the problem gets removed, IMO, because it’s at best approximately unified, and never perfectly unified, and the reason we perceive our consciousness as unitary is because the human body is small enough, and the latency is optimized such that you can almost immediately get conscious awareness of something, because there are hard latency constraints from nature, and if you fail, you get eaten (BTW this also explains why humans use so much energy, since latency needs to be optimized hard, and thus you have to go deep into unfavorable territory to use a brain in an external environment):
https://www.lesswrong.com/posts/zTDkhm6yFq6edhZ7L/inference-cost-limits-the-impact-of-ever-larger-models#ZHdTBrezWFagrAbkF
On Blindsight, Harth realizes that Dennett’s arguments are unfortunately not very falsifiable, and the natural strengthening of the claim is basically false as an empirical matter, which is a problem here.
On the Color Phi phenomenon, Harth is correct to question Dennett’s certainty, and this is a problem that the book as a whole shares, where it’s too certain of it’s own conclusions, and while the models it’s arguing against are pretty cartoonishly wrong, it’s too certain of itself.
Overall, I view Dennett as someone who made productive mistakes on the way to a better theory of consciousness, but also was too certain of itself in retrospect, combined with the bad theory of mind makes me agree with Harth that the books should not be read, and instead the reviews should be read.
However, the productive mistakes were productive enough, combined with the review being good at it’s job that I do have to give it a +4 here.