Sabine Hossenfelder

To find the way forward in the foundations of physics, we need results, not null-results. When testing new hypotheses takes decades of construction time and billions of dollars, we have to be careful what to invest in. Experiments have become too costly to rely on serendipitous discoveries. Beauty-based methods have historically not worked. They still don’t work. It’s time that physicists take note.

6 thoughts on “Sabine Hossenfelder

  1. shinichi Post author

    When Beauty Gets in the Way of Science
    Insisting that new ideas must be beautiful blocks progress in particle physics.
    by Sabine Hossenfelder


    The biggest news in particle physics is no news. In March, one of the most important conferences in the field, Rencontres de Moriond, took place. It is an annual meeting at which experimental collaborations present preliminary results. But the recent data from the Large Hadron Collider (LHC), currently the world’s largest particle collider, has not revealed anything new.

    Forty years ago, particle physicists thought themselves close to a final theory for the structure of matter. At that time, they formulated the Standard Model of particle physics to describe the elementary constituents of matter and their interactions. After that, they searched for the predicted, but still missing, particles of the Standard Model. In 2012, they confirmed the last missing particle, the Higgs boson.

    The Higgs boson is necessary to make sense of the rest of the Standard Model. Without it, the other particles would not have masses, and probabilities would not properly add up to one. Now, with the Higgs in the bag, the Standard Model is complete; all Pokémon caught.

    The Standard Model may be physicists’ best shot at the structure of fundamental matter, but it leaves them wanting. Many particle physicists think it is simply too ugly to be nature’s last word. The 25 particles of the Standard Model can be classified by three types of symmetries that correspond to three fundamental forces: The electromagnetic force, and the strong and weak nuclear forces. Physicists, however, would rather there was only one unified force. They would also like to see an entirely new type of symmetry, the so-called “supersymmetry,” because that would be more appealing. Oh, and additional dimensions of space would be pretty. And maybe also parallel universes. Their wish list is long.

    It has become common practice among particle physicists to use arguments from beauty to select the theories they deem worthy of further study. These criteria of beauty are subjective and not evidence-based, but they are widely believed to be good guides to theory development. The most often used criteria of beauty in the foundations of physics are presently simplicity and naturalness.

    By “simplicity,” I don’t mean relative simplicity, the idea that the simplest theory is the best (a.k.a. “Occam’s razor”). Relying on relative simplicity is good scientific practice. The desire that a theory be simple in absolute terms, in contrast, is a criterion from beauty: There is no deep reason that the laws of nature should be simple. In the foundations of physics, this desire for absolute simplicity presently shows in physicists’ hope for unification or, if you push it one level further, in the quest for a “Theory of Everything” that would merge the three forces of the Standard Model with gravity.

    As an ex-particle physicist, I understand the desire to have an encompassing theory for the structure of matter.

    The other criterion of beauty, naturalness, requires that pure numbers that appear in a theory (i.e., those without units) should neither be very large nor very small; instead, these numbers should be close to one. Exactly how close these numbers should be to one is debatable, which is already an indicator of the non-scientific nature of this argument. Indeed, the inability of particle physicists to quantify just when a lack of naturalness becomes problematic highlights that the fact that an unnatural theory is utterly unproblematic. It is just not beautiful.

    Anyone who has a look at the literature of the foundations of physics will see that relying on such arguments from beauty has been a major current in the field for decades. It has been propagated by big players in the field, including Steven Weinberg, Frank Wilczek, Edward Witten, Murray Gell-Mann, and Sheldon Glashow. Countless books popularized the idea that the laws of nature should be beautiful, written, among others, by Brian Greene, Dan Hooper, Gordon Kane, and Anthony Zee. Indeed, this talk about beauty has been going on for so long that at this point it seems likely most people presently in the field were attracted by it in the first place. Little surprise, then, they can’t seem to let go of it.

    Trouble is, relying on beauty as a guide to new laws of nature is not working.


    Since the 1980s, dozens of experiments looked for evidence of unified forces and supersymmetric particles, and other particles invented to beautify the Standard Model. Physicists have conjectured hundreds of hypothetical particles, from “gluinos” and “wimps” to “branons” and “cuscutons,” each of which they invented to remedy a perceived lack of beauty in the existing theories. These particles are supposed to aid beauty, for example, by increasing the amount of symmetries, by unifying forces, or by explaining why certain numbers are small. Unfortunately, not a single one of those particles has ever been seen. Measurements have merely confirmed the Standard Model over and over again. And a theory of everything, if it exists, is as elusive today as it was in the 1970s. The Large Hadron Collider is only the most recent in a long series of searches that failed to confirm those beauty-based predictions.

    These decades of failure show that postulating new laws of nature just because they are beautiful according to human standards is not a good way to put forward scientific hypotheses. It’s not the first time this has happened. Historical precedents are not difficult to find. Relying on beauty did not work for Kepler’s Platonic solids, it did not work for Einstein’s idea of an eternally unchanging universe, and it did not work for the oh-so-pretty idea, popular at the end of the 19th century, that atoms are knots in an invisible ether. All of these theories were once considered beautiful, but are today known to be wrong. Physicists have repeatedly told me about beautiful ideas that didn’t turn out to be beautiful at all. Such hindsight is not evidence that arguments from beauty work, but rather that our perception of beauty changes over time.

    Physicists must, first and foremost, learn from their failed predictions. So far, they have not.

    That beauty is subjective is hardly a breakthrough insight, but physicists are slow to learn the lesson—and that has consequences. Experiments that test ill-motivated hypotheses are at high risk to only find null results; i.e., to confirm the existing theories and not see evidence of new effects. This is what has happened in the foundations of physics for 40 years now. And with the new LHC results, it happened once again.

    The data analyzed so far shows no evidence for supersymmetric particles, extra dimensions, or any other physics that would not be compatible with the Standard Model. In the past two years, particle physicists were excited about an anomaly in the interaction rates of different leptons. The Standard Model predicts these rates should be identical, but the data demonstrates a slight difference. This “lepton anomaly” has persisted in the new data, but—against particle physicists’ hopes—it did not increase in significance, is hence not a sign for new particles. The LHC collaborations succeeded in measuring the violation of symmetry in the decay of composite particles called “D-mesons,” but the measured effect is, once again, consistent with the Standard Model. The data stubbornly repeat: Nothing new to see here.

    Of course it’s possible there is something to find in the data yet to be analyzed. But at this point we already know that all previously made predictions for new physics were wrong, meaning that there is now no reason to expect anything new to appear.

    Yes, null results—like the recent LHC measurements—are also results. They rule out some hypotheses. But null results are not very useful results if you want to develop a new theory. A null-result says: “Let’s not go this way.” A result says: “Let’s go that way.” If there are many ways to go, discarding some of them does not help much.

    To find the way forward in the foundations of physics, we need results, not null-results. When testing new hypotheses takes decades of construction time and billions of dollars, we have to be careful what to invest in. Experiments have become too costly to rely on serendipitous discoveries. Beauty-based methods have historically not worked. They still don’t work. It’s time that physicists take note.

    And it’s not like the lack of beauty is the only problem with the current theories in the foundations of physics. There are good reasons to think physics is not done. The Standard Model cannot be the last word, notably because it does not contain gravity and fails to account for the masses of neutrinos. It also describes neither dark matter nor dark energy, which are necessary to explain galactic structures.

    So, clearly, the foundations of physics have problems that require answers. Physicists should focus on those. And we currently have no reason to think that colliding particles at the next higher energies will help solve any of the existing problems. New effects may not appear until energies are a billion times higher than what even the next larger collider could probe. To make progress, then, physicists must, first and foremost, learn from their failed predictions.

    So far, they have not. In 2016, the particle physicists Howard Baer, Vernon Barger, and Jenny List wrote an essay for Scientific American arguing that we need a larger particle collider to “save physics.” The reason? A theory the authors had proposed themselves, that is natural (beautiful!) in a specific way, predicts such a larger collider should see new particles. This March, Kane, a particle physicist, used similar beauty-based arguments in an essay for Physics Today. And a recent comment in Nature Reviews Physics about a big, new particle collider planned in Japan once again drew on the same motivations from naturalness that have already not worked for the LHC. Even the particle physicists who have admitted their predictions failed do not want to give up beauty-based hypotheses. Instead, they have argued we need more experiments to test just how wrong they are.

    Will this latest round of null-results finally convince particle physicists that they need new methods of theory-development? I certainly hope so.

    As an ex-particle physicist myself, I understand very well the desire to have an all-encompassing theory for the structure of matter. I can also relate to the appeal of theories such a supersymmetry or string theory. And, yes, I quite like the idea that we live in one of infinitely many universes that together make up the “multiverse.” But, as the latest LHC results drive home once again, the laws of nature care heartily little about what humans find beautiful.

  2. shinichi Post author

    Lost in Math: How Beauty Leads Physics Astray

    by Sabine Hossenfelder

    In this “provocative” book (New York Times), a contrarian physicist argues that her field’s modern obsession with beauty has given us wonderful math but bad science.

    Whether pondering black holes or predicting discoveries at CERN, physicists believe the best theories are beautiful, natural, and elegant, and this standard separates popular theories from disposable ones. This is why, Sabine Hossenfelder argues, we have not seen a major breakthrough in the foundations of physics for more than four decades.

    The belief in beauty has become so dogmatic that it now conflicts with scientific objectivity: observation has been unable to confirm mindboggling theories, like supersymmetry or grand unification, invented by physicists based on aesthetic criteria. Worse, these “too good to not be true” theories are actually untestable and they have left the field in a cul-de-sac. To escape, physicists must rethink their methods. Only by embracing reality as it is can science discover the truth.



    They were so sure, they bet billions on it. For decades physicists told us they knew where the next discoveries were waiting. They built accelerators, shot satellites into space, and planted detectors in underground mines. The world prepared to ramp up the physics envy. But where physicists expected a breakthrough, the ground wouldn’t give. The experiments didn’t reveal anything new.

    What failed physicists wasn’t their math; it was their choice of math. They believed that Mother Nature was elegant, simple, and kind about providing clues. They thought they could hear her whisper when they were talking to themselves. Now Nature spoke, and she said nothing, loud and clear.

    Theoretical physics is the stereotypical math-heavy, hard-to-understand discipline. But for a book about math, this book contains very little math. Strip away equations and technical terms and physics becomes a quest for meaning—a quest that has taken an unexpected turn. Whatever laws of nature govern our universe, they’re not what physicists thought they were. They’re not what I thought they were.

    Lost in Math is the story of how aesthetic judgment drives contemporary research. It is my own story, a reflection on the use of what I was taught. But it is also the story of many other physicists who struggle with the same tension: we believe the laws of nature are beautiful, but is not believing something a scientist must not do?

  3. shinichi Post author


    by ザビーネ・ホッセンフェルダー

    translated by 吉田三知世

    物理学の基盤的領域では30年以上も、既存の理論を超えようとして失敗し続けてきたと著者は言う。実験で検証されないまま理論が乱立する時代が、すでに長きに渡っている。それら理論の正当性の拠り所とされてきたのは、数学的な「美しさ」や「自然さ」だが、なぜ多くの物理学者がこうした基準を信奉するのか? 革新的な理論の美が、前世紀に成功をもたらした美の延長上にあると考える根拠はどこにあるのか? そして、超対称性、余剰次元の物理、暗黒物質の粒子、多宇宙……等々も、その信念がはらむ錯覚の産物だとしたら?研究者たち自身の語りを通じて浮かび上がるのは、究極のフロンティアに進撃を続けるイメージとは異なり、空振り続きの実験結果に戸惑い、理論の足場の不確かさと苦闘する物理学の姿である。「誰もバラ色の人生なんて約束しませんでしたよ。これはリスクのある仕事なのです」(ニマ・アルカニ=ハメド)、「気がかりになりはじめましたよ、確かに。たやすいことだろうなんて思ったことは一度もありませんが」(フランク・ウィルチェック)著者の提案する処方箋は、前提となっている部分を見つめ直すこと、あくまで観測事実に導かれること、それに、狭く閉じた産業の体になりつつあるこの分野の風通しをよくすることだ。しかし、争点はいまだその手前にある。物理学は「数学の美しさのなかで道を見失って」いるのだろうか? 本書が探針を投じる。







  4. shinichi Post author

    Lost in Math: How Beauty Leads Physics Astray

    by Moira I. Gresham

    Lost in Math: How Beauty Leads Physics Astray, Hossenfelder Sabine

    In brief. Read this book. It is fascinating. Don’t expect it to fit a mold of other books on theoretical physics for a popular audience. The core strand of this book is an insightful argument of import to philosophy of and best practices in physics. It also offers much more. Read it with good humor, generosity, skepticism, and care. Also read Frank Wilczek’s response in Physics Today.

    Introduction. The author, Sabine Hossenfelder, is a German theoretical physicist who works on quantum gravity and physics beyond the standard model. She is also an excellent writer. She’s been funded through short-term research grants and postdoctoral appointments since she received her PhD in 2003. An impetus for her book is Hossenfelder’s self-admitted personal crisis: “And while I witnessed my profession slip into crisis, I slipped into my own personal crisis. I’m not sure anymore that what we do here, in the foundations of physics, is science. And if not, why am I wasting my time on it?” (p. 2). Hossenfelder goes on to explore, analyze, and generalize from the supposed crisis. In a style inspired by that of the book, I begin by describing my own personal micro-crisis in deciding whether to review it.

    I was approached about this book review late last summer. After using my superior research skills to locate and read the blurb, “A contrarian argues that modern physicists’ obsession with beauty has given us wonderful math but bad science,” the table of contents, e.g., “Chap. 5: Ideal Theories:… In which… I fly to Austin, let Steven Weinberg talk at me, and realize just how much we do to avoid boredom,” and the first part of the preface, “They were so sure, they bet billions on it… But where physicists expected a breakthrough, the ground wouldn’t give. The experiments didn’t reveal anything new,” I started freaking out a little bit. I can’t believe she just painted the LHC as a waste of money. The politicians are going to run with this and use it as ammunition against science. And how could she speak so apparently irreverently about an interaction with Steven Weinberg? This is really going to piss some people off, and if I’m associated with it in any way, I could be kicked off the island.

    However—I thought, after having calmed down—even the table of contents is hilarious. I’ve heard that Hossenfelder’s blog is one of the best written and most read in the business. (Although I don’t read it, nor do I regularly read any blogs.) I’m nervous about the time this could take from precious research hours, but it could be intellectually stimulating. I’m on sabbatical this fall, and if I don’t say yes to a review like this now, I never will. Okay, I’ll do it.

    What is this book and should I read it? Lost in Math is organized around about a dozen captivating interviews with physicists—mostly theoretical physicists. Several of the interviewees are acknowledged luminaries, while a few are working on their own at the fringes of science. Hossenfelder describes why and how she interviews each physicist. The interviewees reveal interesting ideas and insights. They also provide a window on the way theoretical physicists think and work. She weaves in impressively clear, concise, and untechnical explanations of the theoretical physics topics referenced in the interviews. Hossenfelder also weaves in histories of physics. While the physics ideas and histories are interesting on their own—and could serve as a nonspecialist-appropriate introduction to a slew of contemporary theoretical physics topics—they are not the main point of the book. Though entertaining, the table of contents and “In Brief” summaries at the end of each chapter give no real clue to the physics described in any given chapter. They’re not designed to. To satisfy your curiosity, in the table below, I have summarized the physics topics discussed along with physicists interviewed, by chapter.

    Hossenfelder highlights social elements of physicists’ interactions, along with her own and other physicists’ feelings. She also weaves some of her personal story throughout the book, including how and why she got into the business of theoretical physics, and difficult realities that she faces due to the structure of her employment. Some physics students might particularly benefit from reading and thinking about these human aspects of being a physicist. But Hossenfelder spends substantial space addressing human aspects of doing theoretical physics for reasons other than entertainment value and outreach: she thinks that our failure to recognize and address our social/human biases (related to factors that almost kept me from writing this review) is hurting the science we do. She also demonstrates some of the kinds of metacognition that might help to keep our human biases from hurting progress in science.

    One might rightly point out—as Frank Wilczek does in his Physics Today review of the book—that Hossenfelder paints an overly dire picture of the state of foundational physics as a whole (encompassing high energy physics, cosmology, quantum gravity, and foundations of quantum mechanics). She omits details in her accounts of the contemporary history of various physics subfields that might paint a more optimistic picture. Chapter 8 provides a key that alerts us to these choices. In it, she presents two histories of string theory—one pessimistic history and one romantic history of similar length. Then, she writes, “Both stories are true. But it’s more fun if you pick one and ignore the other,” (p. 176). Hint: Hossenfelder generally chooses one. We should not lose sight of this. I think that we should not be too upset about the pessimistic choices, either. (I was upset, at first—especially in Chap. 9.) First, she’s providing a counterpoint to the romantic stories that most other physicists tell (at least in public), where they typically leave out a different set of details. Second, she’s trying to create a productive tension that might move some of us scientists to do better science. Hossenfelder delivers some heavy blows, but they are moderated by sarcasm, witticisms, expressions of self-doubt, and (more rarely) glimmers of optimism.

    The central strand of Lost in Math is an argument that (some) theoretical physicists and philosophers of science have lost the track of where science ends and philosophy begins. This is potentially damaging. This is not to say that physicists ought not value philosophy, do philosophy, or engage with philosophers. In fact, we theoretical physicists could use philosophers’ help in thoroughly understanding the subtle boundary between philosophical assumptions, mathematics, and scientific reasoning at the foundations of physics. Hossenfelder argues that one mechanism by which (some) physicists’ delusions developed is through the mathematical formalization of aesthetic ideals over the past several decades. Particular ideals of beauty have become entrenched and even elevated in some circles to constituting solid scientific criteria—able to justify rather than merely to guide scientific work, even in the absence of observational evidence. Furthermore, even if serving merely as guides, Hossenfelder worries that entrenchment and uniformity of aesthetic ideals and ideas, more generally, could stifle innovation and lead us astray. At various points throughout the book and in advertising it, she makes the stronger assertion that the field as a whole has been led astray. Whether or not Hossenfelder’s characterization of exactly who has been led astray is correct, her arguments deserve serious attention. Hossenfelder argues that other important factors in the entrenchment and elevation of aesthetic ideals and ideas within theoretical physics (and other sciences) are rooted in cognitive biases that have been exacerbated by recent shifts in the way science is funded, disseminated, and done. That it’s always worked out in the past shouldn’t give us too much comfort. The final chapter ends with a summary of three main lessons for theoretical physicists to draw from her arguments. Appendix C contains further recommendations relevant to all of science—written for scientists, administrators, editors, and others—all of which seem sensible to me.

    Going back to my initial anxieties, I still worry that politicians might be emboldened to further cut science funding after misunderstanding this book or accounts of it, but I’m glad that Hossenfelder wrote it. (If you are a politician, please read rather than skim, or skip to Appendix C.) She is insightful and makes arguments that deserve careful attention. She makes it easy to give that attention by writing so well and humorously. I’m no longer worried about being kicked off the island.

    To conclude, you should read this book. And don’t give it short shrift. Nor should you take it entirely at its word. Recommend Lost in Math to your philosophically inclined physics students—especially those interested in theoretical physics. Also recommend it to any critical reader interested in the philosophy of science or science policy.

  5. shinichi Post author



  6. shinichi Post author

    The Hidden Rules of Physics
    In which I realize I don’t understand physics anymore. I talk to friends and colleagues, see I’m not the only one confused, and set out to bring reason back to Earth.

    The Conundrum of the Good Scientist


    After twenty years in theoretical physics, most people I know make a career by studying things nobody has seen. They have concocted mind-boggling new theories, like the idea that our universe is but one of infinitely many that together form a “multiverse.” They have invented dozens of new particles, declared that we are projections of a higher-dimensional space and that space is spawned by wormholes that tie together distant places.

    These ideas are highly controversial and yet exceedingly popular, speculative yet intriguing, pretty yet useless. Most of them are so difficult to test, they are practically untestable. Others are untestable even theoretically. What they have in common is that they are backed up by theoreticians convinced that their math contains an element of truth about nature. Their theories, they believe, are too good to not be true.

    * * *

    第1章 物理学の隠れたルール






Leave a Reply

Your email address will not be published.