Hold Fire! This Epic Vessel Has Only Just Set Sail…

The August 25th issue of the Times Higher Education Supplement has a feature article by Leonard Susskind about my book and Lee Smolin’s entitled “Hold Fire! This Epic Vessel Has Only Just Set Sail…” (unfortunately only available to subscribers). The bulk of it consists of two parts: an extended analogy designed to show what he thinks the current state of string theory is, and a long ad hominem argument about why people shouldn’t listen to me and Smolin.

In Susskind’s analogy, the current state of particle physics is like the 15th century European view of the world, aware that there was a large Atlantic ocean out there, but with no idea of what lay beyond it. String theorists are like ship-builders, building vessels that intrepid string theorist explorers will courageously pilot out into the risky unknown. I’m a “Chicken Little” figure, telling people that if they do this they’ll fall off the end of the earth. Smolin is a builder of ships that don’t float.

Susskind mostly ignores the contents of my book and Smolin’s, which, in his analogy, both provide detailed analyses of the history and current state of a shipbuilding project, which, despite massive investment, has led only to a huge, overweight vessel which can’t even get out of the harbor. Both of us are arguing that this project needs to be restructured and largely abandoned, and investigation of other ship designs supported and encouraged.

The part of Susskind’s long ad hominem argument that attacks Smolin is just stupid, vicious, and offensive and I won’t repeat it here. Given how limited the successes of string theory have been, his attacks on Smolin’s work as ideas that are not working out is completely indefensible.

Susskind devotes a surprisingly long part of the article to discussing me and my career, and I have to admit that what he has to say is, while less than completely accurate, far more sympathetic than I would ever have suspected, especially given the many harsh things I’ve had to say about him here and elsewhere. He describes me as “one of those Princeton mavericks, who had the guts to work on other questions, in particular modern nuclear physics [by which he means QCD]“, and criticizes (during the mid-eighties) “an unusual degree of hubris in Princeton, a smug, arrogant dismissal of any ideas that didn’t fit the string theory agenda.”

Susskind’s interpretation of my early career is sympathetic, but a bit off. I actually left Princeton in the summer of 1984, just before the string theory “revolution” hit. I spent the early years of the era of string theory dominance at Stony Brook, with limited contact with what was going on in Princeton. Susskind doesn’t quite directly say so, but he strongly implies that my criticisms of string theory are motivated by bitterness at not being able to have a successful career in a physics department due to the domination of string theory. What actually happened is that in 1987, after my postdoc at Stony Brook, I did find myself unemployed, and at the time wasn’t too happy that string theory dominance was one of several reasons no one was much interested in hiring me. I spent a year as an unpaid visitor in the Harvard physics department and got a part-time job teaching calculus as an adjunct instructor at Tufts. During this year I had plenty to live on, but did face an uncertain future and wasn’t so happy about it. People at Harvard and at Tufts were quite helpful, and in the spring I was offered an excellent job for the next year at MSRI, the math institute in Berkeley. After that I came to Columbia, and from my time at MSRI on, I have no complaints whatsoever about my career, feeling I’ve probably done better than I deserved, living in the places I most want to live, working with excellent colleagues under good conditions. So, as far as the embittered part of my career goes, it was pretty much limited to a short period of about a year, almost 20 years ago, during 1987 and 1988.

Susskind ends his discussion of me with something positive:

But Woit is correct to remind us how important diversity and humility are in the face of the vast sea of ignorance.

and ends his review by quoting ‘t Hooft as a sceptical critic of string theory, finishing with:

This leads ‘t Hooft to another important point: diversity of viewpoints is to be cherished, not suppressed. This is something that Woit and Smolin have properly reminded us of, and string theorists should not be allowed to forget it.

So, all in all I’ve quite mixed feelings about this piece. Susskind’s attack on Smolin is highly reprehensible, and the way it ignores discussion of real issues, concentrating on dubious analogies and ad hominem argument, is disappointing. But, I have to admit that in his more than charitable discussion of one of his fiercest critics he shows a capability for gentlemanly behavior I wouldn’t have suspected (and wish he had shown Smolin), and, in the end he recognizes and admits that Smolin and I are making an important point that string theorists need to take note of.

Update: Several people have pointed out that the same issue of the Times Higher Education Supplement also includes a quite positive review of Not Even Wrong by Philip Anderson. On the whole it’s accurate, although I think Anderson neglects to mention that, lacking experimental results, I’m much more of a believer in the possibility of using mathematics to make progress in particle theory than he is. There are quotes from and discussion of the review at a new blog here.

Update: The THES in a later issue has a letter about Susskind’s article, which correctly points out that answering criticisms of string theory by claiming they come from a “mid-level theoretical physicist” or a member of the “Chicken Little Society”, didn’t address the fact that in the same issue these criticisms were coming from an extremely distinguished theorist and Nobel Laureate (Anderson). The letter writer’s reaction to Susskind’s article was:

Moreover, Susskind’s defence of string theory not only failed to address Anderson’s key criticism of string theorists – namely that their theorising is not grounded “on the acute observation of nature” – but rather reinforced this impression.

This entry was posted in Not Even Wrong: The Book. Bookmark the permalink.

73 Responses to Hold Fire! This Epic Vessel Has Only Just Set Sail…

  1. A says:

    adding to “amused” above:

    Philosophy, after trying many epic vessels, remained at the starting point.

    Physics, thanks to near coast navigation, progressed even too far.

  2. Anon says:

    “The history of the no-go theorems you’re talking about goes back to the mid-eighties. I’m not going to waste my time getting together references … ”

    We’re all capable of using SPIRES to find the exact references. How about some names and approximate dates to get us all started?

  3. Chris W. says:


    Your characterization of the situation reminds me of what is now happening to science within NASA; it is being sacrificed in favor of massive engineering boondoggles—a manned return to the moon (and establishment of a base there) and a manned expedition to Mars. These projects will probably collapse into a mess of cost overruns and missed milestones before they attain their stated objectives, but the aerospace and defense contractors involved (and the federal officials and members of Congress promoting their interests) will certainly benefit in the meantime.

    [By the way, you nearly ruined a great comment by contaminating it with the pseudo-word "irregardless".]

  4. UCSB says:

    Hi Anon,

    Let me help. The papers did indeed exist, but the precise issue is when `string theorists were going on about them’, to use PW’s words. They existed but in fact were so unknown that Maldacena and Nunez were not aware of them (and did not cite them) when they rediscovered and generalized their results. One paper is

    B. de Wit, D.J. Smit (Utrecht U.) , N.D. Hari Dass (NIKHEF, Amsterdam) . NIKHEF-H-86-15, May 1986. 44pp.
    Published in Nucl.Phys.B283:165,1987

    This was cited a grand total of 6 times in the ten years before the dark energy was found, then 4 more in the next two years, and then 78 times when it was rediscovered after the MN paper (which itself was cited 289 times). So this clearly shows at what point string theorists were `going on about this’. By the way, MN themselves discuss the limitations on their assumptions, and cite a list of aready-existing papers that lie outside these assumptions and are not constrained by the theorem (beginning with Strominger in 1986, long before the dark energy was found). So there was no point in time at which those actually working on the construction of superstring vacua regarded these theorems as a constraint.

    The other early paper was

    Aspects Of Supergravity Theories.
    G.W. Gibbons (Cambridge U.) . Print-85-0061 (CAMBRIDGE), Jun 1984. 53pp.
    Three lectures given at GIFT Seminar on Theoretical Physics, San Feliu de Guixols, Spain, Jun 4-11, 1984.

    which was even less cited and less known (it appeared only in a conference proceedings, before the days of the arxiv).

  5. woit says:

    I’ve wasted my time and found the precise reference I was thinking of when I wrote the lines that so upset AC from UCSB, they’re from a paper but also correspond to what I remember hearing in a talk around this time

    “This means that there is no classical way to get de Sitter space from string theory or M-theory… In fact, classical or not, I don’t know any clear-cut way to get de Sitter space from string theory or M-theory. This last statement is not very surprising given the classical no go theorem. For, in view of the usual problems in stabilizing moduli, it is hard to get de Sitter space in a reliable fashion at the quantum level given that it does not arise classically”.

    Funny that we’re told that the no go theorems were “well-known to be irrelevant”, I guess this particular string theorist wasn’t very well informed. It’s true that he didn’t claim that string theory “predicted” a non-positive CC, he’s not the sort to go on about bogus string theory “predictions”. And, by the time of this quote, the experimental evidence was in favor of a positive CC.

    People who are very concerned with the issue of whether I’ve got this right still seem to insist on ignoring the main point: is Susskind’s “prediction” a bad joke or not?

  6. Anon says:

    “I’ve wasted my time and found the precise reference …”

    And the precise reference is …?

    “Funny that we’re told that the no go theorems were ‘well-known to be irrelevant’”

    They weren’t well-known at the time.

    And by the time they became well-known, they were already understood to be irrelevant (in the sense that there were already well-known string backgrounds that violated the assumptions of the theorems).

    UCSB has the history right.

  7. Anonymous Coward from UCSB says:

    E. Witten, hep-th/0106109, 3 years _after_ the supernova data.
    So you did distort the chronology: you stated that string theorists made a wrong prediction and then backtracked when the data came out the other way, whereas this was an honest statement of a research puzzle. And Witten was indeed out of the loop on this subject, since the first dS solutions (E. Silverstein hep-th/0106209) appeared at virtually the same time.

  8. woit says:

    Anon and AC from UCSB,

    I think the Witten quotation speaks for itself, your statement that the no-go theorem was “well-known to be irrelevant” is simply wrong. Witten was giving talks at places like Strings 2001 saying what I quoted. He did not consider these theorems “irrelevant”, and if there’s one thing Witten rarely is, it’s “out of the loop”. Yes, it’s true that the way I stated things exaggerated the true situation. As I said, it was somewhat of a joke made in the context of responding to Susskind’s absurdity.

    For the N’th time, will you or won’t you address the issue of whether or not Susskind’s argument is an absurdly bad joke? I understand very well why you would rather argue about something else, but sooner or later people in this field will have to face up to what is going on. You can devote your days (and nights) to anonymously attacking anyone who tries to point this out, but it won’t change what is happening.

  9. amused says:


    Although I can well understand the temptation I think you’re being too harsh comparing string theorists to sleazy aerospace/defense contractors and congressmen. As individuals they aren’t really doing anything different from other physicists or academics in general by promoting their own area and supporting the people in it to the exclusion of others. But in this singular situation of string dominance the leaders of that field need to realise that with great power comes great responsibility. I.e. the responsibility to act in the best interests of physics. I wonder if they ever give any thought to this – it’s certainly not something they like to talk about. Do they really think it’s in the best interests of physics that the string program is pursued to the exclusion of everything else in formal hep theory?
    The current protestations from string critics are nothing compared to the merciless slamming that awaits them from future historians of science if they don’t exercise their responsibility wisely.

    (And as for the pseudo-word contamination, it was a quite mild contamination level by my usual standards ;))

  10. Anon says:

    “Yes, it’s true that the way I stated things exaggerated the true situation.”

    You wanted to know why none of the string theorist bother to respond to you. 

    That’s why. 

    UCSB chose one recent example, but we could do the same with just about everything you say on this blog.

    If you want them to take you seriously, then you’ll have to take them seriously, too, and not engage in this kind of deliberate distortions and obfuscations.

  11. woit says:

    Yet another anonymous string theorist,

    “we could do the same with just about everything you say on this blog.”

    Whoever you are, hiding behind anonymity, you’re completely dishonest.

    Look, what is going on here is that, in a throwaway line in the comment section here, in the context of discussing the completely insane state that string theory has come to, I made a reasonable analogy to another similar issue in string theory. I wrote two sentences, both of which are completely accurate, but which could be read together as making a misleading accusation. I shouldn’t have done this. I’ve written probably a thousand pages of stuff here. Every bit of it is signed by me, and if I’m wrong about something I will admit it.

    In response to this, I’ve gotten mostly dishonest responses from what appear to be three separate string theorists, none of whom dare put their name to what they write. AC from UCSB’s claim that what I wrote was untrue because these no-go theorems were “well-known to be irrelevant” was shown to be not true by the quote from Witten.

    Sure, sometimes in a comment section here I say something that is not quite right, sometimes for rhetorical effect. Anyone who wants to is welcome to correct me, but doing so in an insulting and dishonest way, hiding behind anonymity, is just kind of pathetic.

    For the N+1th time, will any one of the three of you actually address the question at issue: is Susskind’s “argument” a bad joke, deserving of being made fun of, or not?

  12. UCSB says:

    `is Susskind’s “argument” a bad joke, deserving of being made fun of, or not?’

    Actually, this is an interesting question, thank you for asking it. What if the curvature is measured to be positive? There is natural explanation: tunneling from nothing (Hartle-Hawking, Vilenkin) produces universes with positive spatial curvature, whereas tunneling from a higher dS state produces universes with negative spatial curvature. Either of these processes is sufficient to populate the landscape, so indeed neither observation would falsify the environmental solution to the cc problem, it would tell us something important about quantum cosmology.

    The relation between these two processes has never clearly been understood by quantum cosmologists working in the effective QM-GR: every one has their own person favorite for the wavefunction. A complete theory of cosmology should explain the relation between them, most likely both occur in some form but presumably, barring an accident, one or the other is much more likely. Indeed, providing a theoretically convincing solution to this old area of controversy is a target for string theory, though observational test depends on having few enough e-foldings for the curvature still to be visible.

  13. z3 says:

    Peter finally grudgingly admitted to an inaccuracy in an off-the-cuff, un-researched remark in a comment section — this is actually a pretty hard thing to do, especially in public view. One can respect him for that.

    USBC finally calmed down and start to address physics — for a while there he was latched on to a secondary topic, attacking in full rage, and in so doing exposed the fact that he himself is not quite in full command of the history of the topic either.

    Anon just extrapolated Peter’s one error by 3 orders of magnitude to show that the entirety of Peter is in error. I hope he is more conservative in his main areas of study.

    And we all now see how memory is not 100% reliable and why one must always research the facts before making quick strong statements.

  14. dan says:

    Hello Peter,

    Even if string theory is unable to make predictions at any level, could you accept it as legimate scientific research as a way of explaining (but not necessarily predicting) fundamental physics, such as 3 generations of fermions?

    There are many instances of a scientific theory (i.e meterology, evolutionm, n-body problem, fluid dynamics) that cannot predict (because too much is unknown or equations poorly understood or impossible to solve) but are considered scientific for other reasons.


  15. TheGraduate says:


    I could be wrong but I think that you can get predictions out of meteorology and mathematical analyses of the n-body problem and of fluid dynamics over a small enough time scale and in special circumstances.

    Also in each case, there are actual physical systems that can be catalogued. Collection, arranging and categorization is also a scientific activity. For instance, a meteorologist can observe and classify types of clouds and the conditions under which they are known to form.

    In the case of evolution, you can easily apply selection pressure to bacteria, viruses, protozoa etc and observe the affects on different allelles. eg. exposing bacteria to antibiotics and observing the shift in the population to antibiotic-resistant bacteria

  16. TheGraduate says:


    I think a good way to figure out if one is doing science or religion is to ask yourself whether your approach is vulnerable to the sorts of errors that religious approaches to finding truth are vulnerable. I think in the case of string theory, they certainly are.

    Without accountability, the reasoning process tends to go awry.

  17. dan says:

    so TheGraduate Says:,
    would u c string theory as a scientific theory in the sense of explanatory rather than predictive?

  18. nigel cook says:


    String theory probably “explains” why the science fiction of extra dimensional spacetime, superpartners, gravitons, multiple universes, and time travelling in wormholes is so popular.

    There is a list of hyperspace films at http://en.wikipedia.org/wiki/Hyperspace_(science_fiction) and we owe a great debt of gratitude to Professor M. Kaku, who is the “anchor” of hyperspace, according to the first reference on that Wiki page.

    Without string theory “explanation” hype, where would science fiction be today? It would still be half-plausible fictional stories based upon science. String “explanation” is not just replacing half decent religions (thanks to Susskind’s new book on the anthropic principle, landscape and intelligent design), it is also replacing decent science fiction with brilliant stringy films.

    I wonder if Speilberg and others in Hollywood will become stringy defenders, joining Motl, Distler, and Susskind? ;-)

  19. woit says:


    To have a scientific explanation of anything, you have to be able to check that the explanation is correct. That’s why you need a testable prediction. “Explanations” that can’t be checked and tested are not part of science.

    All the sciences you mention make many testable predictions, meteorology is all about learning how to make better and better predictions.

  20. dan says:

    Nigel, I find what you say amusing.

    Peter, string theorists have claimed that the fact string theory can derive BH-entropy a priori in agreement with Hawking-Berkenstein calculations, predict the spectrum of the standard model and three generations and particle masses, provide a theory of gravity in agreement with general relativity at long distances, are examples which check if the explanation are correct.

  21. woit says:


    This kind of discussion is pretty off-topic, and has taken place a hundred times here before.

    1. The BH entropy calculations don’t apply to physical black holes, so don’t predict anything you can observe, even if you do find a black hole you can examine.

    2. It’s just a complete bald-faced lie to say that string theory predicts the standard model, 3 generations and particle masses.

    3. The idea that it gives GR at long distances is the strongest argument in favor of string theory. This is not a prediction, at best a postdiction (and you can argue about this..), it’s the recent people picked this theory to look at. As Lisa Randall has been known to point out: “Yes, string theory predicts GR. In 10 dimensions.”

  22. dan says:

    Sorry if it’s off-topic, I should say “output is MSSM”

  23. Peter Woit says:


    It’s every bit as much nonsense to claim that string theory predicts the MSSM as to say that it predicts the standard model.