Let be a real function. Then, is continuous at if and only if for all , there exists a such that for all with . Our task today is to investigate the discontinuities of , i.e. the set of points at which is not continuous. Specifically, we want to answer two questions:
What type of set can be?
Given any set , is there a function that is discontinuous exactly on ?
We first need to know what it means for to be discontinuous at a point. By negating the definition of continuity, we can deduce that
is discontinuous at if and only if there exists an such that for all , we have for some with . (*)
The definition should make sense to you: it says that as we zoom in closer to , the function stops getting closer to at some point, instead staying outside a "barrier" around .
Given a set , the oscillation of on is
The geometric meaning of the oscillation is hopefully clear. Note that the oscillation is always nonnegative, and if , then . Thus, as "shrinks" toward , the oscillation decreases to some nonnegative limit. We can now define the oscillation of at by
If is how much "jumps" on the set , then is how much jumps at . The definition (*) above can be rewritten into
is discontinuous at if and only if there exists an such that for all , we have .
We can now conclude that is discontinuous at if and only if .
The preceding observation allows us to write as the set of for which . We need one final detail.
Lemma. For any , the set is open.
Proof. Given such that , there is a bounded interval containing such that . Consequently, is a subset of , since for any , we have .
With this lemma, we can finally prove a characterization of .
Theorem. is the countable union of closed sets in .
Proof. The first part of this characterization is easy enough:
The complement of is , which is open. Thus, is a countable union of closed sets.
This result holds in a general metric space. In a topological space, a countable union of closed sets is called an set.1 We now know that is an set, but is every set a for some ?
The answer to the question above is "yes": any set in is the set of discontinuities for some function . We shall construct by trying to follow the characterization in our proof of the major theorem in the previous section.
Theorem. If is an set in , there is some such that .
Proof. Let , where each is closed. Let and
Note that the are disjoint, , and . Each has a countable dense subset (see later). We now define on as follows:
The function is well-defined since and the are mutually disjoint. We now consider some .
Suppose , then for some .
In any case, we can always find a point in this interval such that . Altogether, if , then , so is discontinuous at .
Suppose . For all integer , the set is closed and does not contain . For any positive integer , is open, so there is an open interval disjoint from , in which case the interval only contains points in or in with . Hence, for any with , we have , and since is arbitrary, this means the oscillation of at is zero. Thus, is continuous at , and we conclude .
There is one detail we need to work out, and that is each has a countable dense subset, i.e. it is separable. Luckily for us, a subset of a separable metric space is always separable, and is indeed separable, so is separable too. This detail will be left an exercise for the reader. After that, it should not be difficult to see that this theorem holds for any separable metric space, not just the reals, and the proof is mostly analogous.
To recap, we answer the two questions put forth from the beginning of the post:
What type of set can be?
It must be a countable union of closed sets, in other words, an set.
Given any set , is there a function that is discontinuous exactly on ?
In a separable metric space , for any set, there is a function whose is that set.
here stands for the French word fermé, meaning closed, while the means a "sum" of sets, i.e. a union. ↩