Re: ordered pairs/n-tuples as collections of sets



Elotemuygrande wrote:

Suppose you define a 2-tuple this way, which I just found in a book.
T=(a,b)={{a}, {a,b}}

Is there a way to express T_1 and T_2, the first and second elements of the tuple in terms of standard operations on sets like intersection and union and such instead of the gibberish found on http://en.wikipedia.org/wiki/Ordered_pair ? Yes, the gibberish makes sense to me, but it doesn't feel right saying it that way. The ideas I've come up with so far break down, especially when a=b. The book I'm looking through just says thank God we're not doing it this way in this book and no detailed explanation. I'm also thinking about how to define n-tuples without nasty nested pairs.

Any links or ideas appreciated, this is just arm-chair math thinking.


William Elliot gave some lovely definitions for T1 and T2 in terms of unions and intersections. A much uglier one:
Note that an ordered pair p has one or two elements. Exactly one of these must be a singleton. If p has two elements, one of these must be a doubleton whereof the singleton is a subset. T1(p) is the element of the singleton. If p is a singleton, T2(p) = T1(p). Otherwise, T2(p) is the sole element of the doubleton less than the singleton (set difference).


While nested pairs is one way to define n-tuples, it is more common to define an n-tuple as a function with domain n = {0, 1, 2,..., n-1}. You can't do that immediately for ordered pairs, since the definition of function uses ordered pairs itself. Thus, formally, ordered pairs and 2-tuples are different things, but there is a natural correspondence.

I find the Wikipedia article quite clear, but I already know this stuff. It is not necessarily a good source for someone learning these concepts for the first time. I strongly urge you to pick up a copy of Halmos's Naive Set Theory. You will find a better explanation of such things there, in less compact form, than anywhere on the 'net. I feel all students of mathematics should read it.

Best regards,

--
Stephen J. Herschkorn                        sjherschko@xxxxxxxxxxxx
Math Tutor in Central New Jersey and Manhattan

.



Relevant Pages

  • Re: Zuhairs ordered pairs
    ... The characteristic of ordered pairs: ... So we only have possibility 1) which leads to a=c and b=d. ... Now lets prove that z-ordered pairs are never singletons. ... Thus is never singleton. ...
    (sci.logic)
  • Re: intersection
    ... > The Hyperspec says the following about intersection: ... > "The intersection operation is described as follows. ... > ordered pairs consisting of one element from list-1 and one element ... hence CLisp (and Common Lisp) are right in returning what they return. ...
    (comp.lang.lisp)
  • Re: Zuhairs ordered pairs
    ... light of your response that you wished to avoid ordered pairs being ... i.e a pair that always have the cardinality of 2 for all x and y. ... singleton is no worse a property for a representation of pairs to have ... simplification of Wiener's pair, and the most complicated pair is the ...
    (sci.logic)
  • Re: Zuhairs ordered pairs
    ... of ordered pairs and the property of being non singleton. ... Which is definitely simpler than the one the poster G.Frege has ... It is even easier to prove than the standard Kuratowski pairs ...
    (sci.logic)
  • Re: intersection
    ... > Common Lisp. ... > "The intersection operation is described as follows. ... > ordered pairs consisting of one element from list-1 and one element ... "If one of the lists contains duplicate elements, ...
    (comp.lang.lisp)