# How big is infinity?

This might be more philosophy than math, but it’s been bothering me for a while.

Question: If there’s an infinite amount of real numbers between $0$ and $1$, shouldn’t there be twice the amount of real numbers between $0$ and $2$? Wouldn’t that be $2 \times \infty$? What about the amount of real numbers between $0$ and $\infty$? Wouldn’t that be $\infty \times \infty$?

I guess the whole concept of infinity just kind of evades me. Any help would be appreciated!

-
There are several different not necessarily compatible measures of size. In the sense of cardinality, the set $[0,1]$ and $[0,2]$ have the same (infinite) cardinality. In the sense of measure, one has measure $1$ and the other has measure $2$. With shapes, there are several measures of size (surface area, volume). They are useful in different contexts (painting, carrying). – André Nicolas Jun 17 '14 at 4:08
@SanathDevalapurkar, define mathematics then. – Fabien Jun 17 '14 at 4:13
"Infinity is pretty huge" ... – DonAntonio Jun 17 '14 at 4:17
"Infinity, is big. Really big. You just won't believe how vastly, hugely, mindbogglingly big it is. I mean, you may think it's a long way down the road to the chemist's, but that's just peanuts to infinity, listen..." – Asimov Jun 17 '14 at 4:24
I always wanted to answer this question with, simply, "bigger than that." – David Z Jun 17 '14 at 5:45

There are many different things that could be called "the infinite" in mathematics. None of them is a real number or a complex number, but some are used in discussing functions or real or complex numbers.

• There are things called $+\infty$ and $-\infty$. Those appear in such expressions as $$\lim_{x\to-\infty}\frac{1}{1+2^x} = 1 \text{ and }\lim_{x\to+\infty}\frac{1}{1+2^x}=0.$$

• There is also an $\infty$ that is approached as $x$ goes in either direction along the line. That occurs in $$\lim_{x\to\infty} \frac{x}{1-x} = -1\text{ and }\lim_{x\to1} \frac{x}{1-x}=\infty.$$ The second limit above could be said to approach $+\infty$ as $x$ approaches $1$ from one direction and $-\infty$ if from the other direction, but if one has just one $\infty$ at both ends of the line, then one makes the rational function a continuous function at the point where it has a vertical asymptote. This may be regarded as the same $\infty$ that appears in the theory of complex variables.

• There are "points at infinity" in projective geometry. This is similar to the "infinity" in the bullet point immediately preceding this one. Two parallel lines meet at infinity, and it's the same point at infinity regardless of which of the two directions you take along the lines. But two non-parallel lines pass through different points at infinity rather than the same point at infinity. Thus any two lines in the projective plane intersect exactly once.

• There are cardinalities of infinite sets such as $\{1,2,3,\ldots\}$ (which is countably infinite) or $\mathbb R$ (which is uncountably infinite). When it is said that Euclid proved there are infinitely many prime numbers, this sort of "infinity" is referred to.

• One regards an integral $\int_a^b f(x)\,dx$ as a sum of infinitely many infinitely small quantities, and a derivative $dy/dx$ as a quotient of two infinitely small quantities. This is a different idea from all of the above.

• Consider the step function $x\mapsto\begin{cases} 0 & \text{if }x<0, \\ 1 & \text{if } x\ge 0. \end{cases}$ One can say that its rate of change is infinite at $x=0$. This "infinity" admits multiplication by real numbers, so that for example, the rate of change at $0$ of the function that is $3.2$ times this function, is just $3.2$ times the "infinity" that is the rate of change of the original step function at $0$. This is made precise is the very useful theory of Dirac's delta function.

• There is the "infinite" of Robinson's nonstandard analysis. In that theory, we learn that if $n$ is an infinite positive integer, then evern "internal" one-to-one function that maps $\{1,2,3,\ldots,n-3\}$ into $\{1,2,3,\ldots,n\}$ omits exactly three elements of the latter set from its image. Nothing like that holds for cardinalities of infinite sets discussed above.

• I'm sure there are other examples that I'm missing here.

-
Great answer. Arguably Euclid's proof of the infinitude of the primes doesn't need anything to do with cardinalities or collections of integers. Saying that there are infinitely many primes is shorthand for saying that no matter what bound you pick there's always a bigger prime. – Dan Piponi Jun 17 '14 at 14:24
Might want to mention the hyperreals and surreals as well. (You allude to them with "Robinson's nonstandard analysis", but describing $\omega$ such that $\forall k \in \mathbb{N} \sum_0^k 1 < \omega$ might be clearer.) – zwol Jun 17 '14 at 19:35
It might also be worth mentioning infinite ordinals. Especially because that is one case where multiplying $\infty$ ($\omega$) by $2$ actually makes sense. – murgatroid99 Jun 17 '14 at 19:53
@Zack : By "hyperreals" do you mean something other than Robinson's theory that I mentioned in my answer? – Michael Hardy Jun 17 '14 at 19:55
@MichaelHardy At the time I wrote that I had forgotten that Robinson's theory and the hyperreals were the same thing. As I tried to explain in my edited comment, I think a gentler introduction to Robinson's theory talks about $\mathbb{R}^\star \supset \mathbb{R}$ and the extended (infinite and infinitesmal) elements of $\mathbb{R}^\star$. Conway's surreals, however, are their own thing AFAIK. – zwol Jun 17 '14 at 20:05

User wrote:

If there is an infinite amount of numbers between $0$ and $1$, shouldn’t there be twice that amount between $0$ and $2$?

No. Every real number in the interval $[0,1]$ can be mapped one-to-one to every real number in $[0,2]$ using the function $f(x)=2x$. So, there are no more numbers in $[0,2]$ than there are in $[0,1]$.

-
+1 - Why can't we all be more succinct? – dcromley Jun 18 '14 at 3:10
-1: I'm sorry, but this does not explain it to me. It makes sense, but if there are infinite natural numbers, and infinite real numbers between each natural number, how can the number of natural numbers and the number of real numbers be the same infinity, and not different orders of magnitude? – ANeves Jun 19 '14 at 12:06
@ANeves No, you're right. There are more real numbers than there are natural numbers. What this answer is claiming is that there are just as many real numbers in $[0,1]$ as there are in $[0,2]$. – David Zhang Jun 19 '14 at 14:18
@ANeves to piggyback on what David was saying, though both the natural numbers and the real numbers are infinite, they are not equivalent in size. One thing many laypeople do not understand or realize is that infinity is not a number, it's not equal to any number, and that two infinities can be different (or the same) in size from one another. – Andrew Coonce Jun 19 '14 at 16:32
@ANeves What does it mean that there are as many elements in $A$ as in $B$? Just that there exists a bijection $f:A\to B$, i.e. the elements of one set can be matched up 1-to-1 to all elements in the other set. – Dan Christensen Jun 19 '14 at 17:32

I have a rule of thumb: anything that says "infinity" (as opposed to $\aleph_0$, $\beth_2$, $\varepsilon_0$, $+\infty$, (projective) point at infinity, infinite, $\omega$, etc.) is probably not mathematics. It's not that mathematicians are afraid of the infinite -- hardly! -- but that it's an imprecise collection of heterogeneous concepts. Math is more precise than that.

In this case you're talking about the number of elements in a set (of real numbers), and so the particular concept you're talking about is called cardinality. The set [0, 1] has the same cardinality or 'size' as [0, 2]: precisely $2^{\aleph_0}=\beth_1.$

If you look at the fractions in [0, 1] and [0, 2] you'd find that, again, they have the same cardinality as each other, but a different cardinality from the above. This time they're 'just' $\aleph_0=\beth_0.$

If you drop all the way down to the integers the cardinality of the integers in [0, 1] is 2 while thee cardinality of the integers in [0, 2] is 3 and so these are no longer the same.

-
The first statement is wrong. Mathematicians work with the infinite just fine, and quite precisely. The problem is that when a layperson uses the word "infinity", they often do so in an imprecise way that only loosely corresponds to a mathematical idea if at all. (and often resist the notion that there can even be mathematical ideas about the infinite) – Hurkyl Jun 17 '14 at 5:35
@Hurkyl To be fair, that could be argued to be a distinction between the infinite and the word 'infinity', which - even in e.g. set theory - very seldom actually crops up (in lieu of more precise terms, of course!). – Steven Stadnicki Jun 17 '14 at 6:56
@Hurkyl: Say what you will, this is my rule of thumb. If I see "$+\infty$" there's no problem, if I see "$\aleph_0$" there's no problem, but if I see "infinity" it's almost never mathematical. In fact this is one of my most useful heuristics, better than the "real math papers use TeX" heuristic. – Charles Jun 17 '14 at 6:57
@Steven: That's because the noun infinity usually crops up in geometry: e.g. positive and negative infinity of the extended reals (the same infinity that appears in "the limit is infinity"), or the points at infinity of the projective plane. – Hurkyl Jun 17 '14 at 8:27
@Charles That is a far more cautious statement than your initial statement in this answer. – Brilliand Jun 17 '14 at 17:31

Intuitively, you are treating the "quantity" called infinity just like an ordinary real number. This is the defect.

You have a definition of $\infty$ and prove whether it satisfies the properties of real numbers, if yes, then you may say that,

If there is an infinite amount of numbers between $0$ and $1$, shouldn’t there be twice that amount between $0$ and $2$?

Note that you have assumed that the quantity $2$ $.$ $\infty$ is defined similarly as we define $x . y$ in $\mathbb {R}$.

Basically you may have assumed $\infty$ to be a very very big real number. This is the basic flaw, since $\infty$ is not a member of $\mathbb {R}$.

-

The ideas of being "of the same size", here , referring to your question we are dealing with cardinality, that we understand in the world of the finite, break down when we have infinite sets. There are formal definitions of what it means for two infinite sets to have the same size/cardinality or, informally, having the same amount of terms. There is a whole precisely-defined arithmetic of infinite cardinalities, with precisely-stated meanings, e.g.: $X^Y$ is the cardinality of all the functions from a set of cardinality represented by $X$ to a set with cardinality represented by $Y$.

Specifically, two infinite sets $X,Y$ have the same size if there is a one-one , onto map between them, i.e., if there is a bijection between $X,Y$ (there are related results like the Cantor-Schroder-Bernstein theorem that tell you that if there is an injection between $X,Y$ and an injection between $Y,X$ then there is a bijection between the two). In our case, since we can find a bijection between the two intervals $(0,1)$ and $(0,2)$, by our definition, the two have the same size (cardinality).

To answer your question, $2 \infty$ is decided by whther we can find a bijection between a set of that same infinite cardinality and the disjoint union of the set with itself.

Like Andre Nicolas stated in the comments, you can consider different measures of size, such as length, or measure in general. So one can have sets of the same cardinality but with different measure.

-
What is the downvote about? – user99680 Jun 23 '14 at 21:33

Infinity can be as big as you want it to be. It can be in the interval between 0 and 1, or it can be in the interval between 0 and 2, or in the interval between $(-3)^7$ to $10^{10^{10^{34}}}$, etc. I take it you've heard the term "infinitesimally small."

Try to count how many rational numbers there are between 0 and 1. Better yet, try to count how many rational numbers there are between 0 and 1 which can be expressed with a denominator that is a power of 2. There is $\frac{1}{2}$. There is $\frac{1}{4}$ and $\frac{3}{4}$. And $\frac{1}{8}$, $\frac{3}{8}$, $\frac{5}{8}$, $\frac{7}{8}$, ... you get the idea. Now try the same thing between 0 and 2. There's 1. There's $\frac{1}{2}$ and $\frac{3}{2}$. And $\frac{1}{4}$, $\frac{3}{4}$, $\frac{5}{4}$, and well, you run into the same problem.

In your mind, imagine a typical 12-inch ruler. But with this ruler, you can make the tickmarks as small as you want, whereas with a real ruler, you would get to a point at which you have to make the tickmarks so small you can't even see them.

The concept of infinity can easily lead to a lot of paradoxes, as well as statements that look nonsensical, like $\infty^2 = \infty^\infty = \infty$. I suggest you read http://www.suitcaseofdreams.net/infinity_paradox.htm

-
Welcome to math.se! I didn't know you had an account here. – Charles Jun 17 '14 at 16:57
Thanks, Charles. Only very recently, after I suggested it to an OEIS contributor. – user153918 Jun 17 '14 at 17:21
In what context is $\infty^{\infty} = \infty$ true? – Tobias Kildetoft Jun 19 '14 at 8:55
Maybe only when a mathematician or a smart-aleck is asking. Assuming Tetyana Butler is right, I think that $\infty^\infty$ follows from her assertion that $\infty^2 = \infty$. – user153918 Jun 19 '14 at 13:59
I can "accept" $\infty^2 = \infty$ since this for example is true if we replace $\infty$ by any infinite cardinal. But I cannot think of a context where we can both make sense of $\infty^{\infty}$ and have it equal $\infty$. – Tobias Kildetoft Jun 20 '14 at 9:01

I think you should not try to tackle understanding "infinity" head on. It's literally too much for human mind. So I'll try to give an "intuitively" understandable answer, thus avoiding "infinity" from now on:

All the numbers in [0, 2] can easily fit into [0, 1]! We can play a game to show this: You provide a number from [0, 2] and I'll return a number from [0, 1]. For any different numbers, I'll lose, if I fail to provide different numbers. I can only win, if I convince you that I won't lose.

You may start with 2. I'll go with 1.

Maybe 1? 0.5

square root of 2! (square root of 2) divided by 2.

Zero, got ya. Nope, zero, too.

Any number. Half that number.

See, what I'm getting at? There are as many numbers in [0, 1] as there are in [0, 2], because I can define a method to get one for every one you provide. I'll assume this is convincing enough and I won the game.

So to measure something that is too huge to measure, you just compare it, piece by piece, to something about the same size. Well, actually you don't compare, but tell how you would compare. If you can find a way to compare, piece by piece, it's not bigger.

The problem is to find a way to compare piece by piece. The only help I can give you in this, is the advice to avoid confusing yourself by the-bad-word. Instead ask yourself: "Where does the huge number of pieces come from?"

• [0, 2] looks twice as big as [0, 1]. But it's not, if you half it!

• [0, x] looks x times as big as [0, 1]. So as long as you can divide by x, everything is fine.

• We all know that dividing by zero isn't allowed. So we found something that's "smaller": [0, 0] has only one element, so it's clearly smaller.

• I disallowed myself from using the-bad-word. So I need to put it like this: If x keeps growing. What will happen to say 1/x?

At some point 1 divided by ever-growing-x will be indistinguishable close to 0. And if you can't distinguish between 0 and 1/ever-growing-x the piece to piece comparison fails. This makes [0, ever-growing-x] "bigger" than [0, 1] or [0, not-growing-x].

The next step would be to have a look at y-growing-way-faster-than-x. But I guess that'd overdo things.

-
Excellent layperson answer. – paqogomez Jun 18 '14 at 20:39
Your example is correct, but you should also note that with this function you need to get all of the elements in [0,1]. I can map {0,1,2,3...} to [0,1] by sticking a decimal point on the left of the number, but [0,1] is bigger. – Carlos Bribiescas Jun 24 '14 at 13:02
@CarlosBribiescas: You're right. But I wanted to keep it as simple as possible. So the comparison does only answer is A bigger than B with "yes" or "no". It does not answer "No, B is bigger than A.". BTW: Yes, it is a layperson answer, because the OP indicated to be a layperson. So thank you, for the compliment. – NoAnswer Jul 8 '14 at 10:19

This is a great anecdotal way to learn about different sizes of infinity.

Wikipedia Article - Hilbert's Hotel

So there are different sizes of infinity. Suppose you have two sets A and B and both of those sets are full of infinite things, that is, each of those two sets has infinite items. Not to say that they are they same size, one of their infinities may be bigger than the other.

If for every single item in A I can map it to exactly one item in B and I have all items in B covered, then the two sets are the same size of infinity.

If in all 1-1 mapping from A to B in which all items in A are used there are always items in B leftover, then the size of B is a bigger infinity than that of A.

-
I didn't use the normal "for all" terminology, but I think I addressed your point. In my defense, its been a while... :-D Thanks! – Carlos Bribiescas Jun 24 '14 at 12:54
Yes, except for the mapping s typo. – Mark Hurd Jun 25 '14 at 2:06

I take this perspective. Mathematicians cannot escape the consequences of the laws of physics; they can only ever perform a finite number of logical operations by applying a finite number of rules. That doesn't stop Mathematicians from interpreting the abstract objects they manipulate in a way that invokes a notion of "infinity", "uncountability" etc. just like the laws of physics won't stop theologians from discussing the Resurrection of Jesus. However, the fact that the laws of physics are computable implies that whenever we invoke a concept that cannot be realized in the real world that there always exists an alternative interpretation of the symbols that is compatible with the laws of physics.

Doron Zeilberger makes the same point here.

-

You need a little set theory to understand this concept well.

If a set s has infinite amount of elements then there exists a function f from S to a subset of S ( say T, note that T != S ).

Just that easy !

Now, Why does N (1, 2, ,3 , 4, ...) have infinite elements?

Consider f(x) = x + 1 it maps whole N to "N - { 1 }" therefore N so .

Why does R have infinite members!? Because we can find functions like f(x) = 1/x which map whole r between a subset of R ( here between 0 & 1 ).

That's all I known ... I hope it helps. If you are interested in set theory I recommend "Lectures in logic & set theory vol2" ( hey I am not a mathematics expert if you ask a more experienced one you shall get better resources )

-

Its been mentioned before. But this TED-Ed video is totally cool ..

The Infinite Hotel Paradox - Jeff Dekofsky

-
Would the downvoter care to comment on the reason for downvote? – square_one Jun 23 '14 at 11:32

Infinity is so large, it can contain multiple copies of itself, within itself. It's not at all like any of the other numbers, so many of the "common sense" approaches to infinity don't make a lot of sense.

That answer that pointed you to Cantor's Diagonalization Argument... yes, that's the way to go...

-

If you want to be able to treat $\infty$ more or less as a real number, look into hyperreal numbers. A hyperreal number is essentially a number defined by a sequence, which may or may not be limited. Therefore it is possible to say that the sequence $\{2,4,6,...\}$ is twice as large as the sequence $\{1,2,3,...\}$, even though both approach $\infty$. However, good luck applying hyperreals to the set of numbers between $0$ and $1$. That particular type of infinity is very difficult to define or measure, and almost completely useless.

-

By definition, "infinity" is so large that "ordinary" orders of magnitude don't change its definition.

To take an example, "a lot" so much that two times "a lot," or even a million times "a lot" doesn't "move the needle." That's basically what infinity is.

-

## protected by Michael Greinecker♦Jun 18 '14 at 0:02

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).