Where have all the geniuses gone? Or: what's the real replication crisis?

This chap has asked the eminently sensible question: why have we basically stopped producing geniuses? 

Let's be clear: we are indeed well on the way to having stopped. Which is a bit odd, given quite how many people there are. The world now has billions of well-fed, disease-free, literate people living in sufficient comfort to turn their minds and hands to creating the Big New Thing and yet we're horribly underperforming the fewer and worse-off people of the past. 

Below is the key graph from the link above. It shows the number of acclaimed scientists (in blue) and artists (in red), divided by the 'effective' population (i.e., the total human population with the education and access to contribute to these fields).

Perhaps you doubt the methodology. Perhaps you don't find this graph as inuitively convincing as I do. If so, here's another experiment you can try. Take a decade of the 19th century at random and spend just a few of minutes on Wikipedia checking what things of note were created or discovered in those 10 years. 

I took the 1860s and this is what I found:

- The decade saw the publication of Les Misérables, War and Peace, Crime and Punishment, Alice's Adventures in Wonderland, Twenty Thousand Leagues Under the Sea, Great Expectations, Our Mutual Friend, The Mill on the Floss, Silas Marner and Das Kapital.
- In art, Monet painted Le déjeuner sur l'herbe, Renoir exhibited at the Paris Salon and was painted by Sisley, William Morris set up a wallpaper company, Rossetti was being upset after the death of his wife and Gilbert Scott was hard at work on the Midland Hotel (at St Pancras station) and the Foreign Office, among many other projects.
- In science, Alfred Nobel invented dynamite (patented 1867), James Clerk Maxwell published his equations that quantify the relationship between electricity and magnetism and show that light is a form of electromagnetic radiation, Lister developed the antiseptic methods for use in surgery in 1867, introducing carbolic acid as an antiseptic, turning it into the first widely used surgical antiseptic in surgery, and publishing Antiseptic Principle of the Practice of Surgery, Gregor Mendel formulated his laws of inheritance, the basis for genetics, in a two-part paper written in 1865 and published in 1866, and Dmitri Mendeleev developed the periodic table.
- In music, the decade saw the composition of Tristan und Isolde, Die Meistersinger von Nürnberg, Brahms' Requiem; Verdi's Requiem and Aida were both commissioned; Bruch wrote his violin concerto; Mussorgsky finished Night on Bald Mountain, Lizst wrote a coronation Mass, while Bruckner wrote three Masses and various motets including Locus Iste.

That's just a few minutes looking at Wikipedia. And it tells a story of creativity and advances across the full spectrum of intellectual and creative life that modern times simply cannot replicate. You can quibble all you want - you can doubt the significance of dynamite or the importance of Rossetti - but this is a hefty list: that science from just the years 1865-1867 was awesome - and that was at the same time as the world was creating two of its best ever requiems. Given that we are standing on the shoulders of such giants, we should be doing far, far more than we are. Or even just given the sheer number of human bodies, we should at least be producing 6 times as many ideas as the world of the 1860s: the population of the world in 1870 was in the region of 1.3bn, while its population now is getting close to 8bn. Forget about flying cars: where's my wallpaper with a better design than William Morris can do? Where are my 6 Bruch violin concertos?   

The author of the piece I linked to above, Erik Hoel, suggests that the reason for this catastrophe is the decline of what he calls aristocratic tutoring. (The word "aristocratic" is intended to refer to the kind of private tuition employed by the well-to-do for the general education of their children and to distinguish it from the kind of tutoring used by pushy parents nowadays to get children into competitive schools.)

That's just a very silly theory and we don't need to do more than look at the names in the list of achievements from the 1860s above to see that it's wrong: by and large, these people simply weren't aristocrats or people who had the education of aristocrats. And as for those people who did have the education of aristocrats, Grand Tour and all, what became of them? Not a lot, on the whole.

Sure, there's the odd Tolstoy on the list above, but he's very much the exception: at the same time, in Britain, Dickens (most definitely not an aristocrat raised by a succession of private tutors) was doing the heavy lifting in the big thick novels stakes. Or let's look at the education of George Eliot: "Thanks to her father's important role on the estate, she was allowed access to the library of Arbury Hall, which greatly aided her self-education and breadth of learning". One can see how that helped her. But we all have the entire internet nowadays! Plus free public libraries! Plus ... I don't need to go on. Imagine what Eliot could have done with the resources available to, say, Zadie Smith today.

Scott Alexander agrees with me (or perhaps vice versa) that aristocratic tutoring is a red herring. But even his analysis is not quite right. He points out that aristocratic tutoring is still very common in the world of musical and chess prodigies and suggest that that is a counter-example to Hoel's thesis. But that fact is not really a counter-example. The standard of performance in classical music and chess today is superb. Musical virtuosity is really quite common: it is not difficult, or at least so it seems to me, to find people alive today who can play the hardest pieces ever written; and it is not really disputed that the best chess players ever are alive right now. I couldn't tell you whether the number of virtuosi or chess grandmasters has tracked or exceeded the growth in the world's population, but it doesn't matter: the fact is that this kind of performance is relatively easily and predictably achieved. We actually have pretty good ways of generating Williams sisters or Polgar sisters, Yuja Wangs and Hilary Hahns. What we can't do is replicate Einsteins, Beethovens or Leonardos. 

It seems to me that there is something about that word "replicate" that is important. We can accurately replicate or reproduce so much nowadays. There is a famous anecdote about Mozart hearing Allegri's Miserere which I won't bother repeating or verifying, but the essence of the story lies in Mozart not being able to record the music. Today, of course, it's available to anyone for free at the touch of a button. Just try to imagine on how few occasions anyone would ever have heard a Beethoven symphony in their life before recording came along: now you are limited only by your amount of free time. Or try to imagine what it was like knowing the Mona Lisa or the Birth of Venus, the Parthenon or the Colosseum, only from a black and white line drawing or perhaps one visit in a lifetime: now we are all familiar with these works even if we never ever actually see them. We live in an age of abundant perfect replication.

And, it seems to me, replication is what we are really good at doing. We produce, again and again, people who perform the great works of the past. We can consistently train people to the highest standards of doing that. And not only in the field of artistic performance. What is the most recent potential successor to Das Kapital? It's Piketty's Capital! Even in political economy, the greatest effort is a self-conscious performance in the shadow of a greater ancestor.

It strikes me as entirely symptomatic of our age that the worry that nags away at so much science - perhaps not a worry so much as an almost existential concern - is what we call the replication crisis. We spend so much of our time living among reproductions of the glories of the past - polishing the glass on the great old paintings and re-staging operas in different clothes, painstakingly teaching millions of children to re-perform the advances of Leibniz and Newton in maths classes and those of Mendel in biology classes - that we throw up our hands in abject horror when we find that some of the latest findings in the cutting edge of science can't be replicated. 

I don't dispute that it's a Bad Thing that lots of science is wrong. But: 'twas ever thus, guys! If you want a replication crisis, try replicating Isaac Newton's cure for the plague: "the best is a toad suspended by the legs in a chimney for three days, which at last vomited up earth with various insects in it, on to a dish of yellow wax, and shortly after died. Combining powdered toad with the excretions and serum made into lozenges and worn about the affected area drove away the contagion and drew out the poison". The latest scientific thinking has always included wrong ideas - but when we had geniuses to push it forward nonetheless then that wasn't the greatest worry. Even now, when someone comes up with something as good - something as clearly breakthrough-y - as mRNA vaccines then we don't worry about the replication of results. 

Is it possible that trying to replicate findings that various small things make various small differences in diverting attention from finding the big things? I don't know. There are all sorts of reasons why coming up with new ideas is hard. Scott Alexander recites the boring but probably correct ones in his piece, and here are some other interesting ideas.  

But I do know this: failure to replicate what has gone before is not our biggest challenge today. Our problem is not too many mistakes made or too many blind alleys explored. Our problem is not that music is being produced that the public can't understand, or buildings are built that shock or astound us. If we had a mass of brave new theories in physics then we would be seeing an almost equally large mass of brilliant demolitions of their flaws and misconceptions, exposing today's phlogiston and ether - but we aren't. No, our real replication crisis is that the successes of our past are too glibly and slickly replicated, repackaged and reproduced for us to re-consume. 

When recordings and reproductions were almost as good as they are today, they were described as "high fidelity". Perhaps, at least in the intellectual and creative fields, it is time for some promiscuity.