Inter Caetera

Posts About

Books, debts, and delicatessen

Published on 18 October 2020

The trope namer for the "Matthew effect" is a verse from the Bible. It's been ripped out of context and appropriated by economists, social scientists, politicians, apparatchiks and representatives of other unfortunate professions to mean something completely different than when it is read in context. However, we will skip the spiritual and typically Catholic interpretations of the parable of the talents (I'm not nearly qualified enough to talk about that) and focus just on the material meaning of the verse.
The verse in question comes from Matthew 25:29 and reads as follows:
"For to all those who have, more will be given, and they will have an abundance; but from those who have nothing, even what they have will be taken away."
It seems unfair, that statement. However, one thing that the Bible is especially good with and why it should be read by both Christians and non-Christians alike (especially the gospels) is the ability to timelessly describe the state of the world and the way things work. And it seems fitting that this adage was adopted by economists — because it is very true in terms of material wealth and trade.
If you have something then it is possible to grow your wealth. Starting from one red paperclip the sky is the limit. But problems appear when you get to zero. At zero, there is nothing. And below zero — when you're in debt — things are even worse.
Games mirror real life like that, too. When you start a Monopoly game, everyone has equal amounts of in-game wealth, but over time one person gets considerably ahead and the remaining players get to zero. Once you're at zero, the game is over. Similarly in poker, everyone starts with an equal number of chips and in the end only one remains who holds them all. But in poker as in real life, if you have something, you can bounce back. All you need is a chip and a chair.

Mortality and immortality

A certain application of Matthew effect can be observed in mortal — or perishable — entities. One such entity is human life expectancy. Let's consider someone who is 35 years of age and has no visibly terminal illnesses. According to US SSA actuarial life tables, that man has a life expectancy of additional 46 years — meaning on average he is expected to die at 81. However, if we consider a man who also is not terminally ill, but 81, his average life expectancy is additional 8 years. If he made it that long, there's a chance that he'll make it even longer. At 89, it's 4 years more. At 93 it's another three. And so on.
This seems like an instance of Zeno's paradox, but actually we are just working with averages and statistics, so it's never going to precisely conform to reality. There are no averages in real life. And once you understand that, it makes perfect sense.
What's a little less intuitive — although is obviously true when one thinks about it for a moment — is that this also applies to non-perishable or intangible entities, but in exact reverse. This effect is well understood in sales, where the common belief is that the "first sale is the hardest". Once you have made the first one and it ends up being successful, further sales are more likely to happen than the first one.
Another such example is with books. Why do we still read the classics? Why are such works as Plato's Republic widely read during philosophy courses and Adam Smith's Wealth of Nations when studying economics? It's quite certain that there were many books written by their contemporaries, but being granted far less recognition — either due to inferior quality or other factors — have been relegated to the depths of obscurity.
If a book has been in print for six months, it might continue to be in print for another three. But if a book has been in print for a decade, it might continue to be in print for another decade. If a book has been printed for fifty years, it will likely continue for another century, and so on. (The numbers might not be right, but the idea remains.)

Pound pastrami and six bagels

This observation has been made most convincingly by Benoit Mandelbrot and later recalled by Nassim Nicholas Taleb in The Black Swan and Antifragile. It was named "Lindy effect" after a delicatessen in New York, where allegedly the original idea was coined, with respect to comedians' careers and repeat gigs.
This fact can help us make value judgements on non-perishable things — technology, ideas and so on. Time is a good determinant, because if something has been around for a long time, it implies that this thing has served the needs of the people well and will most likely continue to do so. Time is a natural filter for meaning. If a book has been popular for two thousand years, it's likely that it addresses some kind of universal truth or problem that has been relevant for that time. Likewise, if a piece of technology has only recently appeared, there is little chance that it will survive the next ten years. But there is no reason to think that, short of some kind of global catastrophe, we will stop using ubiquitous technology like the Internet or automobiles in those ten years.
What is important to remember is that there is a very important obstacle to these calculations, and it is that there is not a factor that greatly reduces our qualifier. Actuarial life tables are only relevant for people with no terminal illnesses — a 35-year-old man with colon cancer does not have life expectancy of 81 years. A four-hundred-year-old horse riding manual is likely going out of print after the first Ford Model T rolls off the assembly line. Cigarettes face a steep decline in sales once they are linked to lung cancer.
This leads me to a point I made already, a universal rule of things — once you identified a rule that governs a system, the things that conform no longer matter, and exceptions should instead be examined. Time is a great tool when it comes to making value judgements — but those judgements ought to be considered with regards to change and changing circumstances as well.

Time for space

Many of us, in our hubris, think that we live in some kind of special times, when the mysteries of nature have been laid bare before us and we have answers to all the questions. That we should not concern ourselves with tradition, that we know better than our fathers and their fathers before them. This mode of thinking is not unique to our generation, I think all or most thought the same of their own times. But, those who think so, do it at their peril.
In the early sixteenth century, an issue of great concern for the philosophers and theologians of the time was whether to grant humanity to the savages of the New World. It seems obvious to us now, from the vantage point of a half-millenium of hindsight, but we do not realise that a similar debate will soon rage about humanity of AI. And, without a doubt, the knowledge and wisdom of the past generations will be used to determine that, too.
We think of computing and programming like it's something new, something that we still have yet to explore, without understanding that aspects of it have always been with us. The principles of craftsmanship have been applied as successfully to masonry, goldsmithing and barrel-making as much as they are applied to software engineering. Those who can recognise these parallels and successfully apply past principles to modern problems will get ahead.
A software developer is not much different from a philosopher, both professions deal with making sense and applying structure to a set of abstractions. Only Aristotle and Aquinas did so theoretically, without direct influence on the world, and we do it practically.
"Computers, for the first time in history, facilitate the field of pure applied philosophy."
For every project that is a number of days behind its deadline, it will be behind even more. For every book that has readers will be given more, and it will have an abundance, and whichever does not have, it will be burned down and forgotten. For every app that has users, it will have more, and those which do not have any will be forever buried in the depths of Google Play store.
And there will be syntax errors and failing tests.