Category: science

The dual engines of modern science

Posted on by Razib Khan

A few years ago Armand Leroi wrote The Lagoon: How Aristotle Invented Science. Some people immediately made a critique that actually, science, as we understand it, is really the creation of early modern Europe. That Aristotle and his fellow Ancients, or physicians and astronomers of early medieval Islam, or the scholastics of the high Middle Ages, didn’t “really” do “science.”

I think most of us understand where this critique is coming from. But, even if you grant the objection if Aristotle was alive today, would he be a scientist? Of course, he would go into science! And, he would probably a good one. Perhaps a great one. Why? Because he had the curiosity, cognitive skills, and, there is a culture that would allow him to flourish. To me, the biggest difference between early modern Western science, as it emerged in the 17th and 18th centuries, and what came before, is that it was a cultural concert of thinkers, a vast constellation of minds and minions.

In contrast, much of ancient science was driven by singular geniuses.

This brings me to the massive replication effort that just got published:

There are lots of angles to this story. Mostly good. But Jonathan Haidt pointed out how important this makes collaboration and a culture of truth-seeking within the enterprise. Alexandra Elbakyan has stated that her scientific activism is driven by “communist ideals.” And though I dislike Communism, I do think there is something fundamentally communistic about science. In Uncontrolled Jim Manzi points out that within the world of science there are very strong norms about honesty. A major issue with scientific fraud is scientists are trusting.

But then there is the von Neumann factor: geniuses can accelerate and open up whole landscapes of research. They do a “different kind of science.” It’s less culturally embedded, and less social and incremental. They are the sparks which fly in the darkness. Ratiocination machines of a different kind and ken from mortal man.

The moral of the story, if there’s any, is that modern science is a synthesis of these two aspects. There is the “industrial” aspect of scale, efficiency, and incrementalism. One step at a time into the darkness, cautious and continuous. This is science that Adam Smith would be proud of. International, specialized, and efficient. Optimized.

And then there are the startling breakthroughs. Sometimes those breakthroughs are genius and insight. Consider the story of the emergence of String Theory outlined in Lee Smolin’s The Trouble with Physics. Smolin is a skeptic of String Theory, but in the book, he describes how rapidly it took the scientific world by storm, just by force of its insight and elegance.

In contrast, there are cases such as CRISPR, where several different groups seem to have “stumbled” onto it (this resembles the “rediscovery” of genetics in the early 20th century). The genius here is less in the humans than in what nature had invented, and what humans discovered through trial, error, and luck. Nevertheless, in a few years, CRISPR radically transformed the possibilities in “genetic engineering.”

Going forward, big collaborative science will keep lumbering on. It will play the role that it has played for decades, driving translation, laying the seedbed for innovation. Normal science. But every now and then a spark will fly, and a new flame will explode. Genius still has a role to play in the firmament of human advancement.

The day we saw Stephen Hawking

Posted on by Razib Khan

In the spring of 1996, several of my dormmates decided to trek north to the University of Portland, to attend a speech by Stephen Hawking. We were still in that phase where we barely left campus, so intense was our social world. So this was a major undertaking. I don’t recall how we found out about the speech. This is before the internet was a widespread means of distribution of this sort of information (though I think we found out from my dormmate who was a journalism student).

I remember the anticipation and excitement. It was like we were going to a rock concert.

The talk Hawking gave was a typical one about cosmology. He also gave some shout outs to Linus Pauling, who was a native Oregonian.

Like many people, I had read A Brief History of Time. Also, perhaps like most people, I didn’t recall much from that book (I read the book years before going to the talk, in my defense).

Even by the mid-1990s, I was aware that Stephen Hawking was part of a somewhat out of control hype-machine. Though he was an eminent physicist, he was not necessarily the most brilliant physicist since Einstein (one of the claims on one edition of A Brief History of Time I saw at one point). We didn’t have Wikipedia, so I didn’t know about his somewhat messy personal life.

What we did know about Hawking was that he was a man of incredible brilliance who didn’t let his medical condition stop him. We admired him. We admired his achievements. He was heroic. By the time my dormmates and I saw Hawking in the flesh, he was already very frail. The only movement that we could perceive was that you could see he was breathing because of some barely perceptible movement around his neck.

At the end of the talk, people mulled around during the Q & A, trying to get as close as possible. I still have a vivid recollection of seeing Hawking up on the dais, in bright light.

Afterward, on the trip home, we reflected that it seemed unlikely that the great physicist had much time left, seeing as how he was nearly immobile. We felt privileged to be in his presence when he gave a talk. That was enough. Of course, he lived on for more than 20 years.

The postdoc salary range with cost of living (situation probably worst than reported)

Posted on by Razib Khan


Nature has an article, Pay for US postdocs varies wildly by institution. True, but as Matt Hahn, professor of biology at Indian University in Bloomington (cost of living 93% of the USA average) observed there isn’t any correction for cost of living. The researcher who dug through the data actually posted it online, so I decided to correct that oversight.

I took the institutions with N > 20, and looked up the cost of living in Best Places. The plot above is messy, but you can see that lots of institutions are paying a standard median salary of around $47,500, no matter the cost of living.

The correlation between cost of living and postdoc salary is 0.39. The weighted correlation is 0.48. These are pretty modest. That means you can find a really good situation, or a really bad one (also, institution reputation matters, there are some gems which pay well and have great reputations from what I can tell!).

Also, I’m pretty sure that the situation is worse than the numbers above suggest. Looking at the list of universities it seems there’s a bias for institutions at high cost of living locations not to want to report their salary data I think. Aside from UCSB the whole UC system denied the attempt to get data, and I don’t see Stanford, Columbia, or Harvard on the list.

The full table is below the fold, but adjusted for cost of living UCSB postdocs get $20,866 per year. In contrast, Michigan State, University of Maryland, Baltimore, and Wayne State University postdocs make more than $60,000 per year when you adjust. Stanford isn’t on the list, but online it says Stanford postdocs make between the low $50,000 to low $60,000 range, which seems reasonable for life sciences, though definitely poverty wages where the university is located (though if you are in a lucrative field it can be more, and depending on your supervisor outside consulting is a possibility, though good luck living in Silicon Valley on a $100,000 yearly gross income if you have a family, as many postdocs do).

Read More