Skip to the Main Content

Note:These pages make extensive use of the latest XHTML and CSS Standards. They ought to look great in any standards-compliant modern browser. Unfortunately, they will probably look horrible in older browsers, like Netscape 4.x and IE 4.x. Moreover, many posts use MathML, which is, currently only supported in Mozilla. My best suggestion (and you will thank me when surfing an ever-increasing number of sites on the web which have been crafted to use the new standards) is to upgrade to the latest version of your browser. If that's not possible, consider moving to the Standards-compliant and open-source Mozilla browser.

February 12, 2023

Talk on the Tenfold Way

Posted by John Baez

There are ten ways that a substance can have symmetry under time reversal, switching particles and holes, both or neither. But this fact turns out to extend far beyond condensed matter physics! It’s built into the fabric of mathematics in a deep way.

I gave a talk on this at Nicohl Furey’s seminar Algebra, Particles and Quantum Theory, and you can see a video of my talk here.

You can also watch another version, where I explain this stuff to my friend James Dolan.

I like the idea of being able to watch an official talk but also watch the speaker chatting about that talk with a friend. It gives another view of the material. I skim over stuff Jim already knows, explain things I didn’t have time to get into in the actual talk, and emphasize the things I don’t understand. And he points out lots more patterns lurking in the tenfold way!

You can see my slides and more notes here.

Posted at February 12, 2023 7:19 AM UTC

TrackBack URL for this Entry:   https://golem.ph.utexas.edu/cgi-bin/MT-3.0/dxy-tb.fcgi/3448

17 Comments & 0 Trackbacks

Re: Talk on the Tenfold Way

Presumably this is related to the four connected components of the Lorenz group O(1,3), cf

https://en.wikipedia.org/wiki/Wu_experiment

and maybe even axions

https://en.wikipedia.org/wiki/Peccei%E2%80%93Quinn_theory

oh my?

Posted by: jack morava on January 31, 2023 5:47 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Hi!

There’s definitely a connection to C, P, and T symmetry and the 4 components of the Lorentz group (containing 1, P, T, and PT)…

… but the point of my talk is that the 10-fold way doesn’t rely on any concept of spacetime: it shows up as soon as you study /2\mathbb{Z}\!/\!2-graded Hilbert spaces. I’ll explain how any unitary representation of a group on a /2\mathbb{Z}\!/\!2-graded Hilbert space that’s decomposable into irreducibles is automatically X\mathrm{X}-graded, where

X={0,1,2,3,4,5,6,7,0,1} \mathrm{X} = \{0,1,2,3,4,5,6,7,\mathbf{0},\mathbf{1} \}

is a 10-element set with a certain monoid structure. This monoid has the groups /8\mathbb{Z}\!/\!8 and /2\mathbb{Z}\!/\!2 as submonoids. It unifies real and complex Bott periodicity!

By the way, it’s the ‘Lorentz’ group, not the ‘Lorenz’ group. Lorenz was a physicist who worked on electromagnetism; so was Lorentz. The second was obtained from the first by a transformation that mixed the tt and zz coordinates.

Posted by: John Baez on January 31, 2023 7:03 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

thanks, sorry, my bad. then there’s Weil and Weyl…

but… could π 0O(1,3)\pi_0 O(1,3) interact in interesting ways with representation theory graded somehow by this ten-element monoid?

Posted by: jack morava on January 31, 2023 11:58 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

It certainly interacts in some way!
π 0O(1,3)/2×/2\pi_0 O(1,3) \cong \mathbb{Z}\!/\!2 \times \mathbb{Z}\!/\!2 is generated by time reversal T and space reversal, or ‘parity’ P. Physicists like to think of it as a subgroup of /2×/2×/2\mathbb{Z}\!/\!2 \times \mathbb{Z}\!/\!2 \times \mathbb{Z}\!/\!2 where the third generator, C, switches particles and antiparticles.

In physics this larger group often acts on a Hilbert space H\mathbf{H} in such a way that P acts as a unitary operator while C and T act as antiunitary operators. Sometimes H\mathbf{H} is /2\mathbb{Z}\!/\! 2-graded in such a way that ‘particles’ are even and ‘antiparticles’ are odd. Then C acts as an odd antiunitary operator while T acts as an even antiunitary operator.

In my talk slides I discuss a more general situation where we have an irreducible representation ρ\rho of a group on a /2\mathbb{Z}\!/\! 2-graded Hilbert space. This superficially has little to do with the story I’ve just been telling: yes there’s /2\mathbb{Z}\!/\! 2-graded Hilbert space, and a group, but we don’t necessarily have O(3,1)O(3,1) around. Nonetheless, for any such representation, we have:

The Tenfold Way. The irreducible unitary representations ρ\rho of a Z/2\Z/2-graded group GG on a super Hilbert space H\mathbf{H} come in 10 types, based on the commutant: the real-linear operators that commute with ρ(g)\rho(g) for all GG.

In 9 of these types the commutant contains:

  • an even antiunitary TT with either T 2=1T^2 = 1, T 2=1T^2 = -1, or no such TT

and

  • an odd antiunitary CC with either C 2=1C^2 = 1,C 2=1C^2 = -1, or no such CC.

In the 10th type the commutant contains:

  • no such TT or CC, but an odd unitary SS; we may assume S 2=1S^2 = 1.

In short, even when TT and CC aren’t put in via the Lorentz group O(1,3), they magically appear (or not) on their own, having the properties we expect in the O(1,3) case!

Posted by: John Baez on February 1, 2023 11:01 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

IANAPhysicist, so there seems to be something funny about

https://en.wikipedia.org/wiki/CP_violation

that I don’t understand…

Posted by: jack morava on February 2, 2023 8:38 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

I don’t know if I’m a physicist, but I’ve studied the subject for a long time, so tell me what seems funny about CP violation and maybe I can help you out.

CP violation is a big deal — and a big surprise, back around 1964.

Posted by: John Baez on February 3, 2023 12:12 AM | Permalink | Reply to this

Re: Talk on the Tenfold Way

I guess it’s off-topic w.r.t. condensed matter physics but the strong CP problem

“… is a puzzling question in particle physics: Why does quantum chromodynamics (QCD) seem to preserve CP-symmetry?”

I don’t know the current status of the sci-fi story at

https://arxiv.org/abs/1610.08297

but think the idea of an axion with de Broglie wavelength of the order of kiloparsecs is hilarious.

Posted by: jack morava on February 4, 2023 2:58 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Yes, the strong CP problem is very interesting. It’s a ‘problem’ if you believe, as most particle physicists do, that you need a good explanation for any renormalizable term that you can cook up from the fields in the Standard Model that’s missing from the Standard Model Lagrangian. I.e. just setting a coupling constant equal to zero without explanation counts as ‘cheating’ for them.

The only such missing term is the CP-violating term tr(FF) \tr(F \wedge F) where FF is the curvature of the SU(3)SU(3) connection describing the strong force. This is a ‘topological term’: if we treat spacetime as compact its integral gives the 2nd Chern class of the principal SU(3)SU(3) bundle over spacetime… up to a constant factor that I’m too lazy to recall.

Posted by: John Baez on February 7, 2023 5:34 AM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Hadn’t understood that - it seems very intriguing, thanks!

Posted by: jack morava on February 7, 2023 1:05 PM | Permalink | Reply to this

C is a unitary

In physics this larger group often acts on a Hilbert space H in such a way that P acts as a unitary operator while C and T act as antiunitary operators.

In relativistic quantum field theory, C and P are implemented as unitary operators and T as an anti-unitary operator. (CPT is anti-unitary.)

Posted by: Jacques Distler on February 3, 2023 2:35 PM | Permalink | PGP Sig | Reply to this

Re: Talk on the Tenfold Way

Jacques wrote:

In relativistic quantum field theory, CC and PP are implemented as unitary operators and TT as an anti-unitary operator.

Good point! A lot of condensed matter physics uses an antiunitary CC. See for example

For a more mathematical approach, see:

I just noticed a discussion of this here, where Niall writes:

The interesting thing is that the particle hole operator is unitary and linear in the second quantised picture, but antiunitary and antilinear in the first quantised picture.

I think that’s important.

Posted by: John Baez on February 3, 2023 8:34 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

You have “division algebra” defined in the slides to include associativity so as to limit the options to three, \mathbb{R}, \mathbb{C}, \mathbb{H}.

Of course, elsewhere you had us think about the four normed division algebras (allowing non-associativity) and their relation to supersymmetry, (cf. nLab).

If the “super” version of the former gives us the tenfold way, is there a “super” version of the latter?

Posted by: David Corfield on February 9, 2023 4:55 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Good point! I think the reasonable definition of a normed division superalgebra A=A 0+A 1A = A_0 + A_1 is one with norms on A 0A_0 and A 1A_1 such that

ab=ab \|a b\| = \|a\| \|b\|

whenever aa and bb are homogeneous elements, i.e. aA ia \in A_i and bA jb \in A_j for some i,ji,j.

With this definition it’s easy to show that all the associative division superalgebras become normed division superalgebras, and that every associative normed division algebra is of this form. So, there are just 10 associative normed division superalgebras.

In 2005 Todd Trimble wrote up a classification of the associative division superalgebras. I asked him if he could classify the alternative division superalgebras, and he did. He found that besides the 10 associative ones there were a few coming from the octonions. Unfortunately I’ve lost his results. But by now I understand Todd’s arguments well enough that I could redo this classification in an hour.

Copying the arguments you can find at the link, you’ll get one alternative division superalgebra with A 0=𝕆,A 0={0}A_0 = \mathbb{O}, A_0 = \{0\} and one or more with A 0=𝕆,A 1=𝕆A_0 = \mathbb{O}, A_1 = \mathbb{O}. And, I’m pretty sure all of these will be normed division superalgebras.

Further, I predict that these will give us all the normed division superalgebras, except for the 10 associative ones we already know.

I would pursue this more assiduously if I could think of something interesting to do with alternative or normed division superalgebras.

Hmm.

Maybe we could use them to build one or more ‘exceptional Jordan superalgebras’ analogous to the usual one defined using octonions. I think someone has already looked into Jordan superalgebras.

Or maybe we could use them to build exceptional simple Lie superalgebras. I think Kac has classified those. I see that in 1998 he classified infinite-dimensional exceptional Lie superalgebras, finding exotic delights like “E(3,6)”. But that’s too fancy for me right now. Someone must have done the finite-dimensional ones earlier.

Posted by: John Baez on February 9, 2023 5:38 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Thanks!

I think someone has already looked into Jordan superalgebras.

Someone has even started an nLab page on them: Jordan superalgebra, where we learn that over an algebraically closed field of characteristic 00

The only exceptional finite-dimensional example is the 10-dimensional Jordan superalgebra K 10K_10.

Posted by: David Corfield on February 10, 2023 9:23 AM | Permalink | Reply to this

Re: Talk on the Tenfold Way

That someone was David Corfield himself two and a half years ago:

https://nforum.ncatlab.org/discussion/11491/

Posted by: Madeleine Birchfield on February 16, 2023 1:17 AM | Permalink | Reply to this

Re: Talk on the Tenfold Way

Dear professor John Baez,

I’ve watched your lectures on the ten fold way organized by Cohl Furey, but I am really curious about the missing Octonions.

The sequence stops at H+H (wikipedia says these are split biquaternions instead of split quaternions), instead of O.

But if you keep track of the real minimal representation, from week 104:

“C0 is 1; C1 is 2; C2 is 4; C3 is 4; C4 is 8; C5 is 8; C6 is 8; C7 is 8; C8 is 16;

The sphere S0 admits 0 linearly independent vector fields. The sphere S1 admits 1 linearly independent vector fields. The sphere S3 admits 3 linearly independent vector fields. The sphere S7 admits 7 linearly independent vector fields. ” On https://math.ucr.edu/home/baez/octonions/node9.html

you said

“On First of all, as a smooth manifold KP 1\KP^1 is just a sphere with dimension equal to that of K\K

I can notice that there is enough space for you to make a 8 fold way to get Octonions somewhere by successive embedding of hopf fibrations, by picking up spheres related to protective algebras around the 8 path if you defined. I am not sure if that is possible.

Posted by: Daniel Diniz on January 6, 2024 5:03 PM | Permalink | Reply to this

Re: Talk on the Tenfold Way

One Week 106, you said:

“So we have a fascinating situation: an infinite tower of RPn’s, an infinite tower of CPn’s, an infinite tower of HPn’s, but an abortive tower of OPn’s going only up to n = 2 and then fizzling out.”

So, the bott periodicity doesn’t work with OPN, but given that S7 only exist as an “auxiliary” of the clock, either as a total space or a bundle. This could be a type of operator doubling and splitting of within the 8 fold way.

Posted by: Daniel Diniz on January 6, 2024 5:17 PM | Permalink | Reply to this

Post a New Comment