17

In elementary school, one learns that $2+2=4$ by experiment (putting two apples next to two other apples), and maybe also from some addition table to be memorized.

But is there any approach that proves $2+2 = 4$? If so, an example of such a proof would be good.

  • By the way, actually serious about this. It's not homework or anything, but I'm honestly curious. Thank you in advance. – Anoni Moose Aug 14 '14 at 03:19
  • 10
    Yes, see this paper for a proof. – JimmyK4542 Aug 14 '14 at 03:20
  • That's pretty cool. Thanks for finding this! – Anoni Moose Aug 14 '14 at 03:23
  • 11
    For this question to make sense you must specify what you mean by $2$, what you mean by $4$, and what you mean by $+$. Before you do that, you did not actually ask a question. – Ittay Weiss Aug 14 '14 at 03:25
  • 3
    We can rigorously construct the natural numbers, as well as define what we mean by addition. Statements like $2 + 2 = 4$ can then be proved from those definitions. You may find it helpful to read about the Peano axioms. – manthanomen Aug 14 '14 at 03:33
  • If you start with a set of axioms for the integers, as is typically done in number theory courses, then you could certainly prove it from the axioms. You would define $2 = 1 + 1$, then define $3 = 2 + 1$, then define $4 = 3 + 1$, then give a short proof. – littleO Aug 14 '14 at 03:34
  • 1
    If something wasn't equal to $2+2$, we usually wouldn't call it $4$. –  Aug 14 '14 at 03:35
  • 1
    But things that are $2 + 2$ sometimes get called $0$ or $1$. – nomen Aug 14 '14 at 03:36
  • 9
    that paper is just a joke – Gaston Burrull Aug 14 '14 at 03:37
  • 1
    @IttayWeiss No, he is asking if there are definitions that allow us to prove the statement; the question makes sense as it stands. – DanielV Aug 14 '14 at 07:44
  • What do you mean by "not relying on other facts of addition"? Obviously if you can not lean on any definition of addition you can not prove that addition has any given property. – Taemyr Aug 14 '14 at 07:47
  • 3
    I'm not sure what this has to do with logic. It's also impossible to tell just from the current question. Such open-ended questions (without saying what is the language and the axioms being used) were discussed before in the "almost the same" 1+1=2 questions: this one and that one and generally see this list. If all those discussions still haven't answered your question, please clarify your question, and I will vote to reopen it. – Asaf Karagila Aug 14 '14 at 12:39
  • 3
    I gave it a plus 1 because you are brave enough to ask such question. Kudos to you sir! – Sniper Clown Aug 14 '14 at 14:50
  • @AsafKaragila It may be "open ended" and vague but it has EVERYTHING to do with logic. – Sniper Clown Aug 14 '14 at 14:52
  • @SniperClown: EVERYTHING? Are you Gary Oldman in "Leon: The Professional"? I agree that this question has something to do with logic. But here's the kicker, when you ask a question about proving $f'$ doesn't exist when $f$ is not continuous, does it have to do with logic? Of course, but it's not tagged as [logic] (certainly not as the only tag) because there is a natural context to it. So I expect questions tagged under [logic] to have reasonably defined context to them, to prevent 8 answers of pure guesswork. So let me ask differently, how does this question relate to logic? – Asaf Karagila Aug 14 '14 at 14:55
  • @Asaf Karagila You might want to refrain from the ad hominem labelto live up to your preciously coveted "reputation". Also as a courtesy "we" expect that the community doesn't unilaterally belong to you when you start making assumptive gestures as "I expect" ex cathedra It's not your or my forum, it's our forum. How does it NOT relate to logic? Wanna define "logic" for me with Goedelian rigor? – Sniper Clown Aug 14 '14 at 15:02
  • At any rate, I had enough of this oligarchial mafiosi like moderation.

    I leave you all with your self-patting badges to feel good about yourselves instead of seeing the broader picture how this innocuous question would actually benefit education and widening of knowledge.

    So miopic. Sheesh.

    – Sniper Clown Aug 14 '14 at 15:03
  • @SniperClown: I have no idea what you want from me. I'm not sure you have either. I respectfully ask you not to use me to ventilate any bottled anger you have on other topics. Have a pleasant day. – Asaf Karagila Aug 14 '14 at 15:10
  • Voted to reopen. What could be more clear than, "Is it possible to prove 2+2=4?" Several postings here have outlined how it is possible. – Dan Christensen Aug 14 '14 at 15:21
  • @Dan: Since you well understood this question. What is the language being used? Is $2$ part of the language, or is it a predecided shorthand for a term in the language, e.g. $S(S(0))$ or $1+1$ or $S(0)+S(0)$ or $S(0)\cdot S(S(0))$ or is it ${\varnothing,{\varnothing}}$ which itself is a bit of a "cheat"? Next, what is $4$? What is $+$, in case the language does not include $+$? Finally, what are the axioms concerned? Are we using set theory, field theory, Peano axioms, Modular Arithmetic, second-order arithmetic, group theory, ring theory? And don't get me started on inference rules. – Asaf Karagila Aug 14 '14 at 18:05
  • 2
    @AsafKaragila The OP just wants to know if it possible to actually prove that 2+2=4 and to get some idea how you would go about it (without a lot of technical language). Why are you going on like this? – Dan Christensen Aug 15 '14 at 02:00
  • @Dan: Because mathematics is technical, as you may have noticed by now. And it is using a lot of fancy terms like "Peano axioms" and "first-order logic" vs. "second-order definitions" (which you have used and missed respectively in your answer), and whether or not $1,2,3,4$ and $+$ need to be defined at all. And so on and so forth. In hindsight I should have closed this one as a duplicate of one of the several $1+1=2$ threads which have comprehensive answers. In either case it's not clear at all what the OP wants to find out, or whether or not they are able to understand a proper answer. – Asaf Karagila Aug 15 '14 at 02:07
  • 2
    @AsafKaragila The answer you give to a high school student is going to be different from one you would give to a professor of mathematics. Not everyone wants or needs the same level of detail in an answer. You have to use some judgement. I could have posted a link to a formal proof that was several hundred lines in length and the OP, in all likelihood, would have gotten nothing out of it. – Dan Christensen Aug 15 '14 at 02:55
  • @Logan: I don't agree that all the answers of mine you linked are similar, and the questions are not all similar either. Yes, there are some similarities, but some say outright that the OP has no knowledge; others are interested in a particular system; and another really just asks what qualifies as a theorem. I do not like being called a hypocrite and I have taken quite an offense from your comments. I hope that you will amass 3500 more answers, so you would see why sometimes you feel like closing questions similar to those you previously answered. I hope you have a lousy day. I'm out of here. – Asaf Karagila Aug 15 '14 at 04:24
  • 1
    @AsafKaragila Yes I was having a bad day. Thanks for being the better man here. – Sniper Clown Aug 15 '14 at 19:05
  • @JimmyK4542 - Here's another one http://us.metamath.org/mpegif/2p2e4.html – Derek 朕會功夫 Oct 10 '14 at 17:53

6 Answers6

37

There are many ways of defining $\mathbb{N}.$ In the context of this question, the details aren't that important; what matters is that $\mathbb{N}$ ends up being a set equipped with a distinguished function $S : \mathbb{N} \rightarrow \mathbb{N}$ and a distinguished element $0 \in \mathbb{N}$ subject to a theorem that says "definitions by recursion work." This allows us to prove the existence and uniqueness of a binary operation $+$ on $\mathbb{N}$ satisfying the following specifications. $$n+0 = n, \quad n+S(m) = S(n+m)$$

Now write $4$ as shorthand for $S(S(S(S(0))))$ and write $2$ as shorthand for $S(S(0)).$ Then we have

$$2+2 = 2+S(S(0)) = S(2+S(0)) = S(S(2+0)) = S(S(2)) = S(S(S(S(0)))) = 4$$

Extra Information.

For completeness, here's several ways of defining the naturals.

  1. The algebraic structure $\mathbb{N}$ can be defined as the sole (up to unique isomorphism) model of the Peano Postulates (which are second order).

  2. It can also be defined as the free monounary algebra generated by the singleton set $\{0\}$ (I suggest googling this term if you do not know it).

  3. Set theorists like defining it as the least set $\omega$ such that firstly, $\emptyset \in \omega,$ and secondly, $x \in \omega$ implies $x \cup \{x\} \in \omega$. The entity $\emptyset$ ends up being our $0$, and the function $x \mapsto x \cup \{x\}$ ends up being our successor function.

goblin GONE
  • 67,744
  • You could simply define $+$ as you suggest (as did Peano in the original version of his axioms) or you can prove the existence a unique binary function $+$ with the required properties using the rules of logic and set theory if you have them available. – Dan Christensen Aug 14 '14 at 04:58
  • 1
    While this is a good answer in itself, I think it's probably incomprehensible to the OP, who in particular may not know the words "isomorphism," "second order," or "recursion" or be able to read explanations of what a free monounary algebra is. – Kevin Carlson Aug 14 '14 at 05:14
  • @KevinCarlson, you're right. I've reordered the information to emphasize what really matters. – goblin GONE Aug 14 '14 at 05:20
  • @KevinCarlson Not to mention the use of notation. I hope the OP knows what $\mathbb{N}.$ means and what $S : \mathbb{N} \rightarrow \mathbb{N}$ means. – trlkly Aug 14 '14 at 10:13
9

The usual approach for formally proving that $2+2=4$ is to start from Peano's axioms (which define the set $N$ of natural numbers , $0\in N$ and a successor function on $N$). Using these axioms, along with the rules of logic and set theory, you can formally prove that there exists a unique binary function $+$ such that

$x+0 = x$

$x+(y+1) = (x+y)+1$

where $1$ is the successor of $0$, and $n+1$ is the successor of $n$.

This is a long and tedious process. (Earlier versions of Peano's axioms gave you the above definition to start.)

Then you define 2, 3 and 4 such that

$2=1+1$

$3=2+1$

$4=3+1$

Then you have $2+2=2+(1+1)=(2+1)+1=3+1=4$

  • By "Peano's axioms" I mean the usual five axioms without $+$, i.e. not the first-order version which is more like Peano's original version of his axioms. I think you mean $1=S(0)$. Perhaps this is the source of your confusion. – Dan Christensen Aug 14 '14 at 04:36
  • You raise a legitimate point. I did leave out a few details for ease of reading, e.g. the implicit universal quantifiers, and the successor of $x$ being $x+1$, but I believe my definition is essentially correct. – Dan Christensen Aug 14 '14 at 04:50
  • This was not meant to be a formal proof. (I can supply one if you like, but it is several hundred lines long.) In the context of the OP, it is a stretch to say that my definition is somehow "misleading." – Dan Christensen Aug 14 '14 at 05:06
5

We assume the Peano axioms. Specifically:

  1. Zero is a number.
  2. If a is a number, the successor of a is a number.
    (We denote the successor of $x$ as $x'$.)
  3. Zero is not the successor of a number.
  4. Two numbers of which the successors are equal are themselves equal.
  5. (induction axiom.) If a set S of numbers contains zero and also the successor of every number in S, then every number is in S.

We then define addition recursively as follows:

$$a+0 = a$$ $$a+b' = (a+b)'$$

Now, we will name some numbers. We will denote:

$$\begin{align} 0' &= 1 \\ 1' &= 2 \\ 2' &= 3 \\ 3' &= 4 \end{align}$$

We could keep going, but we only need to be able to denote the numbers $0$ through $4$ (inclusive).

Now, showing that $2+2 = 4$ is a simple application of the recursive formula for addition: $$\begin{align} 2 + 2 &= 2 + 1' \\ &= (2+1)' \\ &= (2+0')' \\ &= ((2+0)')' \\ &= ((2)')' \\ &= 3'\\ &= 4 \end{align}$$

apnorton
  • 17,706
  • 1
    While (5) is true, it is not quite definitive. I suggest phrasing it as (5) "if $P(0)$ and $P(x) \rightarrow P(x')$ then $\forall x : P(x)$. The difference can be seen by considering the following structure, let the domain $D$ be $\mathbb{N} \cup {A, B, C}$ and successor be extended by defining for $n$ in $\mathbb N$, $n = n'$, also $A* = B, B* = C, C* = A$. $,{\mathbb N, '}$ satisfies both (5) and (5), but ${D, }$ satisfies (5) and not (5*) by taking $P(x)$ to be $x \in \mathbb N$. Besides functionality, a purpose of the inductive axiom is to exclude structures like the above. – DanielV Aug 14 '14 at 07:55
3

The proposition "2+2 = 4" is a theorem of the Peano arithmetic (the five Peano's axioms). For example: $$1 := 0',$$ $$2 := 1' = 0'',$$ $$3 := 2' = (1')' = 0''',$$ and so on.

Yes, it depends on to what meanings we assign "+" and the numerals.

Yes
  • 20,719
2

The most commonly used method to define sum in $\mathbb{N}$ is derived from Peano Axioms.

$0\in\mathbb{N}$ and $s:\mathbb{N}\to \mathbb{N}\setminus \{0\}$ is a given bijection, we can define $1:=s(0)$, $2:=s(1)$ and so on...

We can define a sum: $n+m:=s^{m}(n)$. In that case $2+2=s(s(2))=s(3)=4$. This sum satisfies the properties which we are accustomed.

Gaston Burrull
  • 5,449
  • 5
  • 33
  • 78
0

well this kind of proving depands on a set of axiom system of group like Z for example . and these are axioms that been used to prove that 2+2=4

Axiom 1. Algebraic Properties of Z (Properties of + and .)

Properties of Addition

  • A1. Associativity. For every x, y, and z in R, ( x + y ) + z = x + ( y + z ).

  • A2. Commutativity. For every x and y in R, x + y = y + x.

  • A3. Identity. R contains an additive identity, 0, such that for every x in R,
    x+0=x.

  • A4. Additive Inverses. For every x in R, there is an additive inverse, (-x ), in R such that x + (-x ) = 0.

Axiom 2. Order Properties of Z (Properties of <)

  • (i ) Transitive Property. For every a, b, and c in Z, if a < b and b < c, then a < c.

    (ii ) Trichotomy Property. For every a and b in Z, exactly one of the following holds: a =b, a < b, or b < a.

    (iii ) Additive Property. For every a, b, and c in Z, if a < b, then a + c < b + c.

    (iv) Multiplicative Property. For every a, b, and c in Z, if a < b and 0 < c, then ac < bc.

    (v) Order of Identities. 0 < 1.

Axiom 3. Th e Well-Ordering Principle

  • For any integer n, there is a next integer n + 1 that comes immediately after it, with no other integers in between.

with these sets of axioms you can easily prove now that 2+2=4. just folow these steps

1+1=2

(1+1) + 1 =2+1 =3

(1+1+1)+1=3+1=4

hence

(1+1) +(1+1 )=2 +2 =4 done.

Bswan
  • 421