Exercise 17.5

Let X be an ordered set in the order topology. Show that (a,b)¯ [a,b]. Under what conditions does equality hold?

Answers

Proof. Since (a,b) [a,b] closed, and by the definition of closure,

(a,b)¯ = K(a,b)closedK [a,b].

(a,b)¯ = [a,b]a,b (a,b)¯ any basis elements A a,B b intersect (a,b) by Theorem 17.5(b). We claim that this is equivalent to the fact that there is no immediate successor α of a and no immediate predecessor β of b. If either are the case, say for a, then choosing A with upper bound α would not intersect (a,b), and so equality doesn’t hold since a(a,b)¯; in the other direction, if neither are the case, we see that, say for a, the upper bound of A, α would be such that (a,α) is non-empty, and so A (a,b), satisfying the condition for Theorem 17.5(b). The same argument applies when considering b and β, and so our claim holds. □

User profile picture
2021-12-21 18:29
Comments

Proof. First, the closed interval [a,b] is closed (hence why it is called such!) because clearly, its complement is

X [a,b] = (,a) (b,)

and we know that open rays are always open so their union is as well. Clearly also [a,b] contains (a,b). Hence [a,b] is a closed set containing (a,b). Since (a,b)¯ is defined as the intersection of closed sets that contain (a,b) clearly we have that (a,b)¯ [a,b] as desired. □

The conditions required for equality are such that [a,b] is also a subset of (a,b)¯ and, in particular, both a and b must be in (a,b)¯. Since clearly a,b(a,b), it has to be that they are both limit points of (a,b). This is equivalent to the condition that a has no immediate successor and b has no immediate predecessor. We show only the first of these since the second is analogous.

Proof. (⇒) We show the contrapositive of this. So suppose that a does have an immediate successor c. Then the open ray (,c) is an open set that contains a but does not intersect (a,b). This is easy to see, because if they did intersect, there would be an x (a,b) where also x (,c). From these, it follows that a < x < c, which contradicts the fact that c is the immediate successor of a. Hence by definition, a is not a limit point of (a,b).

(⇐) Suppose that a is not a limit point of (a,b). Then there is an open set U containing a that does not intersect (a,b). From this, it follows that there is a basis element B containing a such that B U, and thus B also cannot intersect (a,b) (as, if it did, then so would U). Suppose that B is the open interval (c,d) so that c < a < d. It also must be that d < b for otherwise, for any element of x of (a,b), we would have c < a < x < b d so that x (c,d) = B and B and (a,b) would not be disjoint. We claim that d is the immediate successor of a. If this is not the case then there would be an x such that c < a < x < d and hence x (c,d) = B. Also a < x < d < b so that also x (a,b). Therefore B and (a,b) would not be disjoint. Similar arguments can be made if B are other types of basis element in the order topology. (Actually, B cannot be of the form (e,f] for the largest element f of X since then any element of (a,b) would also be in B and they would not be disjoint.) □

It is also worth noting that the Hausdorff axiom (and therefore also the T1 axiom since it is implied by the Hausdorff axiom) is not sufficient for general equivalence of [a,b] and (a,b)¯. For example, the order topology on results in the discrete topology so that every subset is both open and closed. Thus for any pair x1,x2 in , the sets {x1} and {x2} are neighborhoods of x1 and x2, respectively, that are disjoint. This shows that this topology is a Hausdorff space. However, the fact that a has an immediate successor in + is sufficient to show that [a,b](a,b)¯ per what was just shown above.

User profile picture
2019-12-01 00:00
Comments