# Tensor products as isomorphic functors in category theory

+ 4 like - 0 dislike
1391 views

An earlier question that I posed sought to define a category with a set of quantum channels as arrows and the C$^{*}$-algebra that these channels map from and to as the object. So, for example, my channels might be the set of all channels with 2 x 2 Kraus operators mapping some C$^{*}$-algebra to itself. Since the channels are always $d \times d$ square matrices, I'll refer to them as having dimension $d$. So far the answers to that question seem to have indicated that my formulation is correct.

Now suppose we have one such category whose channels have dimension 2 that I'll call Chan(2). Suppose we also have a category whose channels have dimension 4 that I'll call Chan(4). A tensor product of the channels in Chan(2) with themselves produces a channel that is in Chan(4), i.e. it is a mapping from Chan(2) to Chan(4). The tensor product is known to be a functor so this isn't unexpected.

But here's my question: is there a way to define an isomorphism between Chan(2) and Chan(4)? In other words, can the action of the tensor product be "undone," i.e. if the tensor product is the functor going from Chan(2) to Chan(4), is there a functor going from Chan(4) to Chan(2) and, can the two together define an isomorphism? If I make a larger category out of all these little categories, it seems like I could make the tensor product and anything that undoes it, arrows in the larger category and then I'd have my isomorphism. Can I do this and, if so, how does one undo a tensor product?

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Ian Durham
retagged Sep 23, 2014
Are you familiar with multicategories, because they seem relevant to what you're looking for.

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Harry Gindi
Also, about your question, the tensor product cannot be undone unless you tensor over Z, because this is the direct sum of commutative rings. By construction, the tensor product "cuts out" too much stuff to recover everything about the rings.

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Harry Gindi
What exactly is the use of category theory adding to the problem here?

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user j.c.
Harry: I am not familiar with multicategories. Is there a reference you could point me to or are they found somewhere in Saunders and MacLane? jc: This is a smaller part of a larger problem I'm working on related to a quantum extension of Birkhoff's theorem. A quantum extension already exists for 2-d and reversible channels. By wrapping a channel up inside a category (while still trying to maintain the structure somehow), I hope to be able to exploit the category theoretic definition of isomorphism which will eventually allow me to represent n channels by a combination of unitaries.

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Ian Durham
Ian: there's an OK wikipedia article on multicategries and, um, a reference there that you can follow up. They first appeared in a paper of Lambek in the late 1960s. They're not in Mac Lane (and incidentally, Saunders was Mac Lane's first name -- they're not two separate people).

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Tom Leinster
Tom: OMG, I can't believe I did that. I swear I really did know that his name was Saunders Mac Lane. I have read the book (or portions of it, anyway) and Steve Awodey studied under him (I think).

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Ian Durham
Ian: yes, he did!

This post imported from StackExchange MathOverflow at 2014-09-15 18:27 (UCT), posted by SE-user Tom Leinster

 Please use answers only to (at least partly) answer questions. To comment, discuss, or ask for clarification, leave a comment instead. To mask links under text, please type your text, highlight it, and click the "link" button. You can then enter your link URL. Please consult the FAQ for as to how to format your post. This is the answer box; if you want to write a comment instead, please use the 'add comment' button. Live preview (may slow down editor)   Preview Your name to display (optional): Email me at this address if my answer is selected or commented on: Privacy: Your email address will only be used for sending these notifications. Anti-spam verification: If you are a human please identify the position of the character covered by the symbol $\varnothing$ in the following word:p$\hbar$ysicsO$\varnothing$erflowThen drag the red bullet below over the corresponding character of our banner. When you drop it there, the bullet changes to green (on slow internet connections after a few seconds). Please complete the anti-spam verification