Deutsch and the Institute of Art and Ideas

Physicist David Deutsch recently did an interview with the Institute of Art and Ideas (IAI). There are many very good parts but I noticed some problems; I will focus on those.

On the topic of realism, Deutsch says:

Traditionally, science was just straightforwardly understood […] as being a project of understanding the world […].

This is the conception of science he agrees with, as he explains. He continues:

But in the 20th century, for various bad philosophical reasons, people started to question that, and they […] tried to champion ideas like ‘The world doesn’t really exist’, [by which] I mean the physical world, or that, if it does exist, we can’t have any knowledge of it, or that we are only playing with words, or that there isn’t a difference between a truer theory and a less true theory, and so on.

Deutsch is describing a misconception – anti-realism – and various related misconceptions in opposition to his own views. Among those views he holds that there can be truer theories and less true theories. I think there are several problems with this view.

First, it violates the law of the excluded middle: ideas can be either true or false. There’s no room for a third alternative, let alone for an infinity of degrees of ‘truthiness’.

Second, thinking in terms of degrees of truth is a form of justificationism: an epistemology which Deutsch spends a good part of his book The Beginning of Infinity criticizing.

Third, even if it was possible to justify ideas as more true than others, how would one objectively do that? Would it mean assigning a score of some sort? Based on what? How would one quantify that?

Einsteinian physics is better than Newtonian physics in the sense that the former solves problems the latter cannot solve while still allowing the latter to live on as an approximation, to put it in Popperian terms. There’s progress there; errors were corrected. But both theories are false, as far as we know. I think it’s fine to speak of progress toward truth and improvements but not more or less true.

A bit later, Deutsch says: “All human knowledge is false.” So then how can some of it be ‘truer’ than the rest? That sounds like an inconsistency. Also, I disagree that all our knowledge is false. Deutsch quotes Popper as saying we’re all the same in our infinite ignorance but that doesn’t mean that all our knowledge is false. This is a cynical view of human knowledge; it’s a result of taking fallibilism too far. Much of our knowledge is true, by which I mean 100% true, without any errors. We just don’t know which of our theories are true because there’s no criterion of truth. Deutsch breaks significantly in this regard with Popper who said, as quoted in the linked article, that “very many of the statements which we hold for truth are true” – but then Deutsch quotes Popper to substantiate a view Popper would disagree with.

Deutsch also says that false theories (such as Newton’s) can contain a ton of knowledge. I think that phrasing is fine. But I would avoid invoking comparatives of ‘true’. To be sure, this isn’t just a matter of using the right words: truth is an all-or-nothing thing. Even ‘slightly’ (whatever that means in this context) falling short of truth means utter falsehood. There’s a difference.

Next, the interviewer asks Deutsch:

You used the phrase “bad philosophical reasons” there. When you say that, are you saying more than just people being wrongheaded – are you saying that there was some malign motivation behind that, and what would you say that is?

Deutsch responds:

Not malign. I’m drawing a distinction between philosophy or philosophical theories that are false – like that there’s no such thing as infinity or that sort of thing – and philosophies that are bad, which are philosophies that intend to sabotage the debate or to prevent progress […]. So I call that ‘bad philosophy’.

‘Malign’ means “Evil or harmful in nature or effect.” Aren’t the sabotaging of debate and the prevention of progress harmful effects?

On the topic of final truth, Deutsch says:

[…] We definitely can’t expect to reach a final truth. And again, as Karl Popper said, if we did find it, we wouldn’t know, because, say, a true scientific theory has infinitely many consequences, and we can’t explore them all; we can’t even explore a finite fraction of them.

Well, it’s true that any fraction of infinity is always 0, but if you explore and corroborate ten predictions of a given theory, say, then you have still explored those ten. And that’s more than none. That’s progress, especially if you find that the theory is false. People could confuse the underlying mathematical notion for, again, epistemological cynicism.

On the topic of false memories:

Dan Dennett points out many cases of experiments on the brain where, by […] attaching electrodes to the place in the brain that receives tactile signals, we can not only make people experience things that aren’t happening, we can make them remember their own thoughts of a past that never happened. That’s because the brain makes up the differences in timings of nerve signals by confabulating what it itself thought. And all memory is like that, ultimately.

This is the first I’m hearing about these experiments so take my thoughts with a grain of salt, but I always get a bit suspicious when people don’t sufficiently differentiate between the brain and the mind. It’s not the brain that confabulates what it thought: it’s the mind, the person who does that. On the level of the brain, ie the hardware, the appropriate concepts to invoke are nerves and signals, things like that. But ideas, memories, the mind etc live on a higher level of emergence. It’s dehumanizing to reduce those to the underlying hardware; it’s a denial of the self (and of free will, which Deutsch invokes earlier in the interview) and casts people as nothing more than their brains, as machines.

The view of people as nothing more than their brains is part of a long-standing attack of the neurosciences on reason. In reality, people falsely remember things all the time. The high-level reason for that is simply their fallibility: people have ideas about how to interpret their sensations, and sometimes those ideas are simply wrong. The more detailed explanation for the unreliability of memory involves self-replicating ideas and their inevitable mutations.

If you’re skilled enough, you can plant false memories in people just by talking to them, no electrodes required; people who are dishonest with themselves conveniently remember what they want to remember; and so on. Memories are notoriously unreliable, to the point witness statements in court often blatantly disagree with objective records such as video footage. On a fundamental level, that has nothing to do with the brain: the substrate independence of computation dictates that your memories would be just as unreliable in this sense if your mind was transferred to a computer made of metal and silicon.

On the topic of ‘special access’ to one’s mind (as opposed to the outer world), Deutsch says:

[…] it’s not true that the interior of our brains and the interior of our thoughts is more accessible to us than the world we perceive through our senses […].

I agree. Compare my thoughts on mind reading.

[…] mathematical truths are based on conjecture.

No; they’re true independently of what anyone conjectures. Ideas about math are conjectures.