At a 1990 conference on cosmology, I asked attendees, who included folks like Stephen Hawking, Michael Turner, James Peebles, Alan Guth and Andrei Linde, to nominate the smartest living physicist. Edward Witten got the most votes (with Steven Weinberg the runner-up). Some considered Witten to be in the same league as Einstein and Newton. Witten was and is famous for his work on string theory, which unifies quantum mechanics and relativity and holds that all of nature's forces—including gravity--stem from infinitesimal particles wriggling in a hyperspace consisting of many extra dimensions.

Witten: "I hope the landscape interpretation of the universe would turn out to be wrong, as I would like to be able to eventually calculate from first principles the ratio of the masses of the electron and muon (among other things). However, the universe wasn't made for our convenience."

Even then, string theory—which some enthusiasts (not including Witten) called a "theory of everything"--was extremely controversial, because there seemed to be no way to confirm experimentally the existence of strings or the extra dimensions they supposedly inhabit. I was thus thrilled a year later when Witten agreed, albeit reluctantly, to an interview at the Institute for Advanced Study in Princeton. After our encounter, I wrote a short profile of Witten, "The Pied Piper of Superstrings," published in Scientific American in 1991; and a longer profile for my 1996 book The End of Science. Here is an excerpt from the latter:

I asked Witten how he responded to the claims of critics that superstring theory is not testable and therefore is not really physics at all. Witten replied that the theory had predicted gravity. "Even though it is, properly speaking, a post-prediction, in the sense that the experiment was made before the theory, the fact that gravity is a consequence of string theory, to me, is one of the greatest theoretical insights ever."

He acknowledged, even emphasized, that no one has truly fathomed the theory, and that it might be decades before it yielded a precise description of nature. He would not predict, as others had, that string theory might bring about the end of physics. Nevertheless, he was serenely confident that it would eventually yield a profound new understanding of reality. "Good wrong ideas are extremely scarce," he said, "and good wrong ideas that even remotely rival the majesty of string theory have never been seen." When I continued to press Witten on the issue of testability, he grew exasperated. "I don't think I've succeeded in conveying to you its wonder, its incredible consistency, remarkable elegance and beauty." In other words, superstring theory is too beautiful to be wrong.

Recently, Witten won the prestigious Kyoto Prize, created by Japan's Inamori Foundation to honor "those who have contributed significantly to the scientific, cultural and spiritual betterment of mankind." Witten won 50 million yen, or about $500,000, in the category of basic sciences. The Kyoto Prize website states: "Dr. Witten has made significant contributions to theoretical physics for more than 30 years as a leader in the dramatic evolution of superstring theory. Moreover, by applying his physical intuition and mathematical skills, he has advanced mathematics, and prompted the cutting−edge research of many mathematicians."

My profiles of Witten were pretty snarky. So when a publicist for the Kyoto Prize asked recently if I would like to interview Witten, I was pleasantly surprised. I submitted some questions, and Witten graciously answered them by email.

Horgan: When I interviewed you in 1991, you said that "good wrong ideas that even remotely rival the majesty of string theory have never been seen." Are you still confident that string theory (or its descendant, M theory) will turn out to be "right"?

Witten: I think I will stick with what I said in 1991. Since then, we have lived through the second superstring evolution and many surprising developments in which string theory has been used to get a better understanding of conventional theories in physics (and math). All this makes most sense if one assumes that what we are doing is on the right track. (My 2005 article ``Unravelling string theory,'' gives somewhat more detail.)

Horgan: Do you see any other rivals for a unified theory of physics?

Witten: There are not any interesting competing suggestions. One reason, as remarked in "Unravelling," is that interesting competing ideas (twistor theory, noncommutative geometry, ...) tend to be absorbed as part of a larger picture in string theory. The competing interesting ideas have been very fragmentary and have tended to gain power when absorbed in string theory.

Horgan: There are an estimated 10500 versions of string theory, each of which "predicts" a different universe. Leonard Susskind and others propose that these hypothetical universes actually exist, forming a multiverse "landscape." But physicists such as Peter Woit, Paul Steinhardt and George Ellis complain that string and multiverse theories may be untestable and hence not truly scientific. How do you respond to these criticisms?

Witten: Personally, I hope the landscape interpretation of the universe would turn out to be wrong, as I would like to be able to eventually calculate from first principles the ratio of the masses of the electron and muon (among other things). However, the universe wasn't made for our convenience. Plenty of leading physicists -- prominent examples being Steve Weinberg and Martin Rees -- have taken the acceleration of the cosmic expansion seriously as a hint that a landscape interpretation of the universe may be correct. This was a monumental discovery, which was announced in 1998 (two years after publication of your book The End of Science -- unfortunate timing!). It was so shocking that it was several years, and a lot more data, before I personally was convinced of that finding. [Horgan: See my take on cosmic acceleration in Addendum below.]

If the landscape interpretation is correct, can we get additional clues that would make this more believable? One obvious possibility involves the outcome of Large Hadron Collider experiments. I explain why in my 2004 article, "When symmetry breaks down." What I wrote there wasn't original but provides a succinct explanation. The literature is filled with other suggestions about how we might conceivably get more clues about a landscape interpretation if that is correct (for example seeing a signature of a prior phase transition in the cosmic microwave radiation). It is hard to summarize these suggestions for you as it is hard to know which proposals are most worth describing.

Another possibility is that a theory that predicts a landscape would become well-established because of other predictions it made. The trouble with criticizing string theory because it plausibly predicts a landscape of vacua is that the landscape interpretation of the universe might be correct. 200 years from now, if more clues have emerged, possibly including some that are unforeseeable now, it might seem obvious that the landscape of string vacua was necessary to make string theory viable.

Horgan: Do you agree with Sean Carroll that falsifiability is overrated as a criterion for distinguishing science from pseudo-science?

Witten: Scientists aim to get as reliable and precise an understanding of nature as we can. The gold standard is a precise prediction that can be tested in a precise way in a laboratory experiment. Experiments that disprove theories are an important part of the scientific process.

With that said, it is a little too narrow to claim that science consists of trying to falsify theories because a lot of science consists of trying to discover things. (Chemists who attempt a new synthesis could say they are trying to falsify the hypothesis that this new synthesis won't work. But that isn't what they usually say. People who search for life on Mars could say they are trying to falsify the hypothesis that there is no life on Mars. Again, people don't usually talk that way.)

Horgan: Are multiverse theories—whether inspired by string theory or other ideas—falsifiable? If not, should they be taken seriously?

Witten: If the landscape interpretation of the universe is correct, it certainly might be possible to get additional clues supporting this, as I indicated in an answer above.

Horgan: Some proponents of multiverse theories have promoted the anthropic principle as an answer to the question of why we live in this universe and not an entirely different one. Do you think the anthropic principle is a legitimate scientific proposition?

Witten: As I also wrote above, I would prefer an explanation of the universe from first principles, with no anthropic considerations. But once again, the universe wasn't created for our convenience and it is useless to have preconceptions about what the answer is.

Horgan: Do you think science will ever explain exactly how the universe came into being? Why there is something rather than nothing?

Witten: Science has had an amazing amount of success in understanding the early universe, and for example, as you probably know, the evidence for cosmic "inflation" is vastly greater than when we last talked (or when your book appeared). However, when you ask if we will understand "exactly" how the universe came into being, that is a very high bar. [Horgan: I don't accept that the evidence for inflation is "vastly greater" now than in 1996. See my post on inflation under Further Reading. Then see Clarification from Witten.]

I think I would prefer to just say what I can see that might give a better understanding. Obviously, there are forthcoming observations that may help. The most immediate will be more data on the polarization of the cosmic microwave radiation. We should know more there within a few years. On a time scale of a few decades, fundamentally new insight might come from observations of the cosmic 21-centimeter background radiation and probably from other techniques.

However, as a theorist, I would like to also comment on theoretical loose ends that might offer hope of further progress. One involves cosmic inflation. As I've remarked, there is by now quite a lot of evidence for cosmic inflation. But there also isn't a completely consistent, logically sound mathematical framework for the theory of cosmic inflation. If inflation is correct, then I think there is a good hope this problem might eventually be solved and that might lead to a better understanding of the early universe. Likewise, since we last talked, there has been tremendous progress in understanding the behavior of string theory when quantum effects are strong. But there has been very little progress in understanding the behavior when time-dependent effects are large (which would be important in the very early universe).

This is a relatively well-posed mathematical problem and I imagine there is a good chance for eventually getting insight, which again might lead to a better understanding about the early universe.

Horgan: Are there any mysteries that you think science will never explain?

Witten: As in any endeavor, all we can do is to do our best, without knowing at the beginning of the journey how far we will be able to get. The modern scientific endeavor has been going on for hundreds of years by now, and we've gotten way farther than our predecessors probably imagined.

Horgan: Some physicists, such as Stephen Hawking and Lawrence Krauss, have denigrated philosophy as obsolete. Do you agree? Are there any philosophers, living or dead, whose work you find interesting or useful to your work?

Witten: Personally I am most influenced by some of the "natural philosophers" of the past, such as James Clerk Maxwell.

Horgan: Are you religious? Do you think science and religion are compatible?

Witten: I consider scientific explanations to be more interesting and illuminating.

Addendum: Like Witten, I was shocked by—and initially skeptical of--the discovery in the late 1990s that the universe is expanding at a growing rate. Now, I view it as "by far the most exciting advance in physics and cosmology in the last decade," as I put it in 2006. If I wanted to criticize The End of Science, I'd lead with cosmic acceleration. But so far the discovery remains just an interesting twist in the big bang theory, which has not precipitated the kinds of revolutionary theoretical breakthroughs and paradigm shifts that took place in the early 20th century.

Clarification from Witten: I was not alluding to the BICEP findings when I said that the evidence for inflation today is vastly greater than it was when your book appeared. I was referring to the totality of data on cosmic microwave fluctuations which emerged in the decade after your book was published. The evidence for something like inflation in the early universe is indeed vastly stronger. Virtually any physicist or astronomer will tell you that, although there are exceptions. The only implicit allusion to BICEP in my remarks was actually when I wrote (later in the interview, in answer to the question about whether we would ever have a complete knowledge about the origin of the universe) that within a few years we would know more about the polarization of the CMB. What I meant (and I think my phrasing, which reflects questions that have been raised about BICEP, would be clear to colleagues) was that within a few years we will know whether the effect claimed by BICEP exists at a level that can be detected in the CMB.

Further Reading:

"Why I Still Doubt Inflation, in Spite of Gravitational Wave Findings." The BICEP2 results, which I discuss in this post and to which Witten alludes, are looking shakier and shakier.

"Hey Physics, Get Real!"

"If You Want More Higgs Hype, Don't Read This Column."

See also my recent Q&As with physicists George Ellis and Carlo Rovelli, who both comment on strings, multiverses, falsification and other physics-related topics.

Photo: Jewish Currents,