Econobabble - the reason people distrust math
You know, the thing I've noticed is that the proponents of the policies
such as the minimum wage and the proponents of nuclear weapons tend to
be using almost the same tactics when debating. The way of debating they
use is what I call econobabble (a word modelled after
technobabble). Here is how it goes:
However, it seems to me that many if not most of the people in the science of Croatian names of places make the opposite mistake: refusing to look into what some mathematical theory does indeed appear to say about the names of places. I am talking about what the Information Theory appears to say about that k-r pattern in the Croatian river names (Krka, Korana, Krapina, Krbavica, Kravarščica, and two rivers named Karašica). Mainstream linguistics insists that pattern is coincidental. However, using the methods taught at the Information Theory course in Computer Engineering curriculum, you can measure that the collision entropy of the syntax of the Croatian language is 0.214 bits per consonant pair, that the collision entropy of phonotactics is 1.623 bits per consonant pair, that the collision entropy of semantics is 5.992 bits per consonant pair, and that the collision entropy of morphology can at most be 1.572 bits per consonant pair. It appears to follow (whether it actually follows is, if you ask me, an open question), via the Birthday Paradox, that the p-value of that k-r pattern is somewhere between 1/300 and 1/17. The proponents of mainstream onomastics generally appear to be ignoring the issue, presumably because they think it's econobabble comparable to the stories about how the Game Theory proves nuclear weapons are a good thing or that Keynesian economics somehow argues for increasing the minimum wage. But it's different because it gives actual numbers (or, more accurately, confidence intervals). It gives us an actual way to numerically estimate the p-values. That's why it should be taken much more seriously. Especially since the apparent solution to the problem does not seem particularly outrageous. Why would it be so outrageous to suppose that *kurr meant "to flow" in Illyrian and that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix), which was borrowed into Proto-Slavic as *Kъrъrьsьja (regularly giving *Karaša, to which the Slavic suffix -ica may have been added)? Onomasticians seem to think it's an outrageous claim, but they seem to not be able to articulate why. In case you are interested, I made a video about that in the Latin language with English slides (in case your browser cannot stream it, download this MP4 file). So, when somebody tells you that some mathematical theory predicts something, the proper response is probably: "Can you give me the actual numbers?" and then (and only then) to ask: "Do you think there is a way to make your reading of that mathematics at least vaguely compatible with the widely accepted facts from social sciences?" (to make sure it's not a not-even-wrong argument).
I hope you realize now how damaging econobabble is and how prevalent it is.
Now, whether libertarians are committing the same mistake... Well, they are doing something similar when arguing that removing government subsidies to the egg industry would increase the prices of eggs and decrease their supply, and thus mitigate superbacteria (superbacteria are widely agreed to be mostly caused by the egg industry, because, well, 45% of all antibiotics used today are ionophores, and ionophores are antibiotics effective in birds but toxic to mammals). That is obviously (well, obvious to those who have actually bothered to study economics beyond the political trumpet) incompatible with economics, modern economic theories obviously argue that government subsidies have little or no impact on the prices of food. However, the difference is that libertarians here are going by the common sense, they are not asserting something that goes wildly against common sense (such as that putting the whole world into the danger of nuclear holocaust somehow makes us safer) and claiming it follows from some mathematical theory, all the while not even attempting to explain how it follows from it. Subsidies decreasing prices is probably common sense, that's why it's so hard to fight that idea. What really bothers me is the fact that the idea that nukes are a good thing also appears to be hard to fight, even though it is arguably very counter-intuitive. Why is it so?
- They make what appears to be an outrageous claim about something in social sciences.
- You ask them why they think that.
- They point to some mathematical theory which supposedly predicts that.
- You explain to them that the mathematical theory they point to, if anything, makes the opposite prediction.
- Instead of addressing your arguments, they go on to question your competence, almost as an ad-hominem.
Him: The nuclear disarmament of the 1990s is responsible for
the current geopolitical crises. Had the world had as many nuclear
weapons as it had in the 1980s, the world would be at peace.
Me: Why do you think that? Doesn't it seem to you more probable that, if not for the nuclear disarmament, we would have already had a nuclear holocaust? I think there is hardly a statement about politics more weird than the assertion that putting the entire human race into the danger of a nuclear holocaust somehow makes us safer.
Him: But the Game Theory predicts that. We've known this since ancient times. Have you not heard of the Latin saying "Si vis pacem, para bellum!"?
Me: How does the Game Theory predict that? I think the right analogy to this idea of nuclear peace might be the conspiracy theories, the Game Theory of which has been explained in the David Robert Grimes's famous paper On the Viability of Conspiratorial Beliefs. Just like conspiracies fall apart when somebody in it makes a mistake or doesn't feel like participating in the conspiracy any more, so too does nuclear peace fall apart when a nuclear weapon is fired by accident, as has almost happened quite a few times in recent history. And more countries having nuclear weapons and for an extended period of time massively increases the chance of that happening.
Him: Are you any kind of expert in international relations?
Me: Well, no, I am a computer engineer. I've studied some basic Game Theory from the Domagoj Kusalić'es book about programming, where he explained some basic Game Theory which is relevant to understanding Dynamic Programming.
Him: OK, then, let's just take a look at history. See how the violence in the world has been falling ever since the end of World War 2, that is, ever since the nuclear weapons have been invented?
Me: Now, this is an obvious since-when fallacy. The decrease in violence is a long-term trend that's been going on for centuries if not millennia. The 13th century was by an order of magnitude more bloody than the 20th century, there is no question about it among professional historians. Genghis Khan killed almost ten times more people than Hitler did in a world when the human population was by almost two orders of magnitude lower than it was at the time of Hitler. See the data about violence in the world from the Steven Pinker's book The Better Angels of our Nature and try to calculate the p-value of the Chow test of structural break in trend for the end of World War 2 and violence over time, I guarantee you the p-value will be around 50% of more.
Him: What kind of violence? Some forms of violence are relevant to nuclear weapons, while some are not. Domestic violence is not relevant to nuclear weapons, whereas the form of violence in which young men are forced to arm themselves and fight against the other group of young men is relevant.
Me: Then you are going to have to take into account when calculating the p-value the fact that you chose seemingly-arbitrarily one form of violence among many. Go ahead, try to calculate the p-value for that assertion that nuclear weapons made the world more peaceful, I guarantee you it will be very far from statistical significance.
Him: Obviously studying computer engineering has turned you into a lunatic. An incredibly bookish person without common sense. Previous generations valued people such as you because computer programmers were necessary. Now, in the age of Artificial Intelligence, computer programmers are not necessary, and you are nothing but a loser.
Does that conversation sound insane? Well, I had quite a few similar
conversations about the minimum wage:
Me: Why do you think that? Doesn't it seem to you more probable that, if not for the nuclear disarmament, we would have already had a nuclear holocaust? I think there is hardly a statement about politics more weird than the assertion that putting the entire human race into the danger of a nuclear holocaust somehow makes us safer.
Him: But the Game Theory predicts that. We've known this since ancient times. Have you not heard of the Latin saying "Si vis pacem, para bellum!"?
Me: How does the Game Theory predict that? I think the right analogy to this idea of nuclear peace might be the conspiracy theories, the Game Theory of which has been explained in the David Robert Grimes's famous paper On the Viability of Conspiratorial Beliefs. Just like conspiracies fall apart when somebody in it makes a mistake or doesn't feel like participating in the conspiracy any more, so too does nuclear peace fall apart when a nuclear weapon is fired by accident, as has almost happened quite a few times in recent history. And more countries having nuclear weapons and for an extended period of time massively increases the chance of that happening.
Him: Are you any kind of expert in international relations?
Me: Well, no, I am a computer engineer. I've studied some basic Game Theory from the Domagoj Kusalić'es book about programming, where he explained some basic Game Theory which is relevant to understanding Dynamic Programming.
Him: OK, then, let's just take a look at history. See how the violence in the world has been falling ever since the end of World War 2, that is, ever since the nuclear weapons have been invented?
Me: Now, this is an obvious since-when fallacy. The decrease in violence is a long-term trend that's been going on for centuries if not millennia. The 13th century was by an order of magnitude more bloody than the 20th century, there is no question about it among professional historians. Genghis Khan killed almost ten times more people than Hitler did in a world when the human population was by almost two orders of magnitude lower than it was at the time of Hitler. See the data about violence in the world from the Steven Pinker's book The Better Angels of our Nature and try to calculate the p-value of the Chow test of structural break in trend for the end of World War 2 and violence over time, I guarantee you the p-value will be around 50% of more.
Him: What kind of violence? Some forms of violence are relevant to nuclear weapons, while some are not. Domestic violence is not relevant to nuclear weapons, whereas the form of violence in which young men are forced to arm themselves and fight against the other group of young men is relevant.
Me: Then you are going to have to take into account when calculating the p-value the fact that you chose seemingly-arbitrarily one form of violence among many. Go ahead, try to calculate the p-value for that assertion that nuclear weapons made the world more peaceful, I guarantee you it will be very far from statistical significance.
Him: Obviously studying computer engineering has turned you into a lunatic. An incredibly bookish person without common sense. Previous generations valued people such as you because computer programmers were necessary. Now, in the age of Artificial Intelligence, computer programmers are not necessary, and you are nothing but a loser.
Him: Minimum wage is necessary not to make the economy
sluggish. It increases the velocity of money. It gives more money to
the people who will likely spend almost all of their income. If done
properly, it can even increase employment, rather than causing
unemployment.
Me: Is there any credible economic theory that predicts that?
Him: Well, it's the Keynesian economics that predicts that, the only school of economics widely taught at the universities.
Me: Are you joking? The Keynesian economics appears to be based on the idea that the minimum wage causes unemployment. The Keynesian economics explains the fact that the unemployment tends to progressively rise during the periods of economic booms by asserting that prices and wages are sticky, but that prices fall significantly faster than wages (lowering prices attracts customers, whereas lowering wages repells workers, so there is that incentive), so the unemployment is partly this integral of the difference between the curve of the prices through time and the curve of wages through time multiplied by the exponential function with the x=0 (and y=1) set on the current time (We multiply by the exponential function because the previous differences matter a lot less than the current ones.). Now how do you explain why sticky wages cause unemployment without implying that government-imposed minimum wage will do that as well? You cannot, right?
Him: Do you have some relevant education in economics?
Me: Well, I've taken one semester of economics as a part of my computer engineering curriculum.
Him: So, no.
If you see the absurdity in the above discussion about nuclear weapons,
you should see the absurdity in the above discussion about minimum
wage.Me: Is there any credible economic theory that predicts that?
Him: Well, it's the Keynesian economics that predicts that, the only school of economics widely taught at the universities.
Me: Are you joking? The Keynesian economics appears to be based on the idea that the minimum wage causes unemployment. The Keynesian economics explains the fact that the unemployment tends to progressively rise during the periods of economic booms by asserting that prices and wages are sticky, but that prices fall significantly faster than wages (lowering prices attracts customers, whereas lowering wages repells workers, so there is that incentive), so the unemployment is partly this integral of the difference between the curve of the prices through time and the curve of wages through time multiplied by the exponential function with the x=0 (and y=1) set on the current time (We multiply by the exponential function because the previous differences matter a lot less than the current ones.). Now how do you explain why sticky wages cause unemployment without implying that government-imposed minimum wage will do that as well? You cannot, right?
Him: Do you have some relevant education in economics?
Me: Well, I've taken one semester of economics as a part of my computer engineering curriculum.
Him: So, no.
However, it seems to me that many if not most of the people in the science of Croatian names of places make the opposite mistake: refusing to look into what some mathematical theory does indeed appear to say about the names of places. I am talking about what the Information Theory appears to say about that k-r pattern in the Croatian river names (Krka, Korana, Krapina, Krbavica, Kravarščica, and two rivers named Karašica). Mainstream linguistics insists that pattern is coincidental. However, using the methods taught at the Information Theory course in Computer Engineering curriculum, you can measure that the collision entropy of the syntax of the Croatian language is 0.214 bits per consonant pair, that the collision entropy of phonotactics is 1.623 bits per consonant pair, that the collision entropy of semantics is 5.992 bits per consonant pair, and that the collision entropy of morphology can at most be 1.572 bits per consonant pair. It appears to follow (whether it actually follows is, if you ask me, an open question), via the Birthday Paradox, that the p-value of that k-r pattern is somewhere between 1/300 and 1/17. The proponents of mainstream onomastics generally appear to be ignoring the issue, presumably because they think it's econobabble comparable to the stories about how the Game Theory proves nuclear weapons are a good thing or that Keynesian economics somehow argues for increasing the minimum wage. But it's different because it gives actual numbers (or, more accurately, confidence intervals). It gives us an actual way to numerically estimate the p-values. That's why it should be taken much more seriously. Especially since the apparent solution to the problem does not seem particularly outrageous. Why would it be so outrageous to suppose that *kurr meant "to flow" in Illyrian and that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix), which was borrowed into Proto-Slavic as *Kъrъrьsьja (regularly giving *Karaša, to which the Slavic suffix -ica may have been added)? Onomasticians seem to think it's an outrageous claim, but they seem to not be able to articulate why. In case you are interested, I made a video about that in the Latin language with English slides (in case your browser cannot stream it, download this MP4 file). So, when somebody tells you that some mathematical theory predicts something, the proper response is probably: "Can you give me the actual numbers?" and then (and only then) to ask: "Do you think there is a way to make your reading of that mathematics at least vaguely compatible with the widely accepted facts from social sciences?" (to make sure it's not a not-even-wrong argument).
I hope you realize now how damaging econobabble is and how prevalent it is.
Now, whether libertarians are committing the same mistake... Well, they are doing something similar when arguing that removing government subsidies to the egg industry would increase the prices of eggs and decrease their supply, and thus mitigate superbacteria (superbacteria are widely agreed to be mostly caused by the egg industry, because, well, 45% of all antibiotics used today are ionophores, and ionophores are antibiotics effective in birds but toxic to mammals). That is obviously (well, obvious to those who have actually bothered to study economics beyond the political trumpet) incompatible with economics, modern economic theories obviously argue that government subsidies have little or no impact on the prices of food. However, the difference is that libertarians here are going by the common sense, they are not asserting something that goes wildly against common sense (such as that putting the whole world into the danger of nuclear holocaust somehow makes us safer) and claiming it follows from some mathematical theory, all the while not even attempting to explain how it follows from it. Subsidies decreasing prices is probably common sense, that's why it's so hard to fight that idea. What really bothers me is the fact that the idea that nukes are a good thing also appears to be hard to fight, even though it is arguably very counter-intuitive. Why is it so?