The Sum of All the Natural Numbers


by AlfieD
Tags: bosonic, infinite series, natural, numbers, string theory, zeta function
AlfieD
AlfieD is offline
#1
Jan10-14, 05:31 PM
P: 68
Hi lovely people,

I recently came across a video http://www.youtube.com/watch?v=w-I6XTVZXww that said if you add all of the natural numbers from 1 to infinity, the answer is... What do you think it is? Infinity or something like that?

They said it was -1/12. I watched the proof but I don't understand the logic behind it because if you add positive numbers together, how can you get a negative? And if you're adding whole numbers together, how can you get a fraction (not like 5=5/1, you know what I mean)?

I heard them say that it's essential to String Theory and all 26 dimensions coming out. I assume they meant bosonic string theory? Because that's the only subset of string theory I know that has 26 dimensions, I'm pretty sure the original has 10.

I'm hoping that someone can shed some light on this so called 'astounding result'.

Thanks in advance,
AlfieD
Phys.Org News Partner Mathematics news on Phys.org
Researchers help Boston Marathon organizers plan for 2014 race
'Math detective' analyzes odds for suspicious lottery wins
Pseudo-mathematics and financial charlatanism
mfb
mfb is offline
#2
Jan10-14, 06:20 PM
Mentor
P: 10,769
This is not the sum of all natural numbers. He uses formulas in a region where they do not apply.

As a comparison: "all odd numbers between 2 and 8 are prime" is true (as 3, 5 and 7 are prime) and easy to prove, but that does not mean 1 or 9 would be prime as well because they are not in the region where the statement is usable.
AlfieD
AlfieD is offline
#3
Jan10-14, 06:24 PM
P: 68
Quote Quote by mfb View Post
This is not the sum of all natural numbers.
In what way exactly? It says that it's the sum of all the positive whole numbers from 1 to infinity.

1MileCrash
1MileCrash is offline
#4
Jan10-14, 06:33 PM
1MileCrash's Avatar
P: 1,227

The Sum of All the Natural Numbers


Quote Quote by mfb View Post
This is not the sum of all natural numbers. He uses formulas in a region where they do not apply.

As a comparison: "all odd numbers between 2 and 8 are prime" is true (as 3, 5 and 7 are prime) and easy to prove, but that does not mean 1 or 9 would be prime as well because they are not in the region where the statement is usable.
Did you watch the video? Grimes proved it without involving any type of formula, he just did algebra with other series.
mfb
mfb is offline
#5
Jan10-14, 06:34 PM
Mentor
P: 10,769
@AlfieD: Yeah sure, but it is wrong. Entertaining, but wrong.

With incorrect manipulations of infinite sums, you can prove anything.


As an example, consider the sum 1/2 - 1/3 + 1/4 - 1/5 + 1/6 - 1/7 +-...
Clearly the sum is positive everywhere and increases every two steps, so it is larger than zero.
1/2 - 1/3 + 1/4 - 1/5 + 1/6 - 1/7 +- ... > 0
Let's rearrange it a bit:
- 1/3 - 1/5 + 1/2 - 1/7 - 1/9 + 1/4 +- ...
Now -1/3 - 1/5 = -8/15 and 8/15>8/16=1/2, and in the same way -1/7 - 1/9 = 16/63 > 16/64 = 1/4 and so on.
Therefore, this sum is clearly negative.
- 1/3 - 1/5 + 1/2 - 1/7 - 1/9 + 1/4 +- ... < 0

How can the same sum be larger and smaller than zero at the same time?

Well, the answer is bad mathematics - this rearrangement is not valid, it changes the value of the sum. The same is true for the steps made in the video - they are just not valid mathematics.

@1MileCrash: Yes I watched it. Maybe "formulas" is not the best word, let's say "calculations".
1MileCrash
1MileCrash is offline
#6
Jan10-14, 06:40 PM
1MileCrash's Avatar
P: 1,227
Quote Quote by mfb View Post
@1MileCrash: Yes I watched it. Maybe "formulas" is not the best word, let's say "calculations".
OK, you said "in regions where they do not apply" so I thought it sounded like you were referring to Riemann-Zeta regularization, which was not involved.

I'm not very convinced there there was anything wrong with the work he did, but I will look at what you said more closely.
AlfieD
AlfieD is offline
#7
Jan10-14, 06:44 PM
P: 68
Quote Quote by mfb View Post
@AlfieD: Yeah sure, but it is wrong. Entertaining, but wrong. [text...] - this rearrangement is not valid, it changes the value of the sum. The same is true for the steps made in the video - they are just not valid mathematics.
This is essential to bosonic string theory in allowing the 26 dimensions to exist, so are you saying that bosonic string theory can't be correct?
AlfieD
AlfieD is offline
#8
Jan10-14, 06:46 PM
P: 68
Quote Quote by 1MileCrash View Post
OK, you said "in regions where they do not apply" so I thought it sounded like you were referring to Riemann-Zeta regularization, which was not involved.
Yeah, that Riemann-Zeta thingy was in a different proof, but as you said, it wasn't involved in the linked video proof.
Student100
Student100 is offline
#9
Jan10-14, 07:02 PM
Student100's Avatar
P: 448
http://www.youtube.com/watch?v=E-d9m...ature=youtu.be

Is their other 'proof' video, but it's still incorrect.

The first video is just mathematical hand waving.

They're misusing the mathematics.
DrewD
DrewD is online now
#10
Jan10-14, 07:08 PM
P: 427
Quote Quote by 1MileCrash View Post
OK, you said "in regions where they do not apply" so I thought it sounded like you were referring to Riemann-Zeta regularization, which was not involved.

I'm not very convinced there there was anything wrong with the work he did, but I will look at what you said more closely.
I don't know if "wrong" would be the exact term that I would use, but it is definitely misleading. They are using and rearranging divergent sums to get a result. There is a sense in which it is true (from what I have read, but I am not an expert in this area), but what he does is not valid using standard summation. Even from the beginning they are breaking "rules". For example ##\Sigma^{\infty}_n(-1)^n## does not converge. Sure, it can be useful to redefine a sum as the average of the partial sums (which I am pretty sure is what he is doing), but it is misleading to say that this holds true when we are considering a standard summation.

He also rearranges the sums for his second sum (I think-I have watched the video, but not today). Since an infinite sum is usually defined as the limit of the sequence of partial sums, this is not acceptable.

This is not something that I know much about, but I have been told that the truth of this statement is important to some conclusions in math and physics, but his "proof" is extremely flawed.
1MileCrash
1MileCrash is offline
#11
Jan10-14, 07:22 PM
1MileCrash's Avatar
P: 1,227
Quote Quote by DrewD View Post
I don't know if "wrong" would be the exact term that I would use, but it is definitely misleading. They are using and rearranging divergent sums to get a result. There is a sense in which it is true (from what I have read, but I am not an expert in this area), but what he does is not valid using standard summation. Even from the beginning they are breaking "rules". For example ##\Sigma^{\infty}_n(-1)^n## does not converge. Sure, it can be useful to redefine a sum as the average of the partial sums (which I am pretty sure is what he is doing), but it is misleading to say that this holds true when we are considering a standard summation.

He also rearranges the sums for his second sum (I think-I have watched the video, but not today). Since an infinite sum is usually defined as the limit of the sequence of partial sums, this is not acceptable.

This is not something that I know much about, but I have been told that the truth of this statement is important to some conclusions in math and physics, but his "proof" is extremely flawed.
1 - 1 + 1 - 1 +... does not converge by our definitions, but the argument that it is 1/2 is convincing to me. I'm not so sure that our standard notion of convergence is the only way to associate a series with a value.
AlfieD
AlfieD is offline
#12
Jan10-14, 07:27 PM
P: 68
Can I just ask whether the quarrel anyone has is with this particular proof and not the actual answer itself? So, are you fine with it being -1/12, but you just think that the proof is highly floored. If so, do you know of any better proofs?
Student100
Student100 is offline
#13
Jan10-14, 07:34 PM
Student100's Avatar
P: 448
It's only -1/12 when used in the context of ζ(-1). Suggesting that the infinite sum 1+2+3+4+5... is -1/12 is just wrong, it's an abuse of notation.

Someone correct me if I'm wrong, I'm working off a limited knowledge base here.
mfb
mfb is offline
#14
Jan10-14, 07:52 PM
Mentor
P: 10,769
Quote Quote by 1MileCrash View Post
I'm not very convinced there there was anything wrong with the work he did, but I will look at what you said more closely.
None of the sums he uses has a proper value.

As another example:

Assume 1+2+3+.... = -1/12.
Then clearly
0+1+2+3+.... = -1/12.
Taking the difference:
1+1+1+... = 0.
In the same way,
0+1+1+1+... = 0.
Taking the difference again,
1+0+0+..=0
1=0

Wait... no.

Quote Quote by AlfieD View Post
This is essential to bosonic string theory in allowing the 26 dimensions to exist, so are you saying that bosonic string theory can't be correct?
Please give a reference that the actual sum of natural numbers (and not the value of the Riemann Zeta function) is used there.
AlfieD
AlfieD is offline
#15
Jan10-14, 07:56 PM
P: 68
Quote Quote by mfb View Post
Please give a reference that the actual sum of natural numbers (and not the value of the Riemann Zeta function) is used there.

Ummm, if you watch the video the guy shows you the sum inside a book entitled String Theory. Listen closely to what he says as well. I'm pretty sure it's near the start of the video.
Mandelbroth
Mandelbroth is offline
#16
Jan10-14, 08:12 PM
Mandelbroth's Avatar
P: 597
Quote Quote by AlfieD View Post
Hi lovely people,

I recently came across a video http://www.youtube.com/watch?v=w-I6XTVZXww that said if you add all of the natural numbers from 1 to infinity, the answer is... What do you think it is? Infinity or something like that?

They said it was -1/12. I watched the proof but I don't understand the logic behind it because if you add positive numbers together, how can you get a negative? And if you're adding whole numbers together, how can you get a fraction (not like 5=5/1, you know what I mean)?

I heard them say that it's essential to String Theory and all 26 dimensions coming out. I assume they meant bosonic string theory? Because that's the only subset of string theory I know that has 26 dimensions, I'm pretty sure the original has 10.

I'm hoping that someone can shed some light on this so called 'astounding result'.

Thanks in advance,
AlfieD
I haven't watched the video yet, but I'm pretty sure that's using something called a Ramanujan summation. The idea was created by Ramanujan, a famous Indian mathematician. It's not an "actual" summation, but it can apparently be helpful sometimes in number theory.

Edit: Perhaps that's what the idea is, but in the video all I see is mathematical crackpottery. It hurts my eyes.
DrewD
DrewD is online now
#17
Jan10-14, 09:55 PM
P: 427
Quote Quote by AlfieD View Post
Ummm, if you watch the video the guy shows you the sum inside a book entitled String Theory. Listen closely to what he says as well. I'm pretty sure it's near the start of the video.
Be careful. Have you read the book? I am looking for this part of the book, but I bet mfb is right: the book is probably using a value of the analytic extension of the zeta function. This is different (in ways the mfb is probably more comfortable with than I am) than just saying that adding all of the natural numbers together gives this value. It is extremely simple to show that this series, using the normal definitions of summation, is divergent.

The formula is not wrong when it is viewed properly, but the video is ambiguous and uses false mathematics for a "proof" that is meaningless.
atyy
atyy is offline
#18
Jan10-14, 11:25 PM
Sci Advisor
P: 8,004
Although the proof is wrong, some mathematicians have savoured it. In particular, Atiyah writes that the erroneous summation was Euler's, and that it was only in relatively recent times that we understand what Euler's intuition was pointing towards.

Another good fun source that discusses the history of this equation is John Baez's
My Favorite Numbers: 24
http://math.ucr.edu/home/baez/numbers/24.pdf
http://www.youtube.com/watch?v=vzjbRhYjELo

For Atiyah's comment see his article "How research is carried out" in http://books.google.com/books?id=YJ0...gbs_navlinks_s (p213).


Register to reply

Related Discussions
Cardinality of Natural even numbers and Natural numbers Set Theory, Logic, Probability, Statistics 19
Natural Numbers General Math 3
all natural numbers General Math 3
Natural numbers Z? Linear & Abstract Algebra 6
sum of the first n natural numbers is n(n+1)/2 General Math 2