Nathan Myhrvold – the polymath physicist whose passions range from cosmology to cooking – is this month publishing a massive, five-volume book about the science of bread and bread-making. Robert P Crease catches up with this intellectual livewire at his Cooking Lab headquarters in Seattle
Hands down, it is the most dramatic chandelier I’ve ever seen.
I am standing in the foyer of an ordinary-looking office building in the Seattle, US, suburb of Bellevue that’s home to Intellectual Ventures – a company that develops and licenses intellectual property. Suspended above me is a plain copper hoop about a metre in diameter near a small rod attached to a Tesla coil. Nathan Myhrvold – founder and chief executive of the firm – flicks a switch. Bluish-yellow streamer arcs, crackling and humming, burst out of the rod.
But I’m not here for this miniature lightning storm, spectacular though it is. Nor have I come to see any of the other exhibits of Myhrvold’s enthusiasm for science strewn around the building, which include an antique barometer, part of a Saturn launch vehicle and a replica of a Babbage machine. I’ve also paid only a cursory glance at the gleaming, three-storey Van de Graaff generator that stands guard outside. I’m not even here to discuss intellectual property.
Instead, I’ve come to ask Myhrvold about the physics of bread-making, on which he is the reigning authority. The 58-year-old physicist and former chief technology officer of Microsoft has long been interested in food science. In 2011 he published Modernist Cuisine, a six-volume, 2438-page monster of a book, weighing 23.7 kg that Physics World picked as one of the top 10 books of 2011 and the Wall Street Journal declared to be the “cookbook to end all cookbooks”.
This month, Myhrvold releases the sequel: Modernist Bread, a five-volume, 2642-page monster of a book (plus bonus recipe manual), weighing 24.2 kg. Self-published like its predecessor, Modernist Bread is based on about 1600 experiments that Myhrvold and his team have carried out with state-of-the-art equipment. It promises to be the ultimate book on the history and science of bread. But what, I’m wondering, remains to be learned about this common, simple yet sometimes delicious foodstuff that humans have been baking for more than 30,000 years?
Enter the Cooking Lab
Leaving the entrance lobby to Intellectual Ventures, I climb a flight of stairs to Myhrvold’s self-styled Cooking Lab. There I meet Francisco Migoya, a professional chef who has co-written Modernist Bread, and Larissa Zhou, the Cooking Lab’s research scientist, who received her BA in physics from Harvard University, US, in 2011. A large, open, well-lit space in the middle of the building, the Cooking Lab is fully stocked with big ovens, refrigerators and other professional kitchen gear (the lab could easily serve any restaurant or bakery) as well as more homely equipment such as blenders, cutters and cabinets. Next to the lab itself, around one stack of equipment, is a photography studio.
My eye is drawn, however, to an array of centrifuges, rotary evaporators, vacuum chambers and thermocouples of the type more commonly seen in physics labs. Migoya explains that he and others on the team use these devices to create new cooking techniques. “We found you can cook better French fries if you put them in an ultrasonic bath before frying them,” he says, pointing to a device of the sort used to clean jewellery. “The ultrasonic waves create little divets on the surface, exposing more area to the oil and making for a crispier French fry while preserving tenderness inside.” It also works for spaghetti, too, he adds, where the divets make creamy sauces cling better.
Migoya and Zhou launch into a description of the physics of ovens – and how restaurants can optimize bread-making – when suddenly I hear good-natured, boisterous laughter behind me. I turn to find Myhrvold standing next to a technician mixing flour and water. “Looks like you’ve just made a load of snot!” he says, with irreverent glee. Myhrvold, I discover, is rarely without a smile or laugh, and believes that research should be interdisciplinary and fun. Today he’s wearing a T-shirt that’s a Weather Channel send-up: “Deep Space 7 Day Forecast,” it runs, with seven boxes forecasting identical conditions – “Clear. Chance of precipitation 0%. Temperature –270 °C”.
We retire to a conference room, where Myhrvold – whose short, ruddy (albeit whitening) hair juts from his head as if he’s brushed against that Van de Graaff accelerator – does not seem easily able to sit still. Energetic and restless, he seizes every opportunity to get up to show me something, or to draw diagrams on the blackboard. Following him is like trying to keep up with a wave that keeps darting off in new directions. But his curiosity is grounded in a solid scientific pedigree.
Filling the vacuum
After getting a BA in mathematics in 1979 from the University of California, Los Angeles, aged just 20, Myhrvold did a PhD at Princeton University, US, in theoretical and mathematical physics. He then spent a year as a postdoc working with Stephen Hawking at the University of Cambridge in the UK on quantum field theory and quantum gravity, before founding his own software company in 1984. It was bought by Microsoft two years later and Myhrvold spent 14 years at the firm, ending with a four-year stint as Bill Gates’ chief technology officer.
When Myhrvold founded Intellectual Ventures in 2000, he began to spend more time cooking as a hobby (though he’d already taken leave at Microsoft to go to culinary school), and became fascinated by “sous vide” cooking, in which you seal fish or meat in a bag under vacuum and then cook it in a water bath. As the temperature of the bath can be precisely controlled, sous vide (the French for “under vacuum”) lets you cook fish or meat far more evenly than using a hot stone or oven. So rather than having the inside of the food done while overcooking the outside, sous vide preserves the same “done-ness” throughout.
It sounds a great idea, but back then sous vide was a controversial technique, suspected of being unsafe. “I assumed that there were some studies of it,” Myhrvold explains. “Some data that would set it all out for me. There was nothing!” So in his spare time, he began to generate the data himself. As he produced that data, and explored scientific dimensions of cooking, Myhrvold’s curiosity morphed into a passion. His interest coincided with the rise of “molecular gastronomy” – a term coined in the 1980s by the University of Oxford physicist Nicholas Kurti to describe the use of scientific knowledge and techniques to explore new culinary opportunities.
That movement spawned research centres and new restaurants – the most famous being elBulli in Catalonia, Spain – and Myhrvold found his culinary research deeply satisfying both personally and intellectually. “Cooking is the only science experimentation that most people do regularly,” he says. “It’s got variables – some you can control and some you can’t. It doesn’t always turn out, and you can learn to do it better from both your failures and successes. Chemistry describes much of the detail, but physics is a key layer.”
After about a decade researching the scientific basis of cooking, during which he says he spent several millions of dollars of his own money, Myhrvold distilled his findings into Modernist Cuisine, which he co-wrote with chefs Chris Young and Maxime Bilet. The six volumes contained a huge assembly of photographs and information about food and food preparation – from how to wash your hands to a biography of the French mathematician and physicist Joseph Fourier, who discovered the partial differential equation for the conductive diffusion of heat. The book also busts a few food myths, such as insisting that certain red wines taste better if you add salt and that you can aerate wine by whizzing it in a food blender (two sacrilegious notions that had been ridiculed by wine snobs).
Critical reaction to Modernist Cuisine was awed, if occasionally mixed. Reviewing the book for the New York Times, the US food writer Michael Ruhlman called it “mind-crushingly boring, eye-bulgingly riveting, edifying, infuriating, frustrating, fascinating, all in the same moment.” In the end, though, Ruhlman thanked Myhrvold for allowing chefs to ditch “molecular gastronomy” for the more pleasant term “modernist cuisine”, and concluded with a “bow to him and his crew for their work of unprecedented scope and ambition”.
Myhrvold tells me he has no apologies for all the physics in the book. “One interviewer asked me what made me think I could put science in the kitchen. I said, ‘Science is always there! I only took the ignorance out!’”
So what is the ignorance that Modernist Bread is now taking out of baking?
“A lot, it turns out,” he says. “Just because a cooking practice is old doesn’t mean it’s good. In the 1970s, there was an artisanal bread movement that advocated returning to the supposedly good bread-making of the past. Nonsense! The best bread is being baked now!”
Myhrvold leads me to a nearby room that includes every book on bread that his staff could track down – more than 300 ancient and modern volumes. It also has several historical artefacts, such as a stamp used in ancient Rome to identify the provenance of loaves of bread. But when Zhou put the books into a database to discover what similarities there were in bread recipes, it turned out there were surprisingly few.
“Many ancient documents, including the Bible, mention bread,” Myhrvold says. “Was that bread the same as today? Did it even look the same? How can you find out? There were no cameras!” He decided to spend time investigating old paintings, but found that art historians tended not to include foodstuffs in their catalogues of paintings. But Last Supper paintings always had bread in them, and Myhrvold spent a day examining several at the Louvre museum in Paris. “I noticed a funny cultural thing – artists would paint the bread of their time on Jesus’s table. There’s a German painting of the 15th century showing Christ and the apostles eating pretzels! How could God not have had pretzels!”
There’s physics scattered through Modernist Bread, including details of how heat flows through food and of how baking “works” thanks to the presence of tiny bubbles of carbon-dioxide gas (figure 1).
“A Harvard professor once asked me, ‘If I’ve got a roast about this size’ ” – Myhrvold holds up his hands about half a metre apart – “ ‘it takes me a few hours to cook. If I’m cooking the same amount of bread, it’s done in 20–30 minutes. Why?’ It turns out it has to do with heat transport.” As Myhrvold explains, while heat flows into a roast by conduction, in bread it flows more because the bursting bubbles open “heat pipes” through which heat can move by convection. The two have different scaling laws. Indeed, the physics of heat pipes shapes other natural phenomena, including various geological processes such as volcanoes (figure 2).
Bubble size is also crucial in bread-making: in a baguette, for example, you want fewer, larger bubbles. However, surface tension increases nonlinearly with bubble size, meaning it takes far more energy to expand a tiny bubble by a given volume than it does to inflate a large bubble by the same size. “That interests me as a cosmologist, for the same topic crops up in space–time foam research, or in multiverse theories,” says Myhrvold. “Infinitesimal bubbles with infinite curvature are impossible to make!” That’s one of the beauties of Modernist Bread: readers will learn that physics is not only implicated in cooking, but also has a broad interdisciplinary impact on the wider world.
Knead to know
Cases of myth busting, which Myhrvold clearly enjoys, are scattered through this book too. Take kneading. “When I tell people that kneading bread is not required,” he says, “they are horrified. ‘Mom always did it!’ But it’s pretty much a fraud. Kneading has some effect, but it doesn’t do what people thought – mix it [the flour and water] thoroughly somehow. The mechanical reaction does heat up the flour, which is good for mixing with water. But a vacuum sealer [which lets the flour and water react chemically in a vacuum-sealed package] is superior.”
Bread-baking, it seems, is a simple practice that takes advantage of deep physics principles that also appear elsewhere. “Like why is the sky blue,” says Myhrvold. “I was even asked that question at the end of my orals at Princeton, by Val Fitch I think. Do you know it’s for the same reason that non-fat milk is blue? Do this experiment at home: take a glass of water, put in a spoonful of non-fat milk and you’ll see it’s blue-ish. It’s for the same reason.”
“Exactly! Particle size. Air molecules are small enough that you get blue scattering. The colour of the sky tells you the particle size. If you try this with full-fat milk, the fat globules are bigger, and it goes from Rayleigh scattering to Mie scattering. [Gustav] Mie was the German physicist who figured this out. Mie scattering is also the reason why bread is white! Here’s a case where a piece of equipment totally changes your view of something. If you look at bread in a microscope, it’s a clear gel; it looks like shower glass.”
I must look sceptical, for Myhrvold suddenly orders his assistants to bring bread – any kind they can find at a moment’s notice – and they return with four hot-dog buns. He leads me to his private office down the hall. “You see I’m a micromanager!” he says, laughing. For a moment I am unsure what he’s referring to: his ability to get food delivered instantaneously or – most probably – meaning the row of half a dozen microscopes along one wall. Then Myhrvold puts a piece of one under a microscope. Sure enough, it resembles a pile of shattered shower glass.
I ask Myhrvold how his Cooking Lab differs from the long-running US TV show America’s Test Kitchen, in which staff test different ways of preparing a recipe and explain the science to the audience. “I like America’s Test Kitchen,” Myhrvold says. “But we start where they leave off. They’ll try a dozen or so recipes to find the best and most reliable one for people to do at home. We use advanced scientific equipment to try to understand why something happens.”
But if your research requires advanced scientific equipment, I ask, how could it interest the average working person?
“You are not giving the average working person enough credit,” Myhrvold says, in the only flicker of annoyance I see in our hour-long chat. “Don’t dumb people down! People are interested in the why. While I was at Microsoft, Steve Hawking published A Brief History of Time and Madonna published Sex. Who sold more copies? It was not even close – Hawking outsold Madonna more than 10 to one. You don’t have to be a high-end chef to be fascinated by food science.”
The way someone who is not an astronomer might be fascinated by astronomy?
“Right. Plus, the Cooking Lab’s food science also makes it possible for people to do high end and exotic kinds of cooking if they want. Plenty of books have recipes that say, ‘Do this, this, and this.’ If you like doing such things without understanding why or how, buy those books! The unofficial tag line is that our books are for people who are passionate and curious about food. If you are not passionate, a 2640-page cookbook is not for you!”
Myhrvold draws an analogy with engineering, where insight lets engineers build not just mundane safe stuff, but new and extraordinary structures as well. “Most bridges in the world are pretty boring – exercises in building trusses – but deep understanding lets people like [Spanish architect and engineer Santiago] Calatrava make them incredibly beautiful.”
I ask Myhrvold for a simple example of how the knowledge of cooking he has developed might help ordinary home cooking.
“If you have a steak that is twice as thick as the one you cooked the last time,” he asks me, “how much longer is it going to take to cook?”
I say I don’t know exactly. Somewhat longer.
“Most chefs can’t even tell you exactly,” he says, “because even though it’s a really basic question nobody taught them. The answer is four times. Heating in a steak works by conduction, and conduction has a scaling law that goes by the square of the depth.”
So is there then no intuition or fingertip knowledge to cooking?
“Sure there is! A Japanese chef cuts fish more quickly and deftly than I can. But if you talk to the guy at the local steak house, he may have an intuitive sense of how long it takes to cook a steak, but it’s from long experience.”
What’s wrong with that?
“Three things,” Myhrvold says. “First, learning from experience means that you’ve screwed up a lot. That guy has ruined a lot of steaks! Second, learning from experience doesn’t help teaching people. Why not speed things up by telling learners the principles? Third, sometimes the right way of doing something is counterintuitive, as it was with sous vide, and you’ll probably never find it from experience. Active research can uncover new things.” Myhrvold then reminds me of several episodes involving nautical engineering where boats built by extending the tried-and-true, such as the 17th-century Swedish warship Vasa, flipped and sank because of ignorance of principles related to the centre of buoyancy.
Bagels and beyond
So the conversation goes, ranging over many subjects covered in Modernist Bread – the difference between rye and wheat bread, for instance, or why a croissant’s crust is flakey and a bagel’s is not. We talk about heat effects, and why you can reach your arm safely into a dry oven at 250 °C but not over a pot of boiling water. We talk about instruments, and how Myhrvold buys expensive ovens only to cut them apart for pictures.
Before Myhrvold gets whisked off to another commitment, I ask him how food physics differs from other branches of the subject. “Food physics isn’t like particle physics,” Myhrvold admits. As he reminds me, if you want to discover the Higgs, you have to build a big accelerator, with the technology driving the discovery and what you do. “But most science is not like that. Most science is driven by interests and focus, not by the tool. That’s true of the physics in this lab. The research takes us where it takes us.”
As I leave the Cooking Lab and begin my journey back to New York, I realize I’ve got one thing badly wrong. I’d been expecting to learn what physics can teach us about bread and baking. But as Myhrvold has made it abundantly clear, plenty of knowledge flows in the other direction too.