What makes i^i interesting is the four real numbers it generates. (The numbers are +.2078… , -.2078… , +4.8104… , and -4.8104… .)

Can anyone give a geometric reason why an imaginary number raised to the power of an imaginary number generates four real numbers and no imaginary ones? What does \sqrt{-1}^{\sqrt{-1}} even mean? Is there anyone who can visualize a reason why the answers make sense? Are all the answers even correct? Or is only one correct, as any calculator that can do the calculation will tell you?

Complex numbers are two-dimensional numbers that are made by raising the number “e” to the power of an imaginary number—called “i”—times an angle in radians. Complex numbers lie on a circle in the complex number plane. Unless “e” is preceded by a number that stretches or shrinks it, the numbers lie on a unit circle like the one in the picture. Recall that “i” is the square root of minus one. When “i” is raised to the power of “i”, the result collapses onto the real number line—in one of four possible places. Which one? The numbers aren’t found on the unit circle. The process can be demonstrated mathematically, but any physical intuition about why imaginary numbers with imaginary exponents behave the way they do can be elusive.

Abstract math that hides no model that anyone can visualize makes results startling, even unnerving. It’s a lot like the quantum mechanics of entanglement or the physical meaning of gravity. They can be mathematically described and their effects accurately predicted, but no one can explain why.   

Mathematics alone can sometimes describe (or at least approximate) realities of the universe and how it seems to work, but as often as not when humans dive deep into the abyss of ultimate knowledge, math is unable to provide a picture that anyone can understand. How can that be? Things seem to happen that cannot be thought about except by playing around with numbers and being taken by surprise. Intuition is difficult, if not impossible.

Here is the solution of  i^i. Perhaps clues exist in the math that I’ve overlooked. If a model exists in the mind of a reader somewhere, I hope they will share it with me.

(1)       i^i = e^{\ln(i^i)} = e^{i\ln(i)} = cos (ln i) + i sin (ln i)

Now:  e^{i\frac{\pi}{2}} = i

Also:  ln ^{(e^{i\frac{\pi}{2})}} = ln i

Therefore:     ln i =  i (\frac{\pi}{2})

By substitution into line (1):    i^i = cos (i\frac{\pi}{2}) + i sin (i\frac{\pi}{2})

By half angle formulas:             i^i  = (\sqrt\frac{1 + cos (i\pi)}{2}) + i (\sqrt\frac{1 - cos (i\pi)}{2})

Convert 2nd term i to  \sqrt -1 :     i^i  = (\sqrt\frac{1 + cos (i\pi)}{2}) + \sqrt -1 (\sqrt\frac{1 - cos (i\pi)}{2})

(2)     Simplify the 2nd term:     i^i  = (\sqrt\frac{1 + cos (i\pi)}{2}) + (\sqrt\frac{cos (i\pi)-1}{2})

Euler’s cosine identity is:   cos θ =  \frac{e^{i\theta} + e^{-i\theta}}{2}

Therefore:                          cos (iπ) =  \frac{e^{i(i\pi)} + e^{-i(i\pi)}}{2}

(3)     Simplifying:               cos (iπ) =  \frac{e^{-\pi} + e^{\pi}}{2}

Substitute line (3) into line (2) and simplify:
i^i  = \sqrt{{\frac{1}{2} + \frac{e^{-\pi} + e^{\pi}}{4}}} + \sqrt{{\frac{e^{-\pi} + e^{\pi}}{4}} - \frac{1}{2}}

Now it’s just a matter of pulling out an old calculator and punching the keys.

e^{-\pi} = .043214;  e^{\pi} = 23.140693. I rounded off both numbers, because they seem to go on forever like π and “e”; they probably are irrational, because they don’t seem to be formed from ratios of whole numbers. Using these values will enable anyone to compute {i^i} who has a calculator with a square root key.

When square roots are calculated the answers can be positive or negative. Two negatives make a positive, right? So do two positives. So doing the math gives four numbers. See if your numbers match mine: .2078… , -.2078… , 4.1084… , and -4.1084… .  

I don’t know why. The answers aren’t intuitive. Who would guess that imaginary numbers raised to powers of imaginary numbers yield real numbers?—not a solitary number like anyone might expect, but four. Pick one. In nature a unique answer can be arbitrary—determined by chance, most likely.

In this case, no.

It feels to me like the imaginary fairies flying around in complex space are destined to collapse onto the real number line for no good reason, except that the math says they must collapse (maybe from exhaustion?) in at least one of four places. Can anyone make sense of it? 

The ln i is well known. It is i\frac{\pi}{2}, which equals (1.57078… i ). The ln of i^i can be rewritten by the rules of logarithms as i ln i, which is i times (1.57078…i ), which equals -1.57078… (a real number). Right? The ln of the correct answer must equal this number. Only one of the four results listed above has the right ln value: .2078… .

It seems odd that a set of equations I know to be sound should return a set of results from which only one can be validated by back-checking. Maybe there is something esoteric and arcane in the mathematics of logarithms that I missed during my education along the way.

Then again square roots can be messy; there are two square roots in the final equation, each of which can be evaluated as positive or negative. Together they produce four possible answers, but just one result is the right one.

Adding the four numbers is kind of interesting. They sum to zero. That is so like the way the universe seems to work, isn’t it? When everything is added up, physicists like Stephen Hawking claim, there’s really nothing here. Everything is imaginary. Some philosophers agree: everything that is real is at its core imaginary.   

Are there clues in the pictures and models of complex number space that would ever make anyone think? Sure, I totally get it. Yeah, I’ve got this. Real numbers cascading out of imaginary powers of imaginary numbers make perfect sense—like snowflakes falling from a dark sky.

A mathematician told me, Rotating and scaling is all it is. The base must be the imaginary “i” alone; “i” is the key that unlocks everything. The power of the key can be any imaginary number at all; “i” is why the result of every imaginary power of “i” becomes real. 

The explanation calms me; but it seems somehow incomplete; it’s missing something; in my gut I feel like it can’t be entirely right, though it purports to persuade what the math insists is truth. We are being asked to believe, for now at least, and move on.

Billy Lee 

What is e exp (-i π) ?

What is e^{-i\pi} ?

I posted a long answer on Quora.com where it sort of didn’t do well. Answers given by others were much shorter but they seemed, at least to me, to lack geometric insights. After two days my answer was ranked as the most read, but for some reason no one upvoted it. It did receive a few positive replies, though.

I can’t help but believe that there must be nerds in cyberspace who might enjoy my answer. Why not post it on my blog? Maybe someday one of my grandkids will get interested in math and read it. Who knows?

I added some graphics here that wouldn’t post on Quora. The site lacks graphics functionality, apparently. Either that or I’m too dense to figure out how to insert something.

NOTE: 24 Oct 2017: Today Billy Lee finally figured out how to post media on Quora.com. After admonishment and chastisement by the staff, he added a pic and a GIF to his “answer.” Unfortunately, Billy Lee never did get the GIF to run right, so he took it down. Otherwise, Billy Lee is doing good. He really is. THE EDITORIAL BOARD

Anyway, this pic and a working GIF below make a big difference in understanding, I hope. And anyone who doesn’t understand something can always click on a link for more information. (No one ever clicks on links, but I spend a lot of time adding them—maybe so I can click on them myself during times when remembering my name or where I live seems to lie just outside my intellectual skill-set.)

Here is the drawing I added and the answer:

This diagram is excellent but contains a mystery point not on the unit circle—{i^i}. The point is shown at .2078… on the real number line. It has 3 other values on the horizontal line that are not shown: -.2078… and +/- 4.8104… . Its solution can be reviewed in this video link, which Billy Lee has carefully reviewed and confirmed to be sound. An imaginary number raised to the power of an imaginary number yields a result that is four real numbers. How can that be? It’s something to ponder; something to think about. The Editorial Board 

What is e^{-i\pi} ?

The expression evaluates to minus one; the answer is (-1). Why?

Numbers like these are called complex numbers. They are two-dimensional numbers that can be drawn on graph-paper instead of on a one-dimensional number line, like the counting numbers. They are used to analyse wave functions—i.e. phenomenon that are repetitive—like alternating current in the field of electrical engineering, for example.

“e” is a number that cannot be written as a fraction (or a ratio of whole numbers). It is an irrational number (like π, for instance). It can be approximated by adding up an arbitrary number of terms in a certain infinite series to reach whatever level of precision one wants. To work with “e” in practical problems, it must be rounded off to some convenient number of decimal places.

Punch “e” into a calculator and it returns the value 2.7182…. The beauty of working with “e” is that derivatives and integrals of functions based on exponential powers of “e” are easy to calculate.

What is “e” raised to the power of (-iπ) ?

A wonderful feature of the mathematics of complex numbers is that all the values of expressions that involve the number “e” raised to the power of “i” times anything conveniently lie on the edge (or perimeter) of a circle of radius one. This happy fact makes understanding the expressions easy.

I should mention that any point in the complex plane can be reached by adding a number in front of e^{i\theta} to stretch or shrink the unit circle of values. We aren’t going to go there. In this essay “e” is always preceded by the number one, which by the familiar convention is never shown.

The number next to the letter “i” is simply the angle in radians where the answer lies on the circle. Draw a line from the center of the circle at the angle specified in the exponent of “e” and it will intersect the circle at the value of the expression. What could be easier?

For the particular question we are struggling to answer, the number in the exponent next to “i” is (-π), correct?

“π radians” is 180 degrees, right? The minus sign is simply a direction indicator that says to move clockwise around the unit circle—instead of counter-clockwise if the sign was positive.

After drawing a unit circle on graph paper, place your pencil at (1 + 0i)—located at zero radians (or zero degrees)—-and trace 180 degrees clockwise around the circle. Remember that the circle’s radius is one and its center is located at zero, which in two dimensional, complex space is (0 + 0i). You will end up at the value (-1 + 0i) on the opposite side of the circle, which is the answer, by the way.

[Trace the diagram several paragraphs above with your finger if you don’t have graph paper and a pencil. No worries.]

Notice that +π radians takes you to the same place. The value you land on is (-1 + 0i), which is -1. The answer is minus one.

Imagine that the number next to “i” is (π/2) radians. That’s 90 degrees, agreed? The sign is positive, so trace the circle 90 degrees counter-clockwise. You end at (0 + i), which is straight up. “i” in this case is a distance of one unit upward from the horizontal number line, so the number is written (0 + i)—-zero distance in the horizontal direction and “plus one” distance in the “i” (or vertical) direction.

So, the “i” in the exponent of “e” says to “look here” to find the angle where the value of the answer lies on the unit circle; on the other hand, the “i” in the rectangular coordinates of a two-dimensional number like (0 + i) says “look here” to find the vertical distance above or below the horizontal number line.

When evaluating “e” raised to the power of “i” times anything, the angle next to “i”—call it “θ”—can be transformed into rectangular coordinates by using this expression: [cos(θ) + i sin(θ)].

For example: say that the exponent of “e” is i(π/3). (π/3) radians is 60 degrees, right? The cosine of 60 degrees is 0.5 and the sine of 60 degrees is .866….

So the value of “e” raised to the power of i(π/3) is by substitution (0.5 + .866… i ). It is a two-dimensional number. And it lies on the unit circle.

The bigger the exponent on “e” the more times someone will have to trace around the circle to land at the answer. But they never leave the circle. The answer is always found on the circle between 0 and 2π radians (or 0 and 360 degrees) no matter how large the exponent.

It’s why these expressions involving “e” and “i” are ideal for working with repetitive, sinusoidal (wave-like) phenomenon.

In this essay Billy Lee uses θ in place of the Greek letter φ shown in this GIF. “r” equals “one” in a unit circle. The Editorial Board

Some readers might wonder about what radians are. A radian is the radius of a circle, which can be lifted and bent to fit perfectly on the edge of the circle. It takes a little more than three radius pieces (3.14159… to be more precise) to wrap from zero degrees to half-way around any circle of any size. This number—3.14159…—is the number called “π”. 2π radians are a little bit more than six-and-a-quarter radians (radius pieces) and will completely span the perimeter (or circumference) of a circle.

A radian is about 57.3 degrees of arc. Multiply 3.1416 by 57.3 to see how close to 180 degrees it is. I get 180.01… . The result is really close to 180 degrees considering that both numbers are irrational and rounded off to only a few decimal places.

One of the rules of working with complex numbers is this: multiplying any number by “i” rotates that number by 90 degrees. The number “i” is always located at 90 degrees on the unit circle by definition, right? By the rule, multiplying “i” by “i” rotates it another 90 degrees counter-clockwise, which moves it to 180 degrees on the circle.

180 degrees on the unit circle is the point (-1 + 0i), which is minus one, right?

So yes, absolutely, “i” multiplied by “i” is equal to -1. It follows that the square root of minus one must be “i”. Thought of in this way, the square root of a minus one isn’t mysterious.

It is helpful to think of complex numbers as two dimensional numbers with real and imaginary components. There is nothing imaginary, though, about the vertical component of a two-dimensional number.

The people who came up with these numbers thought they were imagining things. The concept of two-dimensional numbers was too radical at the time for anyone to believe that numbers could exist on a plane as naturally as they do on a simple number line of one dimension.

Visit my website for an illustrated and concise explanation of these topics titled, “What is Math?” It is easily found by making a simple entry into the search box on theBillyLeePontificator.com .

Thanks, I was confused about the -i part, so it’s not – sqrt(-1)? My calculator gives me an answer of -1.

Good answer, I approve 🙂


Billy Lee