*i*. I told them it was defined as the square root of negative one.

"But you can't square root a negative."

"That's why we have to create a new definition - we want to see what results from this definition," I said.

One student was particularly enamored by this idea, and I suggested he look at the

*other*forbidden operation (besides square rooting a negative): dividing by zero. He defined 1/0 as

*u*, perhaps for "undefined," but called numbers involving

*u*"blank numbers," since a calculator would return a blank screen after the error.

Over several days I prodded this student to create theorems about blank numbers, and to avoid contradictions by adding "special rules." I have since lost his work, and don't remember the stumbling block that kept him from continuing.

Perhaps you can find it. Let 1/0 =

*u*.

Thm. 1:

*u*² =

*u*

Where does it go from here?

Well, if you are looking for a problem from your theorem, once you have that formula u² = u and you assume you are extending the reals by u, you get that u² - u = 0 which forces u = 0 or u = 1.

ReplyDeleteAnother big problem is that if you let u=1/0 then you get 0 = 0u = 0(1/0) =1 following usual rules for fraction canceling. Just my two cents.

Even though i is defined as the square root of -1, there are other properties (like algebraic closure) that make its existence plausible; whereas, dividing by 0 leads to bad problems almost instantly.

Good points, and the student involved encountered these as well. He was forced to make arbitrary rules, like 0u is neither 0 nor 1. I felt it was a good introduction to groups, which he has probably encountered by now.

ReplyDelete