18
$\begingroup$

I find the following description of Artin motives in Wikipedia. Since these seem to be quite related to number theory, I am interested to learn more in that context. I request the experts available in MO to provide me a reference which can be more useful to me than this terse description. Any comments and explanations also will be helpful. I apologize for asking a seemingly basic question; but I find it impossible to wade through the numerous references available on motives.

Fix a field $k$ and consider the functor finite separable extensions $K$ of $k$ → finite sets with a (continuous) action of the absolute Galois group of $k$ which maps $K$ to the (finite) set of embeddings of $K$ into an algebraic closure of $k$. In Galois theory this functor is shown to be an equivalence of categories. Notice that fields are $0$-dimensional. Motives of this kind are called Artin motives. By $\mathbb Q$-linearizing the above objects, another way of expressing the above is to say that Artin motives are equivalent to finite $\mathbb Q$-vector spaces together with an action of the Galois group.

I have some idea of what are pure motives and mixed motives, in the context of algebraic varieties. What I exactly have in mind is to understand the modern statement of equivariant Tamagawa number conjecture. This would appear to be the simplest instance to keep in mind, if I go ahead.

$\endgroup$
2
  • 2
    $\begingroup$ I remember there was something on Artin motives in Yves Andre's book. $\endgroup$ Jan 15, 2010 at 23:15
  • 1
    $\begingroup$ Thanks, I will look into it(though I am French-handicapped). $\endgroup$
    – Anweshi
    Jan 15, 2010 at 23:19

2 Answers 2

6
$\begingroup$

André's book is the main reference for the "yoga" of motives. You'll find a description of Artin motives in the Voevodsky formalism in

Beilinson and Vologodsky - http://www.math.uiuc.edu/K-theory/0832/

Wildehaus - http://www.math.uiuc.edu/K-theory/0918/

From the tannakian view point, Artin motives are just representation of the usual Galois group. So, as motives, Artin motives are not that interesting. It's just the usual Galois theory of fields.

$\endgroup$
7
  • $\begingroup$ This was very helpful. So it is just a motivic picture of the usual Galois theory; just a different way of looking at a simpler thing? $\endgroup$
    – Anweshi
    Jan 15, 2010 at 23:43
  • 3
    $\begingroup$ Yes it is. Artin motives are motives of 0-dimensional varieties: the (products of) H^0(L/K) for finite extensions L/K. Grothendieck's vision of motives was as a universal cohomolgy theory but also as higher dimensional version of Galois theory. From the tannakian point of view, the inclusion of Artin motives in all motives corresponds to the fact that the absolute Galois group (seen as the pro-algebraic group limproj Sepc(Q[Gal(L/K)]^*)) is a quotient of the motivic Galois group of K. $\endgroup$
    – AFK
    Jan 16, 2010 at 0:56
  • $\begingroup$ Thanks for the explanation. So if I understood correctly, Artin motives introduce a geometric Galois theory, suitably integrated into a bigger picture according to the visions of Grothendieck. $\endgroup$
    – Anweshi
    Jan 16, 2010 at 6:25
  • 2
    $\begingroup$ To YBL: One small nitpick. It seems to me that 'permutation representation' is not good terminology here. As I understand it, this could refer either to a realization as a group of permutations or a linear representation naturally associated to such a realization. Of course many linear representations are not of this form. Importantly, no non-trivial irreducible representation is of this form. Since the category of motives is supposed to induce a decomposition of varieties into 'irreducible objects' in some suitable sense, it's important to draw this distinction. $\endgroup$ Jan 16, 2010 at 10:36
  • $\begingroup$ You're right this wasn't clear. I meant "permutative Q[G]-module" as in Beilinson and Vologodsky's notes par 2.4. $\endgroup$
    – AFK
    Jan 16, 2010 at 12:02
16
$\begingroup$

A motive is a chunk of a variety cut out by correspondences. (If you like, it is something of which we can take cohomology.)

Artin motives are what one gets by restricting to zero-dimensional varieties. If the ground field is algebraically closed then zero-dimensional varieties are simply finite unions of points, so there is not much to say; the only invariant is the number of points.

But if the ground field $K$ is not algebraically closed (but is perfect, e.g. char $0$, so that we can describe all finite extensions by Galois theory), then there are many interesting $0$-dimensional motives, and in fact the category of Artin motives (with coefficients in a field $F$ of characteristic $0$, say) is equal to the category of continuous representations of $Gal(\overline{K}/K)$ on $F$-vector spaces (where the $F$-vector spaces are given their discrete topoogy; in other words, the representation must factor through $Gal(E/K)$ for some finite extension $E$ of $K$).

Perhaps from a geometric perspective, these motives seem less interesting than others. On the other hand, number theoretically, they are very challenging to understand. The Artin conjecture about the holomorphicity of $L$-functions of Artin motives, which is the basic reciprocity conjecture regarding such motives, remains very wide open, with very few non-abelian cases known. (Of course, for representations with abelian image, these conjectures amount to class field theory, which is already quite non-trivial.)

$\endgroup$
7
  • $\begingroup$ Thanks, this was very helpful. As I mentioned, the aim was to get to the equivariant Tamagawa number conjecture. Are there any specific suggestions towards that? $\endgroup$
    – Anweshi
    Jan 16, 2010 at 2:50
  • $\begingroup$ The conjecture on Artin L-functions does not follow from Khare's work on Serre's conjectures? $\endgroup$
    – Anweshi
    Jan 16, 2010 at 3:07
  • 4
    $\begingroup$ The case of 2-dimensional odd representations, in the case $K = {\mathbb Q}$, follows from Khare--Wintenberger--Kisin. This is an example of one of the ``very few non-abelian cases'' that are known. $\endgroup$
    – Emerton
    Jan 16, 2010 at 4:02
  • $\begingroup$ @Emerton. Your answer was extremely helpful to me, and it was very enlightening to read it. However I am accepting YBL's answer since I had asked for references and he gave me some. I hope you don't mind. $\endgroup$
    – Anweshi
    Jan 16, 2010 at 20:16
  • 1
    $\begingroup$ @Emerton. Can we recover a number field from the knowledge of Artin motives attached to it? In general, is it possible to recover an algebraic variety from a motivic understanding, i. e., is it possible to get back to the original thing, if you know the motivic chunks as you called them? $\endgroup$
    – Anweshi
    Jan 17, 2010 at 4:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.