A. Is there any water in the refrigerator?
A. Where? I don’t see it.
B. In the cells of the eggplant.
Was “there is water in the refrigerator” true?1
But we don’t explain why, and it’s not obvious, so we’re being rude and unhelpful.
Believing true things matters. “HIV causes AIDS” is true, and figuring that out saved tens of millions of lives. Some politicians, religious leaders, and “alternative medicine” advocates said they “didn’t believe” HIV causes AIDS. On that basis, they blocked HIV prevention and treatment, causing millions of horrible, unnecessary deaths.
Irrationalists and anti-rationalists dismiss truth: either because true facts contradict their ideological agendas, or out of plain ignorance.
Open-minded rationalists may recognize that meta-rationalists come at truth from some other angle, and try to figure out what we could possibly be talking about. “Why would some smart people, seemingly without evil agendas, think it’s good to believe false things?” they wonder. “Maybe sometimes it’s pragmatically useful to believe false things—and therefore rational in a broader sense? How might that work?” And some rationalists develop interesting theories about that.
But this misses meta-rationalism’s point. It’s not useful to believe false things, for the same reason it’s not useful to believe true things. There aren’t any of either—not in the rationalist’s sense of true and false.
This may sound contradictory, or like an attack on an absurd straw man. Am I spinning facile paradoxes that will turn out to be meaningless word games?
I hope to persuade you otherwise: there are no contradictions here, and a clear understanding of “neither true or false in the rationalist’s sense” has great practical import.
“In what sense?” is a characteristically meta-rational question. “Yes, there is water: in the cells of the eggplant” is true in some sense—probably not a useful one. It’s false in another, more relevant sense.
The sense in which “HIV causes AIDS” is true is more complex, strange, mysterious, and interesting than you’d probably think. Later in this discussion, I’ll explain the relevant biology in detail—and I think you will be surprised! We will see that there are practical public health consequences to the unexpected sense in which “HIV causes AIDS” is true.
This exposition is long; technical in places; abstract and philosophical in others; and discusses less-than-pressing questions like “is the eggplant a fruit?” So you may wonder what the point is. Does “meta-rationality” matter, or is it meaningless ivory-tower philosophical wanking?
Meta-rationality matters because rationality matters. Rule of law, technology, and our post-subsistence economy are all products of rationality. They can’t survive without it. Rationality is under heavy bombardment from irrationalists, newly empowered by the internet, and civilizational collapse could result.
On the other hand, rationalism has conclusively failed in some respects. Enough smart people understand this that maintaining rational institutions by force is probably no longer feasible. Instead, meta-rationalism may deliver all the benefits of rationality, without rationalist errors. That might tip the balance of memetic power against irrationalists and save the world and stuff.
Here I hope to persuade you that wrong ideas about what “truth” and “belief” mean have large practical consequences, so getting a better understanding is important. I want to defend “truth” against irrationalists, but reject “truth” as misunderstood by rationalists.
What do “truth” and “belief” mean? These do sound like classic ivory-tower philosophical questions. Those are mostly nonsense and should be ignored. They are generally unanswerable; and having the “right” answers, if they existed, wouldn’t make any practical difference.
In practical terms, the meanings of “truth” and “belief” seem obvious enough. The sentence “HIV causes AIDS” is true because its meaning corresponds to reality: HIV does cause AIDS. Believing it means that you think the world is that way.
Meta-rationality concerns the relationship between rationality and the world. Rationalism has its own story about that, namely that rationality is the way to find beliefs (mental things) that are true (in the world). On careful investigation, this is empirically false. The world does not work in a way that could make any beliefs true in the sense that rationalism requires. The first main part of In the cells of the eggplant explains why not.
Important decisions, personal and political, often depend on knowing the truth of some matter. Rational methods are often the best way to find that. If rationality, pursued to its logical conclusion, endorsed the impossibility of truth or understanding, it would be really bad. But, whereas rationalism is mistaken, rational methods do often work. Meta-rationality is about how and why, and how to use rationality more effectively.
Recovering accurate, effective senses of “truth,” “belief,” and “rationality” requires a major re-thinking. The second half of The eggplant is about that.
- 1. This dialog is from the meta-rational book Understanding Computers and Cognition: A New Foundation for Design by Terry Winograd and Fernando Flores. I discuss it in detail later.