Evolutionary novelties abound,” writes paleobiologist Douglas Erwin in the introduction to his new book The Origins of the New. Take grasses. They’ve carpeted the continents in undulating blankets of green, blue, and brown for millions and millions of years. They’ve been so successful that they’ve completely altered ecosystems across the planet, modifying not just terrestrial habitat but regional climate. But it took grasses a long time to grow to such dominance. They first appeared 55 million years ago, and they were ecological nobodies for a long time after that. First, they were novel. Then, much later, they were wildly successful. There was a huge lag between the two.
This is more important than it might at first seem, argues Erwin. Today, many scientists equate novelty with innovation, and this poses a problem for our understanding and study of evolution. Erwin draws parallels with how economic historians look at the differences between invention and innovation. The one is patented. The other catches fire and transforms global economies. Erwin suggests that genuine novelty doesn’t just add something new, it restructures what’s possible. Grasses, for instance, radically altered the terrain, opening new evolutionary possibilities that didn’t exist before them. They propelled the evolution of grazing animals, for example, and may have spurred human ancestors to adopt an upright posture.
Erwin has written several popular books about transformative periods in evolutionary history, including the end-Permian mass extinction and the Cambrian Explosion—the former a catastrophic collapse, the latter an explosive burst of diversity. His new book addresses a more abstract idea: What makes something genuinely transformative across biology and human culture?
I spoke with Erwin about where novelty comes from, Jorge Luis Borges’ Library of Babel, and why he doesn’t think a single equation can describe novelty and innovation across disciplines.
Your main argument is that novelty and innovation are very different animals, but are often conflated. Why does this distinction matter?
In 2015, we had a meeting here at the Santa Fe Institute on novelty, and the first thing that the co-organizer said at 9 o’clock on a Wednesday morning was that we weren’t going to discuss definitions of novelty or innovation. The meeting descended into chaos because nobody knew what anybody was talking about. Everybody was using the terms interchangeably. It was a classic Tower of Babel thing over these two words.
One of the things I realized in working on the book was that there are a bunch of different ways people have used the terms novelty and innovation. For developmental biologists, novelty has come to mean, over the last 20 years, the origin of new “characters.” [The vertebrate jaw, the insect wing, the amniotic egg, are all “characters,” discrete, heritable morphological, or anatomical traits that can be defined, observed, and compared across species in a phylogenetic analysis.] But for paleontologists, novelty relates to how things get established, which is an ecological or environmental question, not a developmental question.
Right off the bat, developmental biologists and evolutionary biologists are looking at the question of novelty differently. Then I was reading Northwestern University economist Joel Mokyr’s work on technological innovation and the switch went off in my head. For decades, economic historians have been talking about the difference between invention and innovation—invention being something you can patent, and innovation being something you can make money from. The same thing happens in biology, I realized. The evolutionary biologists care about when something is successful, while the developmental biologists just care about how things form. And one of the things that we’ve discovered since the ’80s is that there’s this decoupling that often happens between the novel thing appearing and the novel thing becoming successful—that has a whole bunch of ramifications.
You tell the story of the Tullimonstrum, Tully’s common monster, a fantastical prehistoric creature that has a sluglike body, a long proboscis ending in two claws, fins to the rear, and an eye at each end of a strange set of protrusions extending from its middle. What does the Tullimonstrum tell us about the nature of novelty?
It’s certainly a novel morphology, and the extent of that novelty is captured by the fact that nobody can agree on what it is. Every couple years another paper comes out on how to classify the Tullimonstrum. It’s kind of a Rorschach test for biologists, because people usually see what they know the best. It’s a mollusk if you study mollusks. It’s an annelid if you study annelids.
It also tells you about the extent to which you can get gaps in evolution, the sort of thing that Darwin didn’t like. There’s nothing close to the Tullimonstrum. We can’t link it easily to anything else. Now, eventually what’s gonna happen is that people will find new fossils that probably do provide a home in the Burgess Shale. Lots of the Burgess Shale organisms lacked homes for a long time until new discoveries made it obvious.
You write in the book that Ernst Mayr, a highly influential 20th-century evolutionary biologist, believed that changes in behavior are one driving force behind novelty, a kind of pre-adaptation to it. Where do these changes in behavior themselves originate?
Mayr had a view that any change in function would produce a change in form. He was following Darwin in this. The changes in behavior come from new opportunities in the environment. For Darwin’s finches, as climate changes, seed side size changes. The change in behavior is that the finches are trying to deal with different seeds.
There’s a whole group of evolutionary biologists who argue that behavioral shifts precede the morphological or developmental changes that we see. And that’s certainly a possibility. But in the fossil record, it’s very hard to identify in most cases.
One of the good cases in the fossil record is that elephants and some other grazing animals shifted from browsing to grazing grasses and needed higher crown teeth, but they made that behavioral shift before the teeth caught up with the new food that they were eating.
You ask in the book whether in the history of life, we’ve been trending toward things that are more adapted, more powerful and just better over time. What’s your answer to this question?
Until the next mass extinction comes along. I mean, brachiopods were doing really well 270 million years ago, and dinosaurs were like, you know, hitting those strides 70 million years ago.
Read more: “Why Darwin Admired the Humble Earthworm”
One key concept you introduce is potentiation, which are these small evolutionary events that open up “radical new spaces” and set the stage for a future novelty. Can you offer an example?
In most organisms, the region just upstream of a gene’s coding sequence on the DNA strand contains a promoter, and near that is something called a proximal enhancer. When the enhancer gets activated, it tells the promoter to start reading the gene. But in complex animals, the enhancers can be a long way away from the gene itself, tens of thousands of sequences away. These are called “distal enhancers,” and they allow this coding sequence to be used in lots of different places. Instead of multiplying the number of genes, which are long, you multiply the number of enhancers, which are short. So it’s energetically less costly. It’s more efficient.
All animals have distal enhancers, but sponges, jellyfish, flatworms, corals—the quote unquote lower animals—don’t use them. They’re there. They’ve had them for 600 million years, but they never really figured out what they were for until what are called bilaterian animals arose. Vertebrates, mollusks, and everything that’s more complicated, use distal enhancers all the time. Whether you’re a fly or a mammal, it helps to have these distal enhancers. That’s an example of potentiation where something arises, but it doesn’t become a novelty until further down the line.
Can potentiation events be identified before the novelty they enable arises?
No. That’s one of the other reasons for distinguishing between novelty and innovation. If a novelty is a change in a character, such as wing shape, those can be identified without knowing whether the change is successful. The same is true of human invention. A lot of inventions in technology never make money. You got a patent for them but don’t make money until 10, 50, or 100 years later. A patent isn’t awarded because a new technology is successful. A patent is awarded because the technology is novel. That’s the mindset that I want people to begin to think about in biology as well.
In your new book, you describe writer Jorge Luis Borges’ Library of Babel. In this library, every possible book exists, including biographies of people who have never lived, scraps of Tolstoy’s War and Peace, books that consist of just the letter M, versions of your own book. But you write that there is a glitch in Borges’ library as a model of evolution. What is the glitch?
People often have used the Borges library as an analogy to Mendel’s library, the set of all possible DNA sequences, where most sequences are nonsensical and only a fraction represent functional living organisms. Selection has to have variation to work on, and Borges’ library has variation, but there’s no map. As Borges himself pointed out, a map of his library is the library. There’s no way of reducing it. There’s not a room that has all different copies of War and Peace differing slightly by what Prince Bolkonsky and Natasha do. There’s a copy of War and Peace next to a mystery by Barbara Pym, next to a play by Shakespeare. The glitch is that you can’t find anything. Natural selection doesn’t work in this scenario because natural selection can't navigate it. Evolution by natural selection needs spaces and topology.
You suggest that the dynamics underlying novelty and innovation apply equally to biological evolution and to human language, culture, and technology. Is there a single principle that unifies these different domains?
In the book, I offer different possibilities for what a theory of novelty and innovation might look like. At the one end, there’s no theory. It’s just natural. It’s just stories of what happens through the history of life. It’s natural history. At the other extreme, which some of my colleagues at the Santa Fe Institute would agree with, is the idea that there’s a mathematical formula that underlies the theory of novelty and innovation. And I don’t think either one of those extremes are correct. There’s a way of thinking about the problem, a conceptual framework of distinguishing between novelty and innovation and identifying potentiation that’s helpful in thinking about the problem without there being some universal formula.
I’m not a physicist. I’m a paleontologist. If I were a physicist, I’d think that there’s some equation for novelty and innovation. I have a number of friends who are physicists. I think they’re deluded, but I also think they’re wonderful people.
In the book, you write we may overestimate the contributions of technology to the scientific revolution, whereas contributions from intellectual tools and conceptual scaffolds are overlooked. Is this something you believe is true for other periods of great change? That we tend to attribute paradigm shifts to technology when concepts and ideas are the primary shapers?
This is sort of a Tullimonstrum problem, because I’m much more drawn to the history of ideas than I am to the history of technology. I’m not disputing that there have been incredible technological advances that have been useful across any science, but just as useful have been new ways of thinking about things. There’s a wonderful quote attributed to both a Hungarian biochemist and to German philosopher Arthur Schopenhauer. They both seem to have said something about how the hard part of science is seeing what everyone else is seeing, but thinking what no one else has thought. I found that quote early in graduate school and that approach has stayed with me. Darwin didn’t have any new technology.
But he certainly collected more data firsthand than many other people had before him.
He had lots of data, but so did Carl Linnaeus, and [Georges-Louis Leclerc, Comte de Buffon] probably had more data than Darwin did. Darwin had an incredible network of people he wrote letters to all over the world who gathered data about pigeons and everything else, because he had this insatiable curiosity. But it was what he was thinking about that made the difference.
Does your framework have anything to say about whether what’s happening with AI today constitutes genuine novelty or innovation?
Replacing coders and making money for Google and Meta because people are searching the web differently is one thing, but those aren’t truly novel contributions in the way some of the AI pundits seem to claim.
But aspects of AI seem genuinely novel. They seem to be creating innovation in the sense that they’re creating what I described in the book as these spaces that can’t be anticipated. So it’s not just that they’re doing something we could already do, but faster. They also seem to be doing things that couldn’t have been done before. That’s the genuinely novel aspect. But I don’t understand enough about it to know whether that’s really true.
If I walked away understanding just one thing from your book, what would you want it to be?
The simplest message is this distinction between novelty and innovation. But the deeper message that I hope some people will start grappling with is the topological issue of how both novelty and innovation in different ways create new spaces within which evolution can operate. 
Enjoying Nautilus? Subscribe to our free newsletter.
Lead image: Dee karen / Adobe Stock
