In 1703, Italian mathematician and monk Guido Grandi posed a deceptively simple-sounding question:

What is the sum of the following infinite series?

1 – 1 + 1 – 1 + 1 – 1 + 1 – 1…

With simple arithmetic, we can easily divide the series using parentheses (brackets):

(1 – 1) + (1 – 1) + (1 – 1) + (1 – 1)… = 0 + 0 + 0 + 0 +… = 0

But what if we changed the way we used the parentheses?

1 + (-1 + 1) + (-1 + 1) + (-1 + 1)… = 1 + 0 + 0 + 0 +… = 1

Because of the way negative numbers work, this solution is equally feasible. Ergo, **both 0 and 1 are acceptable answers**.

How can one series possibly have two different answers? Grandi used the fact that both 0 and 1 are possible from his series as *proof that God exists*, as something (*1*) can be made from nothing (*0*).

Grandi’s series becomes even stranger when a more advanced technique is applied.

Let us say that Grandi’s series is denoted by **S** (S = 1 – 1 + 1 – 1…).

We can then break down the series as 1 – (1 + 1 -1 + 1…), because the plus and minus signs can be inverted together.

Ergo, S = 1 – S → 2S = 1 → **S = ½**

**Now we have three answers to Grandi’s question: 0, 1 and ½.**

For over 150 years, mathematicians fiercely debated the answer to Grandi’s question. By the 19th century, mathematics had evolved and mathematicians had figured out better ways to solve infinite series.

The classic example is the solution to the series: **1 + ½ + ¼ + ⅛…**

To solve this, you can add the ** partial sums**, where you add each number to the sum of the previous numbers to see what number you are approaching (the limit).

1 → 1.5 → 1.75 → 1.875 → 1.9375… until we infinitely approach 2 (or 1.9999999…)

If we apply this method to Grandi’s series, we do not approach a single number because we keep swinging between 0 and 1. (1 → 0 → 1 → 0 → 1…)

So we can apply another method, where we **average the partial sums** as we go instead of adding.

e.g. 1 → ½(1 + 1.5) = 1.25 → ⅓(1 + 1.5 + 1.75) = 1.416 → ¼(1 + 1.5 + 1.75 + 1.875) = 1.531… until we approach 2.

Using this method on Grandi’s series:

1 → ½(1 + 0) = ½ → ⅓(1 + 0 + 1) = ⅔ → ¼(1 + 0 + 1 + 0) = ½…

Eventually, the series appears to converge on ½, showing that the answer to Grandi’s series seems to be ½.

The problem with this method is that **Grandi’s series does not actually have a limit, but we are applying a solution as if it has a limit**. This is similar to using a ** divide by 0 trick** to prove that 1 + 1 = 3. In mathematics, when rules are bent, we end up with weird, paradoxical results.

To show this empirically, consider the thought experiment of *Thomson’s Lamp*:

Imagine a lamp that is turned on after 1 minute, turned off after ½ minute, turned on again after ¼ minute *ad infinitum*.

This incorporates both infinite series discussed above.

Ergo, we know that the sum of time is 2 minutes.

So, at the end of 2 minutes, is the lamp on or off?

If Grandi’s series solves to 0, the light is off; if it is 1, the light is on.

Then what does it mean if Grandi’s series solves to ½? **Is the light on or off**?