64

According the the Wikipedia page, the second generation proof is up to at least nine volumes: six by Gorenstein, Lyons and Solomon dated 1994-2005, two covering the quasithin business by Aschbacher and Smith in 2004, and one by Aschbacher, Lyons, Smith and Solomon in 2011. However, this latter book is really just the second part of an outline of the proof, the first part of which was written by Gorenstein in the 80s (the reason for the delay is, of course, that the quasithin case hadn't actually been settled at the time of the announcement of completion). Hence the last update on the second-generation proof is really 2005.

With the recent formal proof in Coq of the Odd-order Theorem, it would be good to know where the traditional proof is up to.

EDIT 6 August 2013: Any news as to the completion of that seventh volume as mentioned in the comments?

EDIT 29 September 2016 Just a bump to this question in case people know more about where the progress is at. Books 7 and 8 should probably have made some progress since I asked this originally.

YCor
  • 60,149
David Roberts
  • 33,851
  • 21
    I asked this on the group-pub about a year ago, and I received the following email from Mark Lewis (forwarded from Ron Solomon): Richard and I HOPE to complete Book 7 by the end of 2012. Richard, Inna Capdeboscq and I have a lot of work completed on Book 8 as well. We might be able to complete it in one to 1 1/2 years after the completion of Book 7. That still leaves Books 9, 10, and 11. (The current best estimate is 11 total.) Gernot Stroth has a first draft of Book 11 written. Some work has been done on Book 9, but I would not hazard a guess on when that will see the light of day. – Steve D Nov 30 '12 at 01:27
  • 2
    @Steve, would you care to post this as an answer? – David Roberts Dec 03 '12 at 01:48
  • @SteveD - any more news of late? – David Roberts Feb 10 '14 at 23:26
  • 3
    unfortunately, no. Book 7 should be out by now, but I haven't heard anything about an upcoming release. – Steve D Feb 15 '14 at 03:53
  • 1
    Scientific American had a popular article on this topic in its July 2015 issue (Vol 313, No. 1). German version in Spektrum der Wissenschaft, März 2016. – Frieder Ladisch Oct 04 '16 at 14:29
  • 1
    @FriederLadisch Thanks! The article is Researchers Race to Rescue the Enormous Theorem before Its Giant Proof Vanishes by Stephen Ornes (doi:10.1038/scientificamerican0715-68). I quote: "Solomon and Lyons are finishing the seventh book this summer, and a small band of mathematicians have already made inroads into the eighth and ninth. Solomon estimates that the streamlined proof will eventually take up 10 or 11 volumes, which means that just more than half of the revised proof has been published." So I guess that book 7 should be done, but presumably submitted somewhere? – David Roberts Oct 05 '16 at 04:52
  • 1
    I emailed professor Lyons about this post, maybe he could regularly update this or link to a spot where one can gather regular updates. He has retired recently and was then committing all his energies to the revisions last I spoke with him. – Sidharth Ghoshal Jan 10 '18 at 23:53
  • 2
    @frogeyedpeas do you think you could get an update on the estimate on the number of volumes? I think it's surely 13 or 14 minimum by this point, based on comments in the description of Volume 9 and the Notices article mentioned in my answer. – David Roberts Feb 25 '21 at 06:28
  • 1
    his response: "I see that Timothy Chow has edited a previous answer by David Roberts and brought it up to date as well as I could. Volume 9 is just published or just about to be published -- the AMS is advertising it. It is the volume in which the Monster, the Baby Monster, the big Conway group Co_1, and the three other Fischer groups emerge from the analysis. It also has a new co-author, Inna Capdeboscq. " – Sidharth Ghoshal Feb 25 '21 at 20:53
  • 1
    @frogeyedpeas thanks, but that seems like he's keeping his cards close to his chest :-). I was interested in estimates of future content, but of course Lyons is free to keep it quiet if he so chooses. – David Roberts Feb 25 '21 at 20:57
  • 1
    @frogeyedpeas and, amazingly, I noticed the other day Volume 9 on the shelf in the library! (Along with some, but not all, of the other volumes.) – David Roberts Jun 16 '21 at 05:08
  • 1
    @SidharthGhoshal since you seem to have a direct line, is it possible to find out where volume 10 is up to? That is, has it been submitted, as projected? – David Roberts Nov 14 '22 at 11:21
  • 3
    I reached out to him: "Volume 10 is almost done, It will be submitted soon" - RL – Sidharth Ghoshal Nov 14 '22 at 20:57
  • @Sidharth cool, thanks! – David Roberts Nov 14 '22 at 21:04
  • 1
    @SidharthGhoshal and now submitted! – David Roberts Jun 24 '23 at 10:56

2 Answers2

45

There is an interesting review by Ron Solomon of a paper in this area, which has been featured on the Beyond Reviews blog. In particular, he outlines the broad tactics that people are using in CFSG II, and some of the content that will be going into volume 7.

Also, Inna Capdeboscq apparently gave an outline of volume 8, or at least a chunk of it, at the Asymptotic Group Theory conference in Budapest. This was mentioned by Peter Cameron on his blog, sadly with no detail! If anyone can get a whiff of what she said, I would be grateful.

EDIT 15 October 2016 I emailed the group-pub mailing list and was told second-hand that Ron Solomon 'has hopes' volume 7 will be submitted next year.

EDIT 27 March 2018 Thanks to Timothy Chow in a comment on another answer, here is the link to the published version of Volume 7. So now the countdown to Volume 8 starts...

EDIT 22 June 2018 Even better news: Volume 8

...is near completion and promised to the AMS by August 2018. The completion of Volume 8 will be a significant mathematical milestone in our work. (source)

Also (from the same article):

We anticipate that there will be twelve volumes in the complete series [GLS], which we hope to complete by 2023.

Considerable work has been done on this problem [the bicharacteristic case], originally by Gorenstein and Lyons, and more recently by Inna Capdeboscq, Lyons, and me. We anticipate that this will be the principal content of Volume 9 [GLS], coauthored with Capdeboscq.

When p is odd, there is a major 600-page manuscript by Gernot Stroth treating groups with a strongly p-embedded subgroup, which will appear in the [GLS] series, probably in Volume 11. There are also substantial drafts by Richard Foote, Gorenstein, and Lyons for a companion volume (Volume 10?), which together with Stroth’s volume will complete the p-Uniqueness Case.

It would be wonderful to complete our series by 2023, the sixtieth anniversary of the publication of the Odd Order Theorem. Given the state of Volumes 8, 9, 10, and 11, the achievement of this goal depends most heavily on the completion of the e(G) = 3 problem. It is a worthy goal.

EDIT Mar 2019 Volume 8 has been published. The page listing the available volumes, along with links to more details is here.

The summary of this volume is as follows:

This book completes a trilogy (Numbers 5, 7, and 8) of the series The Classification of the Finite Simple Groups treating the generic case of the classification of the finite simple groups. In conjunction with Numbers 4 and 6, it allows us to reach a major milestone in our series—the completion of the proof of the following theorem:

Theorem O: Let G be a finite simple group of odd type, all of whose proper simple sections are known simple groups. Then either G is an alternating group or G is a finite group of Lie type defined over a field of odd order or G is one of six sporadic simple groups.

Put another way, Theorem O asserts that any minimal counterexample to the classification of the finite simple groups must be of even type. The work of Aschbacher and Smith shows that a minimal counterexample is not of quasithin even type, while this volume shows that a minimal counterexample cannot be of generic even type, modulo the treatment of certain intermediate configurations of even type which will be ruled out in the next volume of our series.

EDIT February 2021 Volume 9 has now been published. From the preface:

This book contains a complete proof of Theorem $\mathcal{C}_5$, which covers the “bicharacteristic” subcase of the $e(G) \ge 4$ problem. The outcome of this theorem is that $G$ is isomorphic to one of the six sporadic groups for which $e(G)\ge 4$, or one of six groups of Lie type which exhibit both characteristic 2-like and characteristic 3-like properties. Finally, in Chapter 7, we begin the proof of Theorem $\mathcal{C}_6$ and its generalization Theorem $\mathcal{C}^∗_6$, which cover the “$p$-intermediate” case. $\ldots$ In the preceding book in this series, we had promised complete proofs of Theorems $\mathcal{C}_6$ and $\mathcal{C}^∗_6$ in this book, but because of space considerations, we postpone the completion of those theorems to the next volume.

EDIT September 2021

In response to a question from Hugo de Garis, Ron Solomon sent the following email in January 2021:

Vol. 9 is already submitted, accepted and scheduled for publication. It should be published early this year. As for the rest, my best guess now is that there will in fact be 4 further volumes, not 3. A roughly 800 pages manuscript on the Uniqueness Theorem has been completed by Gernot Stroth. With some additional material, it will fill 2 further volumes. This could probably be readied for publication by a year from now. However, our team (Inna Capdeboscq, Richard Lyons, Chris Parker and myself) are currently focussing on the remaining work to be done for the other two volumes. It is difficult to estimate how long this will take. With luck we might have a first draft completed this calendar year, but it might take longer. It is safe to say that the remaining volumes will not all be published before 2023. I hope it is also safe to say that they will all be published no later than 2025.

(Emphasis added)

EDIT 29 Dec 2021

Richard Lyons maintains an erratum for the whole published second generation CFSG on this page: https://sites.math.rutgers.edu/~lyons/cfsg/

EDIT 05 Apr 2022

In response to a further question from de Garis (see the page linked above), Solomon wrote (in March 2022):

We have been working on the theorems for both Volumes 10 and 11. Just in the past few weeks, we have decided to concentrate on the completion of Volume 10. This is proceeding very well and we should be able to submit Volume 10 for publication this year, I believe. I fear that I may have been a bit overoptimistic in predicting the completion of all the volumes by the end of 2024.

EDIT 09 Mar 2023

From a 23 January 2023 article about Inna Capdeboscq (emphasis added):

The expected length of the Generation-2 proof is of about 5,000 pages published in 12 volumes. At this moment Volumes 1 through 9 are published. Inna has been involved in the Generation-2 project for several years, providing small contributions to Volume 6 and 7. Inna co-authored the recently published Volume 9 and is currently in a process of completing Volume 10.

I don't know how this estimate of 12 volumes sits with Solomon's email from January 2021 (see the Sept '21 edit) saying there would be 4 more volumes after vol 9 was done. And though Stroth's future contribution is mentioned in the short article, I don't think this count of 12 includes his manuscript mentioned above.

EDIT 24 June 2023

I emailed Richard Lyons to double check how things are going given the hopeful progress on volume 10, mentioned above. He replied (and he and Ron Solomon gave permission to relay this):

Volume 10 has been submitted for publication.

I have received Stroth's final manuscript for the Uniqueness Case, which we plan to make the final volume.

[Ron] Solomon and I are currently working on Volume 11, the penultimate volume (provided that it fits into one volume .. it is not clear at this time whether it will or not). This will complete the proof of Theorem C_4 (the last of the seven in the Classification Grid) and begin the treatment of the Uniqueness Case for groups of even type, to mesh with Stroth's work.

So it seems Stroth will contribute one volume, to go at the end, and we will have volume 11 (and maybe vol 12) of the main series before that. So 11+2+1 (or 12+2+1) volumes in total. [edit: the +2 is the Aschbacher–Smith work, the +1 is Stroth]

EDIT 9 October 2023

Commenter colt_browning points out below that Volume 10 is due for publication 26th December, and is now available for preorder: https://bookstore.ams.org/surv-40-10. The title is The Classification of the Finite Simple Groups, Number 10: Part V, Chapters 9–17: Theorem $C_6$ and Theorem $C^*_4$, Case A, with listed authors Capdeboscq, Gorenstein, Lyons and Solomon, and it's 570 pages long.

This book is the tenth in a series of volumes whose aim is to provide a complete proof of the classification theorem for the finite simple groups based on a fairly short and clearly enumerated set of background results. Specifically, this book completes our identification of the simple groups of bicharacteristic type begun in the ninth volume of the series (see Mathematical Surveys and Monographs, Volume 40.9). This is a fascinating set of simple groups which have properties in common with matrix groups (or, more generally, groups of Lie type) defined both over fields of characteristic 2 and over fields of characteristic 3. This set includes 11 of the celebrated 26 sporadic simple groups along with several of their large simple subgroups. Together with SURV/40.9, this volume provides the first unified treatment of this class of simple groups.

Total length of volumes 1–10 is 4511 pages, and Aschbacher and Smith's two volumes fill 1320 pages. Maybe another 1000–1500 pages to go? There's an old manuscript of Stroth from the late 90s that seems to cover the "uniqueness case" (first listed article on this page), which is what his volume will cover. That's 244 pages, but it's not clear how it relates to the draft of what will become the last volume of the published second generation proof.

EDIT 22 February 2024

The preface to volume 10 gives a good outline of where the proof stands. The outstanding results are summarised as follows:

In summary, the theorems to be proved in future volumes are Case B of Theorem $\mathcal{C}^*_4$ and Odd Uniqueness theorems. The change to Theorem $\mathcal{C}^*_4$ has consequences for the proofs of both of these. While we originally conceived of a mainly 2-local proof of Theorem $\mathcal{C}_4$, we now plan to use odd local analysis as well, following the fundamental groundbreaking papers of Aschbacher [A13], [A24]. In particular, Sections 15 and 22 of [I2] no longer fit our plans and can be replaced by Chapter 11 of this volume, which contains more precise details on Theorem $\mathcal{C}^*_4$. Furthermore, our Theorem $\mathcal{C}^*_4$ will also necessitate a strengthened version of the Odd Uniqueness theorems — Theorem and Corollary $U(\sigma)$, which in turn depend on Theorem $\mathcal{M}(S)$.

Just as we have found it necessary in other volumes of this series, we again need to expand the Background Results, this time in connection with the recognition of the sporadic group $Co_2$. The expansion is made precise in (9B) of Chapter 16.

David Roberts
  • 33,851
  • 31
    My personal (vain) hope would be that the CFSG people start and embrace something like the Stacks project for writing up the second-generation proof. It would attract people to finite group theory and help iron out kinks faster with more potential contributors, as well as be a fantastic open resource. I guess the groupprops wiki is meant to be a little something along these lines, but I don't see it approaching the majesty of what de Jong and collaborators have put together. – David Roberts Oct 15 '16 at 01:32
  • 4
    David, I'm just about to edit your answer to give an update, as Volume 8 is now available. I hope you don't mind -- I find it helpful that this page allows people to find out where things are up to. Please roll back the edits if you prefer. – Nick Gill Mar 27 '19 at 14:45
  • @Nick thanks! I don't mind at all. – David Roberts Mar 27 '19 at 21:12
  • As a sample of the contribution of Capdeboscq–Lyons–Solomon to the project (i.e. what will probably be in volume 9), see this talk by Capdeboscq from 2015: https://www.youtube.com/watch?v=rSAKSP1C1-E – David Roberts Nov 13 '19 at 22:46
  • Note to myself: volumes 1,2,3 cover Part I. Volume 4 is (the start of?) Part II. Volumes 5,7,8 cover "Part III: the generic case" (I think all of it). Volume 6 covers "Part IV: The Special Odd Case". – David Roberts Jul 24 '20 at 05:47
  • 2
    Another note to self: volumes 1–9 fill 3941 pages, and there are at least two more volumes with specific plans, if not three, given the comment about volume 9. Then there is the "e(G)=3" problem, which is at least one more volume, but quite likely more. – David Roberts Feb 25 '21 at 06:26
  • 1
    @DavidRoberts Back in 1995, Barry Cipra wrote an article in Science entitled, "At math meetings, enormous theorem eclipses Fermat." At that time, the estimated length of the "second-generation proof" was between 3000 and 5000 pages. If we include Aschbacher and Smith, which seems fair to me, then the current estimated length would seem to be closer to 7000 pages. – Timothy Chow Feb 25 '21 at 14:29
  • 6
    Hugo de Garis' blog is disgusting. It makes me feel ashamed to see something as beautiful as the CFSG be connected with the other material on his blog (sexism, bigotry, and anti-semitism, to name a few.). – Carl-Fredrik Nyberg Brodda Sep 21 '21 at 18:10
  • 3
    @Carl-FredrikNybergBrodda ugh. I was alread getting weird vibes from just the start of the posting I linked to, and I didn't dig further. But I'm loath to supply the email without some sort of provenance. – David Roberts Sep 21 '21 at 22:59
  • 3
    @DavidRoberts Yes, to be clear I think you’re right to put the link there, I just think it’s worth pointing out. His Wikipedia page is awful, too. – Carl-Fredrik Nyberg Brodda Sep 21 '21 at 23:19
  • 1
    @Carl-FredrikNybergBrodda At the risk of derailing the comments (in which case I suppose there is always a chat) I find the page that's linked to and the wikipedia entry painting a rather pathetic picture. Which is not an excuse for the views, but this is not some highly influential mathematician (and AFAICT not a highly influential scientist). We can't police who reads our theorems – Yemon Choi Sep 22 '21 at 04:05
  • 1
    Ron Solomon is due to give an update next week on this subject: http://mca.nsu.ru/kourovkaforum/ – Nick Gill Feb 24 '22 at 15:58
  • @NickGill oh, awesome! Especially the hints to avenues for improvements/optimisation for the CFSG, which, hopefully, will help to enliven the field. – David Roberts Feb 24 '22 at 23:40
  • @NickGill it appears the talk didn't happen? At the least, the last talk I find listed before the forum stopped operating (possibly related to current activities by the Russian government) is the 17th Feb. – David Roberts Apr 06 '22 at 07:44
  • @DavidRoberts, yes, I guess you're right... That's a shame.... And, yes, it seems likely that this cancellation is a result of the war in Ukraine. – Nick Gill Apr 07 '22 at 10:01
  • There was a talk a month ago: Remarks on CFSG and CGLSS (Richard Lyons, April 19, 2023) https://sites.math.rutgers.edu/~weibel/algebra.seminar.html#41923 Alas, I have no idea about its content. – colt_browning May 14 '23 at 20:33
  • @colt_browning thanks. There was also a pair of talks relatively recently (but months before that one—maybe even last year), with the second one given by Capdeboscq, that gave a detailed examination of what is happening about now, and what has been one. I've watched them recently, and meant to give a link and a brief summary, so thanks for the prod. – David Roberts May 14 '23 at 23:20
  • @DavidRoberts My guess is that the "12 volumes" estimate assumes that Stroth's contribution will be a single volume, whereas Solomon's estimate of "13 volumes" assumes that it will be split into two volumes. Also, the estimate of 5,000 pages must exclude Aschbacher-Smith; even so, 5,000 pages sounds like a lowball estimate to me. – Timothy Chow Jun 10 '23 at 03:18
  • @DavidRoberts I don't understand what you mean by "11+2+1 (or 12+2+1)". Don't you mean "10+1+1 (or 10+2+1)"? Or maybe "11+1 (or 12+1)"? – Timothy Chow Jul 07 '23 at 11:59
  • 1
    @TimothyChow (C)GLS (11 or 12) + Aschbacher–Smith (2) + Stroth (1). Lyons said Stroth's work will be the (singular) last volume, and since that manuscript is complete, I think that's an informed judgement. It's volume 11 that might spill over into 12. – David Roberts Jul 07 '23 at 12:18
  • Maybe Inna Capdeboscq's expertise that was used for volumes 9 and (I presume) 10 is orthogonal to what is needed for volume 11, but it's sad to see the team to go back down to two. – David Roberts Jul 07 '23 at 12:31
  • I think, from consulting the various published volumes, that vol 10 finishes off the proof of Theorem $\mathcal{C}_6$, and the last remaining row of the Classification Grid, Theorem $\mathcal{C}_4$, seems to split into two cases: $e(G)\leq 2$, which is covered by Aschbacher–Smith, and $e(G)= 3$. Then there's one more row of the Uniqueness Grid, to be covered by Stroth's volume, and the last volume (or two) by Lyons–Solomon. – David Roberts Aug 21 '23 at 07:37
  • 1
    Meanwhile, volume 10 is available for pre-order: https://bookstore.ams.org/surv-40-10 – colt_browning Oct 08 '23 at 23:00
  • @colt_browning yay!! – David Roberts Oct 08 '23 at 23:02
  • @DavidRoberts If you're starting to count pages, then I'd argue for including the page counts of the "background results." These are mostly listed in Volume 1, but a few additional background results are listed in the introductory pages of Volumes 8 and 9 (and, apparently, in Chapter 16 of Volume 10). – Timothy Chow Oct 13 '23 at 17:16
  • @TimothyChow the Aschbacher–Smith volumes are listed as background knowledge :-) I'd have to really go digging to find all the background stuff and how long it is. One day I will. Maybe after vol 10 comes out, so I can see what's listed in it – David Roberts Oct 13 '23 at 22:18
  • 1
    @DavidRoberts Oh yes, for sure. I'm just saying that I expect one can get another thousand pages or more out of the other background results. Feit-Thompson is 255 pages right off the bat. Volume 1 mentioned several "general group theory" books, and later volumes added two more books by Aschbacher. – Timothy Chow Oct 13 '23 at 22:24
  • 1
    @DavidRoberts Volume 10 is already available. Our library just got a copy. – Timothy Chow Oct 27 '23 at 10:41
  • @TimothyChow cool! Do you know if there it has any detailed discussion about the path forward/what is left for future volumes? – David Roberts Oct 27 '23 at 12:00
  • @TimothyChow Oh, I see the preface https://www.ams.org/bookstore/pspdf/surv-40-10-pref.pdf does indeed have a good discussion of what the books cover so far, and what is left: In summary, the theorems to be proved in future volumes are Case B of Theorem $\mathcal{C}_4^*$ and Odd Uniqueness theorems. The end is in sight! – David Roberts Oct 27 '23 at 12:13
  • Note to self: I asked Richard Lyons, and Stroth's manuscript does fit into the slot as needed by the adjusted theorem $\mathcal{C}^*_4$, not just the original version $\mathcal{C}_4$. – David Roberts Mar 05 '24 at 00:45
29

With respect to the second generation proof you can get an answer `from the horse's mouth' if you like: Ron Solomon gave an update on the program at BIRS recently and a video of his talk is here.

He starts the talk by comparing its progress to `the receding of the glaciers'! In reality, though, they've made very significant headway into the later volumes. (In particular he mentions, around 9:30, that volume 7 is in preparation, and it is mainly this volume that he's discussing in the talk.)

Nick Gill
  • 11,161
  • 39
  • 70