https://arstechnica.com/science/2024/04/philosopher-daniel-dennett-dead-at-82/
World-renowned philosopher Daniel Dennett, who championed controversial takes on consciousness and free will among other mind-bending subjects, died today at the age of 82.
"He was a towering figure in philosophy and in particular in the philosophy of AI," roboticist Rodney Brooks (MIT, emeritus) wrote on X, bemoaning that he'd never replied to Dennett's last email from 30 days ago. "Now we have only memories of him."
Dennett's many books, while dense, nonetheless sold very well and were hugely influential, and he was a distinguished speaker in great demand. His 2003 TED talk, "The Illusion of Consciousness," garnered more than 4 million views. While he gained particular prominence as a leader of the "New Atheist" movement of the early 2000s—colorfully dubbed one of the "Four Horsemen of New Atheism" alongside Richard Dawkins, Christopher Hitchens, and Sam Harris—that was never his primary focus, merely a natural extension of his more central philosophical concerns.
"Dan Dennett was the embodiment of a natural philosopher—someone who was brilliant at the careful conceptual analysis that characterizes the best philosophy, while caring deeply about what science has to teach us about the natural world," Johns Hopkins University physicist and philosopher Sean Carroll told Ars. "At the same time, he was the model of a publicly engaged academic, someone who wrote substantive books that anyone could read and who had a real impact on the wider world. People like that are incredibly rare and precious, and his passing is a real loss."
Dennett was a confirmed compatibilist on the fiercely debated subject of free will, meaning that he saw no conflict between philosophical determinism and free will. "Our only notable divergence was on the question of free will, which Dan maintained exists, in some sense of 'free,' whereas I just agreed that 'will' exists, but maintained that there is no freedom in it," Hoftstadter recalled.
I initially came across Dennett's writings in his book The Mind's I which he wrote with Douglas Hofstader including texts from other authors. It was when I first started to dip my toes into the philosopy of mind. He brings up some fascinating ideas from a rational perspective which always provoke a lot of thought and discussion.
Due to my being in the same philosophical camp as Dennett's great rival David Chalmers, I tend to side with those who called Denett's book, Consciousness Explained, Consciousness Avoided (Dennett actually wrote an epilogue in response to this accusation). This is because, as I understand it, Dennett has always strived to explain consciousness using only axioms derived directly from accepted science, which means it all starts and ends with a third person perspective. Dennett always raised some interesting counterarguments to other philosophers that attempted to discuss the really interesting first person phenomena of consciousness, so whether you agree with him or not, he represented a key school of thought in modern philosophy.
(Score: 5, Insightful) by jelizondo on Tuesday April 23, @02:05AM
A great mind passed away. Those of us who watched his many conferences and debates, were richly rewarded by his wit and intelligence.
It is a sad day
(Score: 4, Informative) by pTamok on Tuesday April 23, @09:57AM (12 children)
...to reach a considered opinion.
As part of the reporting, his 'story' 'Where am I(pdf) [lehigh.edu]" came to my attention. It's a short read.
It illustrates an annoying property of philosophical thought experiments: they are often not grounded in reality - in this case it uses the idea of a disembodied brain (or, of you like, a disembrained body) to examine the question of where the perception of 'I' is seated, and in doing so makes some blithe assumptions about, among other things, copying a working brain.
The philosophical point being made may not depend on the actual mechanics of doing such biologically implausible things, but for me, it gets in the way of the thinking.
Now Dennett argues that consciousness is an emergent property of brains, and an epiphenomenon (a 'side-effect' of having a brain) - a position I mostly agree with (on information theoretic and thermodynamics grounds*). Dennett's ideas have attracted many supporters and critics.
He's likely interesting to read - I'm told his work is easier to read than any others' in the field, but argues for a particular point of view that is not shared universally. Some (not all) of his books are described on Wikipedia, e.g.:
Nature: Cognitive Science: Dennett Rides Again: A review of "From Bacteria to Bach and Back: The Evolution of Minds" by Daniel C. DEnnett [nature.com]
Daniel Dennett (1991): Consciousness Explained [wikipedia.org]
Daniel Dennett (2005): Sweet Dreams [wikipedia.org]
Daniel Dennett (2006): Breaking the Spell [wikipedia.org]
Daniel Dennett (2017):From Bacteria to Bach and Back [wikipedia.org]
*Consciousness requires the processing of information. Processing information that is irreversible requires expenditure of energy (Landauer's principle [wikipedia.org]), so requires a substrate, so consciousness is tied to something physical that expends energy. Amongst other things, if a soul processes information, it means that an eternal soul requires an infinite source of energy.
(Score: 5, Informative) by pTamok on Tuesday April 23, @10:08AM
Long article on the BBC Future web-presence interviewing Daniel Dennett in December 2023: covers his views on Artificial Intelligence:
https://www.bbc.com/future/article/20240422-philosopher-daniel-dennett-artificial-intelligence-consciousness-counterfeit-people [bbc.com]
It's worth reading in its entirety, and following some of the links.
(Score: 3, Funny) by Anonymous Coward on Tuesday April 23, @10:20AM (3 children)
(Score: 3, Interesting) by pTamok on Tuesday April 23, @11:22AM (2 children)
"I am"
What is the 'I' of which you speak? Is it pre-reflective self-consciousness?
https://plato.stanford.edu/entries/self-consciousness/ [stanford.edu]
https://plato.stanford.edu/entries/self-consciousness-phenomenological/#toc [stanford.edu]
'am' is a state of being. How do you know that you exist: only by referencing memories, some of which are older (or younger) than others. Without memory, there is no experience of being.
This is Phenomenology, "the study of structures of consciousness as experienced from the first-person point of view."
https://plato.stanford.edu/entries/phenomenology/ [stanford.edu]
The intersection between neuropsychology and phenomenology is very interesting, and raises all sorts of questions about the nature of 'self', consciousness, and freedom of choice/intentionality.
J. NeuroSci: Predicting Perceptual Decision Biases from Early Brain Activity [jneurosci.org]
The reference bolded by me is to this paper: Spontaneous local variations in ongoing neural activity bias perceptual decisions [pnas.org]
Ann N Y Acad Sci. : Ann N Y Acad Sci. [nih.gov]
(Score: 3, Interesting) by Anonymous Coward on Tuesday April 23, @11:54AM
> Humans are capable of rapidly extracting regularities from the sensory environment 144 , 158 , 160 , 161 and there is strong evidence that the resultant expectations influence sensory processing.
Somewhere I recall this being tied to why we react so strongly (+ or -) to comedy? Things are moving along in some "regular" way until the punch line. This rips up whatever expectations we were assembling from the set-up to the joke.
(Score: 2, Informative) by Anonymous Coward on Wednesday April 24, @07:02AM
Just the barest minimum subjective experience of consciousness itself.
Not including any extra qualia: https://en.wikipedia.org/wiki/Qualia [wikipedia.org]
(Score: 3, Interesting) by VLM on Tuesday April 23, @07:20PM (1 child)
I had to read that as a philosophy elective during EE school back in the 90s. I did not like it. I felt it was deeply anti-engineering, which might be why they made us read it, kind of poke our brains?
An analogy of his story would be trying to ask the reader where is capacitance? Is it in the leads of a capacitor? Clearly not; at 100MHZ those leads are decent inductors and at 100GHz they're longwave antennas. Is it in the metal? Try it, if you snip the leads shorter the measured capacitance will remain about the same. What happens if you don't glue the electrodes to the dielectric and charge a cap and disassemble the plates and the glass insulator, pass them around the class, reassemble and no matter what your prediction, you will be unhappy about the charge on the cap (seen this done with my own eyes as an eye-witness).
But to what end, all this capacitance talk about how we can't recall define where capacitance "is"? Can we disprove the lumped constant model of EE design work? Sure, easy. Take like a 47GHz satellite amp and pump it into an old fashioned non-smd 100 uF power supply capacitor with thru-hole leads and you'll have a shitty dish antenna feed not a capacitor. How does that impact, say, a civilization built on switching power supplies? Does it make the civilization disappear or its power supplies suddenly stop working? No. Does anyone actually learn anything important from complaining about the inadequacies of the lumped constant model of EE circuit design? Nah. The Dennett article is essentially a slap in the face of the concept of a pragmatically useful model, or the concept of using models as a thinking and design tool at all, or even as a method to "learn".
Its kind of like an internet troll. Sure kid you should not use ohms law and not even bother learning it, because its a simplification and even good real-world materials are somewhat nonlinear and "real" knowledge, if it even exists, comes from finite element analysis electromagnetic code that requires a serious computer cluster to run. No kid you need a supercomputer to model a LED indicator on your car with femtosecond precision and even then you don't really understand LEDs. You need the same tools used to design waveguide couplers to make a LED light up on your car dashboard. I mean, technically, yeah, it is possible to do really trivial things with really complicated tools.
Or instead of trolling the kid you can say "well, ohms law says resistance equals volts divided by amps, so if your car runs around 14V on a good day and your LED wants around twenty thousandths of an amp, you want a resistor around 14/0.020 or about 700 ohms (and no need for a calculator if you're born before 1980) in line with the LED and since nothing in electronics is precise or linear just round up for safety to a common value like 820 ohms and call it good enough for government work. And that's real knowledge about electronics, but the approximation and model building about the world would make Dennett very unhappy. Where is the resistance? Is it in the atoms of the resistor or between the atoms of the resistor? Meanwhile the kid wants to light a LED.
Its sad to see philosophy which started out with "how to think the correct way" end up with Dennett's little article.
(Score: 2, Funny) by Anonymous Coward on Tuesday April 23, @09:21PM
At first I was going to ding you, with a smartalec remark like, "Hey, we do car analogies here, not capacitor analogies, are you in the wrong place?" But then, you clever bastard, you worked a car into the rant. Bravo!
(Score: 3, Informative) by acid andy on Tuesday April 23, @09:27PM (4 children)
If a soul processes information. Chalmers would probably argue that a self--best not to call it a soul in case people assume the definition comes with all sorts of religious baggage--that a self doesn't process information, it merely experiences it. Because it seems pretty clear that brains, calculators and abacuses can do all of the information processing required with no need to invoke a self having experiences (See philosophical zombies [wikipedia.org]; link includes Dennett's response to that argument), and yet we do have experiences, suggesting something more is going on. Chalmers solves the problem by suggesting that all matter has a subjective component (a type of panpsychism [wikipedia.org]), so wherever there is information, there is experience, for example a thermostat might have a limited form of consciousness.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 3, Interesting) by pTamok on Wednesday April 24, @10:51AM (3 children)
Chalmers: The Puzzle of Conscious Experience [consc.net]
I agree that the experience of consciousness is a puzzle.
I fundamentally disagree with the idea of positing an entire separate 'parallel universe' of 'experience' that is associated with information - "Perhaps information, or at least some information, has two basic aspects: a physical one and an experiential one. " - - a new 'True Theory of Everything' - that is the equivalent of invoking God when you don't understand what is going on: "Well, God does/did it". It also fails a basic scientific test - no falsifiable predictions (string theory has similar problems).
If you were to ask Searle's Chinese Room if it experiences consciousness, what would you expect the answer to be?
If you were to ask person if they experience consciousness, what would you expect the answer to be?
If the answers are different, why are they different, and if they are the same, how do you know any person other than you is not a philosophical zombie? And, how do you know you are not a philosophical zombie?
(Score: 3, Interesting) by acid andy on Wednesday April 24, @03:39PM (2 children)
It seems to me it's one of the simplest possible theories that fits the available evidence, given that we need to account for the existence of subjective experience. I get a sense you might be subconsciously projecting more attributes onto this 'parallel universe' than Chalmers suggests.
I think it's unlikely humans will ever be able to use objective scientific tests to determine the nature of a subjective self. At best we can rule out things that do not seem to be part of that domain, but that doesn't really help much--I think we're already close to running out of road there unless there's some dramatic new revelation about souls, life and death, with new repeatable experiments to test it. It seems highly unlikely.
I'd also argue that Dennett's theories come up against much the same problem. Why should we believe in the emergence of consciousness from physical matter? How do we prove emergence exists?
You don't. That's kind of the whole point of the thought experiment. From a third person perspective (i.e. any objective analysis) they are indistinguishable. Intuitively this seems correct: I really do have no idea whether anyone else is experiencing their existence or whether they are soulness machines.
That is the only thing I know for sure. The only thing I have direct (subjective) evidence of is the existence of my own subjective experience, therefore I know I am not a philosophical zombie. You can take my word for it, or not, as these words count for nothing from the perspective of anyone else.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 2, Interesting) by pTamok on Wednesday April 24, @09:47PM (1 child)
Occam's razor suggests not. You are right: I could be projecting, but I would suggest the extra entity - the experiential aspect - is unnecessary, or at least explicable with existing entities.
I think we are closer than you think. I suspect consciousness is a by-product of the physiological processes used to make and review memories. The self persists when you dream, when perceptual input is radically different and diminished compared to wakefulness. This implies that the self is not identical with perception. If the self can operate (at least temporarily) with diminished and possibly zero external input, we are pretty much left with it operating on memories. Given the parsimony of evolution and the link between 'self' and memory, I strongly suspect the two are connected quite fundamentally.
Emergence of complex behaviours from a collection of simple building blocks is well known. Emergence itself is uncontroversial. If we can define what consciousness is in physical terms, then demonstrating it can emerge from a simple substrate should not be difficult. I agree that definition is currently beyond us.
We agree.
I would posit that you don't know for sure. A philosophical zombie might inevitably generate a consciousness, whether you want it to or not, if it is 'sufficiently complex' to be a behavioural zombie (i.e. a 'perfect simulacrum', externally, of a human) - sufficient complexity might inevitably generate consciousness and self-awareness. The idea of a philosophical zombie might also be meaningless - in much the same way as trying to generate the colour red with the redness taken out. Alternatively, if a behavioural philosophical zombie is 'perfect', one of the behaviours could be a simulated sense of self, so what you believe to be your 'self' is simply a simulation of what a real 'self' is: and you have no way of telling.
Essentially you hit the problem Descartes had: you cannot demonstrate the existence of anything beyond your own present thoughts. There's a nice debate about whether logic and mathematical laws exist independent of thought - does the concept of 'the empty set' exist if there is no-one around to think it? If it does, what about human concepts like 'justice' and 'morality' - do they exist when people are not around to think about them?
A fun discussion, but unfortunately I can't continue it. I have other things I must be doing. At the end of the day, I have to assume the world exists, and certain behaviours are expected of me. Sigh.
(Score: 3, Interesting) by acid andy on Thursday April 25, @06:02PM
These are common objections which I believe Chalmers addresses--the guy's very exhaustive and methodical in his analyses. For my part I'll just say that the types of emergence we know about tend to be identical to patterns: arrangements of particles in space, patterns of behavior over time. These patterns often manifest as surprising phenomena, yes, but those phenomena can be fully understood in terms of the simpler entities that they emerge from.
I don't believe first person consciousness can be identical to such a pattern. Patterns exist as information. I experience information but I do not believe I consist solely of it. I can observe such information in the brain of another person and yet I do not become that person, therefore I am not the information (OK this last sentence may be logically shakey but like you I'm out of time).
To summarize my take on all this, if first person consciousness was some new alien phenomenon that we didn't posess ourselves, I think it would be uncontroversial to want to define it scientifically as a fundamental property of nature. This is what Chalmers suggests will be necessary. If Chalmers is wrong and conscious experience can emerge from a particular sufficiently complex physical system, in my view there is a very heavy burden of proof for someone to show scientifically how this is the case; there's nothing unscientific about taking something as being fundamental until someone can show otherwise.
I enjoyed it too. Some other time, perhaps.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 3, Interesting) by VLM on Tuesday April 23, @07:33PM
consciousness has to be taught to children as a meme or mind-virus or tradition or however you want to phrase it. Some "primitive" cultures have weird points of view about consciousness, at least weird from current western thought.
As a mind parasite, consciousness as a concept does not have to be constrained by physicality, or by "accepted science".
In a way both sides of the "debate" are correct. A computer virus spreads from computer to computer and does admittedly rely on the hardware being similar-ish but there is no discrete transistor you can point to in a 80386 that is "the transistor that runs computer viruses". The physical hardware exists and it does quite literally "run" everything, but the incorporeal stuff it runs operates outside the physical worlds rules. Actually not much of a surprise that some people think of computers as "magic". Because they are. At least as magic is sometimes thought of WRT mind-body dualism etc.