# Inaccessible cardinals and Andrew Wiles's proof

In a recent issue of New Scientist (16 Aug 2010), I was surprised to read that a part of Wiles' proof of Taniyama-Shimura conjecture relies on inaccessible cardinals.

Here's the relevant bit from the article: "Large cardinals have been studied by logicians for a century, but their intangibility means they have seldom featured in mainstream mathematics. A notable exception is the most celebrated result of recent years, the proof of Fermat's last theorem by the British mathematician Andrew Wiles in 1994............... To complete his proof, Wiles assumed the existence of a type of large cardinal known as an inaccessible cardinal, technically overstepping the bounds of conventional arithmetic"

Is this true ? If so, could someone please outline how they are used ?

-
No, completely false. Perhaps the author (whose mathematical training seems to be far from number theory) has misunderstood or misheard descriptions of parts of the argument. –  BCnrd Aug 16 '10 at 11:12
Does any part of Wiles' proof or general results he may rely on make use of Grothendieck universes? The existence of these is equivalent to the existence of an inaccessible cardinal. –  Joel David Hamkins Aug 16 '10 at 11:20
Actually, Cosmonut misquoted the article by leaving out the rest of the statement: "But there is a general consensus among mathematicians that this was just a convenient short cut rather than a logical necessity. With a little work, Wiles's proof should be translatable into Peano arithmetic or some slight extension of it." –  JS Milne Aug 16 '10 at 13:41
The Stack Project develops a huge amount of Grothendieck style mathematics, including a lot of etale cohomology, using only ZFC (specifically, NOT using universes). If anyone has any doubt that this can be done, I suggest that they look at it. –  JS Milne Aug 16 '10 at 18:48
Dear Pierre: if one wants to write a treatise on a general theory of cohomology on all topoi, including operations like sheafification, enough injectives, derived categories, and Ext-sheaves, there needs to be a way to control the "size" of coverings which arise in these constructions (to replace the implicit role of "power set" for ordinary topology spaces). The universe stuff takes care of such matters in an elegant way, so one can focus attention on the more central aspects of the theory. –  BCnrd Aug 17 '10 at 5:43

The basic contention here is that Wiles' work uses cohomology of sheaves on certain Grothendieck topologies, the general theory of which was first developed in Grothendieck's SGAIV and which requires the existence of an uncountable Grothendieck universe. It has since been clarified that the existence of such a thing is equivalent to the existence of an inaccessible cardinal, and that this existence -- or even the consistency of the existence of an inaccessible cardinal! -- cannot be proved from ZFC alone, assuming that ZFC is consistent.

Many working arithmetic and algebraic geometers however take it as an article of faith that in any use of Grothendieck cohomology theories to solve a "reasonable problem", the appeal to the universe axiom can be bypassed. Doubtless this faith is predicated on abetted by the fact that most arithmetic and algebraic geometers (present company included) are not really conversant or willing to wade into the intricacies of set theory. I have not really thought about such things myself so have no independent opinion, but I have heard from one or two mathematicians that I respect that removing this set-theoretic edifice is not as straightforward as one might think. (Added: here I meant removing it from general constructions, not just from applications to some particular number-theoretic result. And I wasn't thinking solely about the small etale site -- see e.g. the comments on crystalline stuff below.)

Here is an article which gives more details on the matter.

http://www.cwru.edu/artsci/phil/Proving_FLT.pdf

Note that I do not necessarily endorse the claims in this article, although I think there is something to the idea that written work by number theorists and algebraic geometers usually does not discuss what set-theoretic assumptions are necessary for the results to hold, so that when a generic mathematician tries to trace this back through standard references, there may seem to be at least an apparent dependency on Grothendieck universes.

P.S.: If a mathematician of the caliber of Y.I. Manin made a point of asking in public whether the proof of the Weil conjectures depends in some essential way on inaccessible cardinals, is this not a sign that "Of course not; don't be stupid" may not be the most helpful reply?

-
For etale cohomology all of that universe stuff is entirely irrelevant. I say this not as an "article of faith", but because I've read all of the proofs of the theorems of etale cohomology. Perhaps if one wants to make a super-general theory of cohomology for "all" topoi there are these problems, but if one only cares about more "real" examples such as etale cohomology then there are no issues. –  BCnrd Aug 16 '10 at 11:38
Thanks Pete and others. This is very interesting. –  Cosmonut Aug 16 '10 at 11:56
I completely agree with BCnrd opinions here but as far as I know there is no published complete proof of the relevant results that bypasses SGA IV so in that sense it is not an established fact. It is perhaps also interesting to note that (as far as I can remember) Manin poses the question on whether Deligne's proof of the Riemann hypotheses can be made independent of inaccessible cardinals in the problem list in Mathematical Developments Arising from Hilbert Problems: Proceedings (Proceedings of Symposia in Pure Mathematics, V. 28 parts 1 & 2). –  Torsten Ekedahl Aug 16 '10 at 12:06
Section 1 of link shows the author "reasoning" about math he does not understand, and rests on the bogus argument that since there is one general reference on cohomology by Gr. written with universes, appeal to any cohomological results of Gr. has the same dependence. The result of Tate which concerns the author was inspired by Gr. duality but is just a piece of pure comm. algebra. Deligne-Rap. use Gr. duality, but Gr. duality requires no universes. (Improvements in the method removed Gr. duality, by the way.) Proofs exist to impart understanding; the author should have read some of them. –  BCnrd Aug 16 '10 at 12:12
Dear Torsten: For the crystalline theory, the issue is to control the "size" of the nil-thickenings one allows to consider in the sheaf theory. In the etale case, by restricting to small etale site on a fixed scheme (sheaves determined by affines etale over the base) we see that Hom between sheaves is a set (so Tohoku proof of enough inj's via transfinite induction on Hom sets can be applied). For crystalline case, seek analogous "small" site for same needs (e.g., want Ext's) while also retaining a form of the crucial ring fiber product trick early in the theory. Perhaps another 550 pts... –  BCnrd Aug 16 '10 at 13:57

I'm writing a new community wiki answer because it seems to me the consensus in the comments is that the accepted answer doesn't really tell the right story, and since this is something that pops up all the time it'd be good to have a single place to point people without making them read through all the comments. Please improve my answer.

In the most naive sense Wiles proof does depend on existence of Grothendieck universes (and thus on existence of inaccessible cardinals). By this I mean if you took every reference in Wiles proof and read the first published proof of that fact you'd certainly end up somewhere in SGA where, due to Grothendieck's love of generalization, you'd find universes popping up.

However, this certainly doesn't mean the proof really uses universes. It's widely believed (though for some people this belief may not come from much direct evidence) that in any practical situation you don't actually need universes. However, there are some concrete situations (BCnrd mentions some involving sheafification on the crystalline site) where it's not necessarily known how to eliminate the use of universes.

As a result, in order to figure out if Wiles's proof uses universes, or whether it's relatively easy to avoid them, you'd need to either read the proofs yourself or find someone who was both deeply familiar with the details of the proof, and someone who cares a lot about details. One person who comes quickly to mind is BCnrd. BCnrd was one of the mathematicians who proved the Modularity Theorem, which showed that all elliptic curves over $\mathbb{Q}$ are modular. This is a strengthening of Taylor and Wiles' result, which applied only to semi-stable elliptic curves, and the proof involved understanding and building on Taylor and Wiles' work. BCnrd is also famous for his attention to detail and for consulting underlying foundational sources; he is the author of a book dedicated to simplifying and correcting the presentation of Grothendieck duality in Hartshorne's book Residues and Duality.

As explained in the comments to Pete's answer, BCnrd says there's really no issue at all in Wiles's proof. All of the specific things that Wiles uses stay far away from any of the difficult issues where you might be worried about needing to invoke universes.

-
Edited to explain BCnrd's qualifications to those who don't already know them. –  David Speyer Aug 16 '10 at 16:05
Note to BCnrd: It occurs to me that my edits make it trivial for someone to deduce your identity. I assume you have no intention of being genuinely anonymous but, if this bothers you, let me know and I will remove them. –  David Speyer Aug 16 '10 at 16:06
Dear David: No problem. Dear Noah: Nice summary. Galois repns in pf of FLT are weight 2, built via Tate mods of Jacobians (no etale cohom). The part which gets some worked up about universes is Grothendieck duality (eliminated in the subsequent generalizations and improvements of the method, FWIW), but nothing in pfs of Gr. duality leads to universes either. Someone who reads the first proofs of original refs, as you suggest (bad idea: there's been progress!), won't run into universes. Too dramatic to focus on FLT. Would be better if those who speculate on certain proofs read them first. :) –  BCnrd Aug 16 '10 at 16:44
I agree with all of this. I think McLarty's point of view makes some sense when FLT is read as a synecdoche for "modern arithmetic geometry using Grothendieck cohomologies". I am familiar with the foundations of etale cohomology and know that universes are absolutely not required there. However, for some other sites, things are less obviously benign. For an example -- which must be close to the simplest possible example -- of set-theoretic difficulties related to nilpotent thickenings, see Section 6 of math.uga.edu/~pete/apoints11.pdf. –  Pete L. Clark Aug 16 '10 at 17:42

Can I draw attention to the continuation of that quotation? "But there is a general consensus among mathematicians that this was just a convenient short cut rather than a logical necessity. With a little work, Wiles's proof should be translatable into Peano arithmetic or some slight extension of it."

For the record, McLarty's paper in the Bulletin of Symbolic Logic was indeed the source for my claim that Wiles used an inaccessible cardinal (via Universes). If anyone fancies debating the exact definition of 'using' an axiom without 'needing' it, and whether or not that definition applies in the current case, we should probably do it somewhere else (Pedantry Overflow?). But I explicitly opposed the claim that the proof 'needed' large cardinal assumptions, so I don't propose to be too apologetic.

Anyway, it's an interesting question, and I'd be pleased if my article goes some way to bringing an answer into the public domain (even if I do get my head shot off in the process ;) ).

(Incidentally I would rather have left this as a comment than an answer, but don't have enough (any) reputation points. If anyone with superpowers wants to move it, please go ahead.)

-
Incidentally, while I strongly agree that universes are not logically necessary, it not as clear to me exactly what axiom system is necessary for Wiles's proof. (I.e., how "slight [an] extension of" Peano arithmetic, if any, is actually needed.) The reason is that Wiles, like all modern number theorists, argues in a thoroughly second order way (i.e. works with the standard model of the natural numbers, argues with subsets of such, and so on), and I don't how strong an axiom scheme is needed to translate such arguments into first order arithmetic. –  Emerton Aug 16 '10 at 22:16
I now seem to have magically acquired enough reputation to comment... Very recently I have read that Angus Macintyre has announced a proof of FLT in PA. The only reference I can provide is this: cameroncounts.wordpress.com/2010/01/07/mathematics-and-logic-2 –  Richard Elwes Aug 16 '10 at 23:30
Richard, it is more accurate to say Macintyre announced that the proof by Wiles can be developed in PA; the way you wrote it suggests there is some other approach in PA. The blog link says the proof by Wiles follows from his work "on the equivalence of modular forms and elliptic functions". Oops! –  KConrad Aug 17 '10 at 3:32
Ok! Try again... Macintyre has reportedly announced a proof (which I would expect to be non-trivial) that Wiles' proof can be translated into PA. In the intro to his forthcoming book, Boolean Relation Theory and Incompleteneess, Harvey Friedman conjectures that it should be proveable in an elementary fragment $I \Sigma_0(exp)$. This is an instance of his conjecture that "Every theorem published in the Annals of Mathematics whose statement involves only finitary mathematical objects (i.e., what logicians call an arithmetical statement) can be proved in elementary arithmetic." –  Richard Elwes Aug 17 '10 at 7:05
I heard Angus Macintyre talk about it fairly recently. The situation is a little unusual, as to write out a proof fully would entail multiplying the (already considerable) length of Wiles' proof by 5 (or 10 or something), to reduce every single step to PA. Macintyre does not intend to do this in full completeness, but he is continuing to prepare a manuscript to show that the key pieces of higher order machinery used in the Wiles' proof can be translated into PA. The focus is on the central modularity theorem, which he will show is effectively $\Pi^0_1$, and provable in PA. –  Richard Elwes Mar 21 '11 at 12:03

There was some discussion about this on the FOM mailing list: see for example http://cs.nyu.edu/pipermail/fom/1999-April/002983.html, where someone claimed Wiles needed in accessible cardinals, and Harvey Friedman essentially told them to stop being stupid. (Friedman is the guy who created the subject of reverse mathematics, which studies what axioms are necessary for any given result.)

An analogy would be the claim that classifying groups of order 4 needs Grothendieck universes, because groups of order 4 form a proper class, so isomorphism classes of groups of order 4 form a 2-class of classes. This is obviously silly: it is trivial to restate the classification of groups of order 4 without using proper classes, but this makes the statement slightly more complicated: you have to talk about groups that are hereditarily finite sets or something like that, which is just an irrelevant complication. The use of Grothendieck universes is similar: for example, the collection of all etale spaces over a scheme is a proper class so in some sense uses universes to construct it, but is equivalent to a much smaller set so the use of universes is not necessary.

-
The final example above highlights the distinction between working with sheaves vs. presheaves. If one wants to consider presheaves on "all" schemes etale over a base, then the set theory issues emerge (e.g., Hom of presheaves isn't a set...). But we ultimately care about the sheaf theory (= the topos), for which much smaller sites compute the topos. That's why there is no problem with the concept of an etale sheaf even on "all" schemes etale over a fixed scheme. I think Milne's book notes an example of Waterhouse that illustrates the problems that arise if one is too cavalier with presheaves. –  BCnrd Aug 16 '10 at 14:11
An interesting historical data point on the other side of this issue is that when Solovay proved the consistency with ZF+DC that every set of reals is Lebesgue measurable, he used an inaccessible cardinal in the proof, and many researchers at that time reportedly viewed this aspect as a removable defect in the proof. But it was later proved by Shelah that an inaccessible cardinal was necessary. –  Joel David Hamkins Aug 16 '10 at 14:51
"According to the expert, the real reason that the obvious eliminability of Universes is not frequently pointed out in the literature is just that it is so obvious. But the expert says that if there was ever a problem, then a lot of attention would then be paid to it!" –  Junkie Aug 16 '10 at 16:26

At the suggestion of François G. Dorais I am moving this answer from Recent claim that inaccessibles are inconsistent with ZF to here. The text below was written before I read the above, very well-informed answers/comments, so there is considerable overlap.

This is just a quick answer regarding FLT, mentioned by the OP.

Colin McLarty is working on showing a small part of Friedman's grand conjecture, namely that the Fermat-Wiles(-Taylor?) theorem is provable in a weak system of arithmetic. Since originally the semi-stable case of Taniyama-Shimura-Weil that Wiles proved (not to mention the work by others such as Frey, Serre, Ribet to get to that point) required large parts of algebraic and arithmetic geometry developed by the Grothendieck school, which among other things uses sheaves, cohomology and so on, and so a priori requires some foundational care. Universes (~inaccessible cardinals) were introduced to take care of the problem of e.g. forming categories of sheaves on categories of sheaves on a site.

However, one can use a version of set theory much weaker than ZFC+Universe(s), indeed a fair bit weaker than ZFC, as McLarty has shown, and still get pretty much all of EGA/SGA (this is based on general arguments, he hasn't sat down and worked through it all). However, the arithmetic geometry needed for FLT [edit: and indeed a lot of number theory] really only needs to consider countable sites, rather than generic small sites, and so one really only needs to assume much weaker assumptions about infinite objects.

A more recent talk (July 2011 - slides not publicly available as far as I know) by McLarty had the statement that derived functor cohomology is a finitely axiomatised first-order theory, and so not really the complicated logical beast it appears to be.

-
Thanks for moving this answer; I think this is the best place for it. –  François G. Dorais Aug 19 '11 at 3:46