I am new to both category & set theory. From what I have learned of set theory so far:

  • Products between two(+) sets, A×B, create a "grid" of elements which completely covers all combinations of elements a ∈ A & b ∈ B (a,b) or (b,a) depending on which way round they were composed.
  • Co-products between two(+) sets, A∪B, are essentially disjoint unions. Aka, all elements of sets A and B are put into a single set, C, and paired with a discriminator such that any overlaps (e.g. a=b) are still kept and distinguished.
  • (Bi)nary Relations are generalisations of products and co-products in that they represent distinct pairs. However, they do not guarantee every element pair, like products do, nor even that all elements in a given parent set will be used, like co-products.

Now, my question is that products and co-products have their respective symbolic representations in category theory (A×B & A∪B respectively), but I have yet to see how to define a generic (bi)nary relation in notation beyond calling it some arbitrary name (as with any other generic set).

How do I correctly denote a set as being a binary relation between two sets, A & B, in both set theory and category theory?

Thanks in advance.

  • $\begingroup$ Given that there are lots of different relations between any two sets, how could you expect to refer to a generic one other than by calling it some arbitrary name? $\endgroup$ – Eric Wofsey Jul 12 '17 at 20:17
  • $\begingroup$ @EricWofsey For instance via the use of subscripts and possibly some symbol I have yet to learn about? As it feels like a prevalent type of set I figured I've just been looking in the wrong places for it. $\endgroup$ – user3303504 Jul 12 '17 at 20:32
  • $\begingroup$ I wouldn't call them generalizations of products/coproducts by any stretch; a random relation need not be any kind of universal or weakly universal structure. They have a categorical description as subobjects of products, or as the morphisms of an allegory; but not as this purported generalization. $\endgroup$ – Malice Vidrine Jul 13 '17 at 5:27
  • $\begingroup$ @MaliceVidrine Hmm, not heard of allegories before, shall be the next topic on my to-learn list. Thanks for correcting me about the generalisations statement. Without looking into the category of allegories, I am assuming that there isn't any special notation used defining such binary relations? If not then I'll select an answer saying as much. Thanks for your help. $\endgroup$ – user3303504 Jul 13 '17 at 15:16
  • $\begingroup$ @user3303504 - Actually, now that I've had time to think about it, I think I have seen special notation for the morphisms of an allegory (or at least, morphisms in $\mathbf{Rel}$) before; I don't think it's standardized, but let me see if I can dig it up... $\endgroup$ – Malice Vidrine Jul 13 '17 at 16:29

I have not seen a special notation for relations as objects/jointly monic spans, but I have run across a couple of notations for relations as morphisms in an allegory. P.T. Johnstone uses $A \looparrowright B$ in parts of Sketches of an Elephant (writing things like $\phi:A\looparrowright B$ to indicate that $\phi$ is some relation on $A,B$), and Paul Taylor has similarly used a symbol that is like $A\rightleftharpoons B$, except that there was no gap; the harpoons shared the same horizontal line. The latter doesn't seem to be a standard Latex symbol, though.

Neither of these are really standard, at the moment, so if you run across them in other contexts it's a good idea to check what the author intends.

  • 1
    $\begingroup$ Thanks for the answer, I shall have fun looking into allegories later =) $\endgroup$ – user3303504 Jul 14 '17 at 13:36
  • $\begingroup$ They're pretty neat! The relevant chapter of The Elephant is pretty interesting, especially if you have an interest in topos theory at all. $\endgroup$ – Malice Vidrine Jul 14 '17 at 20:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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