The Set Tour, Stage 2: The Real Star

For the second of my little tour of sets that get commonly used as domains and ranges I want to name the most common of them all.


This is the real numbers. In text that’s written with a bold R. Written by hand, and often in text, that’s written with a capital R that has a double stroke for the main vertical line. That’s an easy-to-write way to distinguish it from a plain old civilian R. The double-vertical-stroke convention is used for many of the most common sets of numbers. It will get used for letters like I and J (the integers), or N (the counting numbers). A vertical stroke will even get added to symbols that technically don’t have any vertical strokes, like Q (the rational numbers). There it’s just put inside the loop, on the left side, far enough from the edge that the reader can notice the vertical stroke is there.

R is a big one. It’s not just a big set. It’s also a popular one. It may as well be the default domain and range. If someone fails to tell you what either set is, you can suppose she meant R and be only rarely wrong. The real numbers are familiar and popular and it feels like we know what they are. It’s a bit tricky to define them exactly, though, and you’ll notice that I’m not doing that. You know what I mean, though. It’s whole numbers, and rational numbers, and irrational numbers like the square root of pi, and for that matter pi, and a whole bunch of other boring numbers nobody looks at. Let’s leave it at that.

All the intervals I talked about last time are subsets of R. If we really wanted to, we could turn a function with domain an interval like [0, 1] into a function with a domain of R. That’s a kind of “embedding”. Let me call the function with domain [0, 1] by the name “f”. I’ll then define g, on the domain R, by the rule “whatever f(x) is, if x is from 0 to 1; and some other, harmless value, if x isn’t”. Probably the harmless value is zero. Sometimes we need to change the domain a function’s defined on, and this is a way to do it.

If we only want to talk about the positive real numbers we can denote that by putting a plus sign in superscript: R+. If we only want the negative numbers we put in a minus sign: R. Do either of these include zero? My heart tells me neither should, but I wouldn’t be surprised if in practice either did, because zero is often useful to have around. To be careful we might explicitly include zero, using the notations of set theory. Then we might write \textbf{R}^+ \cup \left\{0\right\} .

Sometimes the rule for a function doesn’t make sense for some values. For example, if a function has the rule f: x \mapsto 1 / (x - 1) then you can’t work out a value for f(1). That would require dividing by zero and we dare not do that. A careful mathematician would say the domain of that function f is all the real numbers R except for the number 1. This exclusion gets written as “R \ {1}”. The backslash means “except the numbers in the following set”. It might be a single number, such as in this example. It might be a lot of numbers. The function g: x \mapsto \log\left(1 - x\right) is meaningless for any x that’s equal to or greater than 1. We could write its domain then as “R \ { x: x ≥ 1 }”.

That’s if we’re being careful. If we get a little careless, or if we’re writing casually, or if the set of non-permitted points is complicated we might omit that. Mathematical writing includes an assumption of good faith. The author is supposed to be trying to say something interesting and true. The reader is expected to be skeptical but not quarrelsome. Spotting a flaw in the argument because the domain doesn’t explicitly rule out some points it shouldn’t have is tedious. Finding that the interesting thing only holds true for values that are implicitly outside the domain is serious.

The set of real numbers is a group; it has an operation that works like addition. We call it addition. For that matter, it’s a ring. It has an operation that works like multiplication. We call it multiplication. And it’s even more than a ring. Everything in R except for the additive identity — 0, the number you can add to anything without changing what the thing is — has a multiplicative inverse. That is, any number except zero has some number you can multiply it by to get 1. This property makes it a “field”, to people who study (abstract) algebra. This “field” hasn’t got anything to do with gravitational or electrical or baseball or magnetic fields. But the overlap in names does serve to sometimes confuse people.

But having this multiplicative inverse means that we can do something that operates like division. Divide one thing by a second by taking the first thing and multiplying it by the second thing’s multiplicative inverse. We call this division-like operation “division”.

It’s not coincidence that the algebraic “addition” and “multiplication” and “division” operations are the ones we call addition and multiplication and division. What makes abstract algebra abstract is that it’s the study of things that work kind of like the real numbers do. The operations we can do on the real numbers inspire us to look for other sets that can let us do similar things.