Steven Pinker the dangers of “pseudo intellectual bullsh t”

Ladies and gentlemen, good evening and welcome. I am thrilled tonight to welcome back to HowTo Academy two genuine superstars of the intellectual world. Professor Stephen Pinker is the author of The Language Instinct, The Blank Slate, many other books and of course the subject of tonight's discussion, Rationality. He's in conversation with Oxford's Professor Marcus de Sautoy. Please give them a huge welcome. Well I guess in this age of kind of fake news, quack kind of remedies for things, the idea of producing a book on trying to get people to think a bit straighter, I mean there seems to be a surge of irrational thinking. Was that the kind of motivation for you writing this book, just seeing that we seem to be losing our minds out there? And like I think probably many mathematicians, certainly many social scientists, I had a feeling that some of the tools that we use in our line of work ought to be given away, exported, that anyone can profit from better intuitions about probability and logic and Bayesian reasoning, namely calibrating your degree of credence in a hypothesis to strengthen evidence, game theory, mental tools that everyone should command. I didn't know of any course that packaged them all in one curriculum. There are a lot of courses on statistics and probability, but they tend not to have logic in their same curriculum or game theory. But of course, as soon as I mentioned to people I'm going to be teaching a course on rationality, so it was adventitiously timely, topical, and I of course couldn't avoid that topic. I guess sitting here as a mathematician, you're slightly preaching to the converted as far as I'm concerned, because rationality is really the way that I think. There was a quote to use of Leibniz, which I think is my kind of philosophy of life, that the only way to rectify our reasonings is to make them as tangible as those of the mathematicians, so that we can find our error at a glance. And when there are disputes among persons, we can simply say, let us calculate without further ado and see who is right. So why am I doomed to failure in trying to reduce everybody to calculations? One of them is that a lot of our knowledge doesn't consist of propositions that are true or false and implications that necessarily follow from them, which is the stuff of classical logic, but it consists of large numbers of probabilistic cues that we kind of add up and weigh to put into context. In logic, people often ask me what's the difference between logic and reason, logic and rationality. Logic is just one of the tools of rationality. And to do logic, you have to do something that can often be quite irrational, namely to get everything you know to concentrate only on what's stated in new chemists. You forget the entirety of your real knowledge. That's what makes logic and indeed academic thinking so foreign to people who haven't been diversified in schooling, non-Western peoples, children, where it never makes sense to forget everything you know in real life. But that's exactly what you have to do in logic. So the syllogism that you're asked to validate is, all plant products are healthy, tobacco is a plant product, therefore tobacco is healthy. Now that is a valid syllogism, but people get it wrong because they can't get out of their head, tobacco healthy? What are you kidding? Now, of course, that is highly relevant to anything you have to do, but in terms of the task stated as a logic problem, it's a logical problem. So logic is obviously an indispensable tool. Computers are logical silicon, and there are cases where you really do want to concentrate only on what's legal reasoning. For example, you may want to deliberately blind yourself to prejudicial factors like race and gender. In biology, sometimes you want to forget about superficial similarities, which we call a porpoise a mammal instead of a fish, where it's closer to a deductive system. So in formal science and in other formal law, you do want the gift of forgetting the entirety of your knowledge and concentrating on the premises. But generally in everyday life, that would be a good thing to do. I think the other point that you make in the book, which is, I think, relevant to this, but as a mathematician, when I'm trying to prove a theorem, I have to define things very explicitly. So that definition will absolutely pin down the object that I'm trying to prove something about. But actually, trying to pin down the definition of things that we use in everyday language is often a real problem. You talk about Wittgenstein's attempt to try and define what a game is. So you're trying to reduce it to a mathematical logical proof, but actually you can't pin down in language what a game is. Is that one of our problems? That's another reason why logic in the formal sense is often kind of useless for everyday reasoning. So a game is an example, a vegetable. Now, of course, there are domains of thinking in which we do have to stipulate rules and ascertain whether a fuzzy category fits the definition. For example, did Bill Clinton and Monica Lewinsky have sex? Sex is a fuzzy family resemblance category. On the other hand, the law has to deal with stipulated formal logical rules just so that the guilt and innocence aren't up to the whim of the judge. And a lot of legalistic reasoning consists of trying to shoehorn our fuzzy probabilistic concepts into these classical categories that have necessary conditional conditions. I guess you talk about Kahneman quite a lot in the book and his ideas of the fact that we have these two systems of thinking seem to be the sort of difficulty that some people are having with applying rationality. Our first kind of reaction, our intuitive idea about how to solve a problem is often not the right one. We need to take slow down, thinking fast and slow. The slow version is applying our kind of analytics. So is this kind of a battle between these two sorts of thinking that we have in our brain? We use this system one, the fast intuitive, which often just is wrong and we have to slow down to do the farthest kind of rationals. Indeed, and we can be fooled by certain intuitions that don't apply to a particular case. Another example from that style of question is if there are lily pads on a pond that double every day and after 30 days the pond is completely covered on what they would have been half covered. People say, well, you know, first they finish with 15, then maybe they adjust upward a little bit and the answer, of course, is page 29 because it doubles every day and it's full on the 30th and half covered on the 29th day. It's an example of how exponential growth is surprisingly unintuitive and it does, even though that might sound like a trick question, a gotcha, there are many cases in everyday life where failure to appreciate exponential growth has real consequences, such as people don't save early enough for retirement, even a small amount of money invested early on because of compound interest, which grows exponentially, can give you quite the nest egg by the time you're ready for retirement, or conversely, people who take out credit card loans where the credit card company charges interest on the interest can quickly wipe out a life savings. Another example, and this is one that tricks even experts, even the kind of person, I'm not going to name any names, but two different experts on statistical reasoning and fallacies that people have made for themselves, early in the days of the COVID pandemic, they said, well, the rate of death from COVID is comparable to stress growth, to mood, other things that we tolerate without shutting down society, you're much more likely to get a car crash, this is a familiar trope in experts on human probabilistic fallacies and biases, but in this case, because the infectious disease can expand exponentially, as every person we cough on not only gets sick, but becomes an infector who coughs on others, who then in turn become infectors, the death rate shot up exponentially and quickly overtook strep throat and so on. It's an unintuitive concept that can have real life consequences, but more generally, the mindset of distrusting your first intuition, your gut feeling, thinking it through to make sure you weren't a fool, is one of the traits that goes into a kind of rationality quotient, which is not the same as the intelligence quotient, although there is some degree of correlation, people who are more resistant to the quick but wrong answer, also tend to be more skeptical of things of the paranormal, they tend to be less susceptible to conspiracy theories, they even are less susceptible to, the technical term is pseudo-intellectual bullshit, namely, if you string together a lot of highfalutin fancy-schmancy words, it sounds like it could be some profundity, but it actually makes no sense, and you ask people, is this meaningful, where people who fall for these simple math problems are more likely to attribute meaning to the pseudo-intellectual bullshit. They're also more likely, on average, there are a number of correlations that neatly but significantly correlate with each other, probably less likely to get into accidents, less likely to get into debt, less likely to get into mishaps like locking their keys in the car, so a general habit of avoiding cognitive impulsiveness is a big component psychologically, it's a major component of rationality insofar as it can be distinguished just from the raw intelligence. I think the example of exponential growth is very relevant to today, I mean our governments just don't seem to learn what exponential growth is, they wait for it to become very big numbers and don't realize that one, two, four, eight is also exponential growth, it's just very small numbers. The classic example I often use is the person who invented chess gets asked for a reward by the king who enjoys the game and says put one grain of rice on the first square of the chessboard, two on the next four, and the king thinks he's got away with a very cheap price for the game of chess, but doesn't realize that the first row of the amount of rice might be a piece of sushi, but by the halfway across you've wiped out the whole rice population of India for 10,000 years. It's a wonderful example, I didn't use it in rationality, I used it in a previous book, I don't want to repeat my example, it is a wonderful example. Well actually I'm wearing my vampire count on counts tonight as well, and I found this one was quite good with kids actually, because it's quite a good proof that there are no vampires, because if there was a vampire then it has to feed on a human every month and then changes them into a vampire, and then you've got two vampires and so the next month they've got to feed on two other humans, so how long would it take to make the whole population of the planet into vampires? It's actually only 30 months or something like that, so that seems to be a good proof, but I used that to try to explain to kids what was happening with this virus, and that one seems to have worked out quite well. It's an interesting question as a psychologist and one who thinks often in evolutionary terms, and we have my friend Richard Dawkins here in the front row, very pleased to see him, but the question that we often ask is, what would be the best adaptive explanation for why we should be blind to experiential growth, given that's what organisms do, that's what pathogens do, that's what, you know, scum on a plumb does, and it's a bit of a mystery. The only thing I could, the best I could do is that probably experiential growth in nature just never goes on for very long. There's a law here called Stein's law, things that can't go on forever don't, and that in any finite environment, because experiential growth really does saturate it very quickly, that any living organism would pretty soon, its growth rate would be checked by waste, by running out of space, by fouling its own environment, so in fact probably most biological processes as we observe them are S-shaped rather than exponential. Yes, yes, and I suppose it's, you know, we don't actually have that good an experience of very large numbers, so, you know, very quickly we're just not aware of, if something does go, you know, a hundred people very well or something like that, and I think, you know, you actually have kind of seven lessons in the book, sort of trying to teach people each of these kind of different approaches to a rational approach to kind of solving problems, and one of the ones you've talked about already, sort of probability, I mean we seem to be very, very bad at probability. Our intuition often catches us out, and you have some lovely examples in the book showing why, you know, everyone's intuitive response is wrong when you Why do you think we are so bad at assessing risk? Yeah, and I, in fact, I don't, I don't put it that way, that we're bad, and that is a natural conclusion to come to, especially from reading the work of Ferski and Kahneman, Mahima's Ferski, Daniel Kahneman's late collaborator, who showed a number of the ways in which we are systematically tripped up. The gambler's fallacy, people think that if the roulette wheel lands on, if the ball lands on red seven or eight times in a row, it's more likely to land black the next time, even though every spin is always independent, and so it's a little less than 50% each time. The infamous conjunction fallacy, immortalized as the Linda problem after, this was in the 80s, so baby boomer protagonist, but Linda, because I don't think girls are named Linda anymore, but I grew up with lots of Lindas, so Linda is a, is very smart, she needs philosophy, she's kind of a socialist warrior, she marches in Black Lives Matter protests. What is the, what is the probability that Linda is a bank teller? What is the probability that Linda is a bank teller who's active in the feminist movement? And people tend to give a higher probability to the second, even though that violates the conjunction rule, namely the probability of A and B must be less than or equal to the probability of A, and the fallacy is like saying you're more likely to draw a red queen from a deck of cards than a red card, and when you think about that, that just can't be right. So that's another example of a systematic probability illusion, and yet another one is the availability bias, namely that people use availability of anecdotes, episodes, narratives from memory as a way of estimating probability. So notoriously they feel that plane travel is a hazard, and so you drive from one place to another, even though per passenger mile car travel is vastly more dangerous, but because every plane crash gets saturation coverage and the news of car crashes kind of dribble in a few deaths at a time, it's easy to remember a plane crash, not so easy to remember a dozen car crashes, and so our sense of probability is disturbed. So those are all examples that lead to the conclusion that we are oblivious to probability, we're probability blind, as one psychologist put it. That can't be right, and again there are many animals who have a keen sense of probability and say forage, and you really, since the world is Einstein to the contrary notwithstanding, God does play dice, or at least it's on scales that are relevant to our well-being and the well-being of all animals, and it'd be quite anomalous to have no sense of probability at all, and they try to resolve that paradox by drawing on the work of other psychologists. The way in which the problems are presented to people make a big difference, and people can be less foolish than you might think if you present probability in a more mind-friendly format. In particular, the concept of probability itself is actually a word with many senses. There isn't actually one rigorous concept called probability, but the word itself can refer to propensity, namely how likely is a physical object to behave in a particular way owing to its makeup. Frequency in the long run, you flip a coin a thousand times, how many of them will be heads? Degree of confidence in a single event, namely for this coin flip, on a scale from zero to one, what is your subjective confidence that it will land heads? And warrant by the evidence, the kind of probability supplied in a court of law, have I given you good reasons to believe, or how good are the reasons that I've given you to believe? People, like philosophers of probability, distinguish them. When problems are presented in asking for the probability of a single event, when you think about it, it's something of an enigmatic concept. What does it mean to ask what the probability is that Linda is a bank teller? Either she is a bank teller, or she isn't a bank teller. You might think, well, it's not a question about probability. That's a sense. The only meaning of probability that is sensible in the context of a single event has to be subjective degree of confidence. You don't have a thousand Lindas, so you can count out how many of them are bank tellers. People's sense of probability might correspond more to frequency in the long run, which is really the way in which we encounter probability relevant information in our lives, in a natural environment. Namely, of all the times that you eat a particular food, how often do you get sick? All the times in which you look for rabbit along this route, how many of them do you actually find rabbits? Indeed, when you present some of these problems to people not couched as probability of a single event, then people do better. Not all of them get the right answer, but instead of saying, here's Linda, you say, imagine 100 women like Linda. How many of them are bank tellers? How very few commit the presumption fallacy? Interestingly, some still do, but far fewer.