A Bit of Physics History: Ed Witten Introduces M-Theory

The Second Superstring Revolution was, like most revolutions, a somewhat messy affair, with a number of pivotal steps along the way: understanding the role of membranes in 11-dimensional supergravity, the discovery of dualities in supersymmetric gauge theories, Polchinski’s appreciation of D-branes as dynamical extended objects in string theory, and of course Maldacena’s formulation of the AdS/CFT correspondence. But perhaps the high point was Ed Witten’s formulation of M-Theory in 1995. And I just noticed that Witten sharing it with the world was captured on video.

Here is Witten’s paper:

String Theory Dynamics In Various Dimensions
Edward Witten

The strong coupling dynamics of string theories in dimension d≥4 are studied. It is argued, among other things, that eleven-dimensional supergravity arises as a low energy limit of the ten-dimensional Type IIA superstring, and that a recently conjectured duality between the heterotic string and Type IIA superstrings controls the strong coupling dynamics of the heterotic string in five, six, and seven dimensions and implies S duality for both heterotic and Type II strings.

Before this result, we knew about five different kinds of string theory, each living in ten dimensions: Type I, two different Type II’s, and two different “heterotic” theories. Then there was the most symmetric form of supergravity, living in 11 dimensions, which some people thought was interesting but others thought was a curiosity that had been superseded by string theory. To everyone’s amazement, Witten showed that all of these theories are simply different limiting cases of a single underlying structure. Nobody knows what that underlying theory really is (although there are a few different formulations that work in some contexts), but we know what to call it: M-theory.


Now Amanda Gefter, author of the new book Trespassing on Einstein’s Lawn (and a recent guest-blogger at Cocktail Party Physics), takes to Twitter to point out something I wasn’t aware of: a video record of Witten’s famous 1995 talk at USC. (I’m pretty sure this is the celebrated talk, but my confidence isn’t 100%.) [Update: folks who should know are actually saying it might be a seminar soon thereafter at Stony Brook. Witten himself admits that he’s not sure.] It’s clearly a recording by someone in the audience, but I don’t know who.

Most physics seminars are, shall we say, not all that historically exciting. But this one was recognized right away as something special. I was a postdoc at MIT at the time, and not in the audience myself, but I remember distinctly how the people who were there were buzzing about it when they returned home.

Nature giveth, and Nature taketh away. The 1995 discovery of M-theory made string theory seem more promising than ever, to the extent that just a single theory, rather than five or six. Then the 1998 discovery that the universe is accelerating made people take more seriously the idea that there might be more than one way to compactify those extra dimensions down to the four we observe — and once you have more than one, you sadly end up with a preposterously high number (the string theory landscape). So even if there is only one unifying theory of everything, there seem to be a bajillion phases it can be in, which creates an enormous difficulty in trying to relate M-theory to reality. But we won’t know unless we try, will we?

  1. I wonder if a hundred years from now this talk will – inadvertently and due to no fault of the brilliant Ed Witten – be seen as the beginning of a period of unrealistic, wild speculation and optimism in physics.

  2. Gotta love that old school blackboard talk (We’ll, the main talk used the overhead projector, but you know what I mean)

  3. I imagine invading space aliens wouldn’t have as much of a deep, guttural voice; but more of an Ed Witten voice. The tone I assigned to him while reading his quotes and bits & pieces of one of his books was very James Earl Jones. I was wrong…

  4. I personally don´t like string theory. These 7 or 8 curled up dimensions go way too much against Occam´s razor in my opinion. And despite that, since the 80´s the theory remains only speculative, making no experimental predictions. Physicists say it (the theory) is the only game in town, but I doubt it. Pehaps Loop Quantum Gravity, who knows ? Or perhaps there is NO such thing as a theory of everything.

  5. Luiz Claudio Weiss:
    Because the theory is incapable of producing any predictions, at least for now, some of it’s adherents have proposed retiring the essential scientific-method requirement of falsifiability. (See a few posts back on this blog. Amazing, if you can’t prove a hag is a witch, change the rules of witch-hunting and burn her anyway, because the show itself has such a great aesthetic value.)

    But watch out, Luiz. This place is for true believers. The sentiments you’ve expressed here might earn you a hailstorm of thumbs-down reserved for the infidels.

  6. What would it mean for there to be “no such thing as a theory of everything”? Do you think that the events we observe are arbitrary? Or are you saying that they are governed by laws, but those laws are somehow fundamentally incomprehensible to us?

  7. sam, “no such thing as a theory of everything” means that there may be different theories, each covering a particular branch of physics, but no unified theory that covers all of physics. Currently, for example, there’s a wonderful theory for the big things—from the universe as a whole to falling apples—and another wonderful one for the little things—from atoms to quarks. But these two theories are incompatible and there isn’t yet a theory that works well for all things. String theory is an attempt at such a unified theory. What Luiz says, and he echoes many thinkers, is that the existence of a unified theory of everything is by no means warranted. Myself, I think there must be one, but for reasons which are beyond science.

  8. RE: what if there is no “theory of everything”?

    Here is Feynman’s view on studying nature for its own sake, and being satisfied with the way it is: http://youtu.be/Kgivi57N8Tc?t=45m13s . To summarize, he said that it would be great if there is a ToE but if not, that’s fine.

  9. If the time it would take to find all the solutions of string theory was longer than the lifetime of the universe, I fail to see how it could be a worthy endeavor…

    If we really wanted to try to find the TOE in the theory, we would need to figure a way to find solutions much faster. I fear that string theory will never go away or see the light of day, and it will be just another possible theory for the rest of time. The founders of string theory will be immortalized due to this aspect of the theory alone.

  10. As far as the theory of everything is concerned, there can not be a theory of literally everything, since this would contradict Goedel’s theorems in logic. However, a quantum gravity theory can be constructed, which describes a limited but sufficiently large set of phenomena. String theory attracted a lot of attention because it was not just a theory of the quantized gravitation field, like loop quantum gravity, but it also unified gravity with other interactions. The problematic feature of string theory is that it does not have a unique solution for the types, masses and charges of the elementary particles. Most people, including me, do not like this, because they believe that a correct theory of Nature should have such a property. However, it is not easy to construct such a theory.

  11. Ed sure looks like his father. (That’s a compliment)

    Ed’s dad was a great teacher and really made physics accessible to generations of students at University of Cincinnati.

  12. I’m 100% sure this video comes from Stony Brook. Prof. Vladimir Korepin does the introduction, Prof. Peter van Nieuwenhuizen is seen asking a question at 42:28. These men work is vastly different areas and really their only similarity is being faculty at Stony Brook. Also, you can see CN Yang near PvN a few moments after.