8
$\begingroup$

3-fold tensor products are usually presented in terms of the natural isomorphism of iterated tensor porducts. Where is there a treatment of 3-fold tensor products without reference to 2-fold?

$\endgroup$
8
  • 2
    $\begingroup$ And the other one is "multilinear algebra". Vote to close as too localized. $\endgroup$ Commented Oct 24, 2012 at 20:15
  • 5
    $\begingroup$ I think it's a good reference-request question. Most linear algebra texts completely skip this definition (and I'm not talking of those who skip tensor products altogether), but papers use it (think of $\bigotimes\limits_{s\in S}V_s$ where $S$ is an infinite, or finite but not canonically ordered, set). But I think the answer should be: "Yes, you should write it up yourself and put the notes online; people will thank you for that." $\endgroup$ Commented Oct 24, 2012 at 20:47
  • 8
    $\begingroup$ Assuming Professor Stasheff is referring to monoidal products, and not just those pertaining to linear algebra, I wouldn't say this reduces to multilinear algebra. I think Qiaochu's answer is more appropriate; there is a circle of ideas here involving unbiased monoidal categories and multicategories, explored in Tom Leinster's book Higher Operads, Higher Categories, which would be a good reference. (By the way, you do realize, Martin, that the work of James D. Stasheff had a lot to do with the very invention of monoidal categories? Show a little respect!) $\endgroup$ Commented Oct 24, 2012 at 23:42
  • 1
    $\begingroup$ @Todd: Sorry for not paying attention to the history behind the poster. I only looked at the question, which is not really precise and elaborate for MO-standards. If the poster was "NoIdeaC3PO", then this question would have already been closed. By the way, the assumption that the question refers to monoidal categories cannot be read off from the question. @Darij: Probably you mean Linear Algebra texts for students in the first semesters? Every complete treatment of Linear Algebra also contains Multilinear Algebra, and there arbitrary tensor products are studied. $\endgroup$ Commented Oct 25, 2012 at 7:43
  • 1
    $\begingroup$ Thanks, Todd - Matho didn't alert me to these answers. And YES Jim is the proper form of address. The question was deliberately ambiguous. I will look into the unbiased monoidal description. A colleague had a particular case in mind; I will look it up and describe more precisely. Meanwhile, how about this: Suppose I have rings R and S AND MODULES A, B and C such that A\otime_R B and B \otimes _S are defined, when does it make sense to talk about a triple tensor product over what? $\endgroup$ Commented Oct 25, 2012 at 13:47

0

You must log in to answer this question.

Browse other questions tagged .