2-node-supercomputer.net 2023-10-21
Today I claim that in algebra the word substitution is really a misnomer, coming from concentrating on the representation on paper rather than the meaning of a mathematical expression. A more accurate description would be to use the word “relabeling” instead of “substituting.”
I got onto this train of thought as I was listening to the podcast Sold a Story by Emily Hanford. In the podcast the narrator says that reading has been taught all wrong in school. Without going into the details of the podcast and whether it is right or not, I was thinking if something similar may be happening in Mathematics education. Is math hard because it is taught wrong? Maybe. The words “solve by subtitution” have always bothered me, as soon as I heard them uttered in the halls of the UC Davis campus years ago, and I think I now know why.
Take the set of equations
Usually, one would say that in order to calculate , you should substitute the value for into the first equation. That is, you erase the and write in its place. Such is the operation performed. Thus, you get .
In this description little is said of why this works or why this is the correct procedure to follow. Indeed, all it does is give you a rule, with no understanding needed. It is a procedure operating purely on the syntax level of the expressions as written on paper.
Indeed, this operational description makes it seem like and may actually be different things.
However, within the context of this example problem, the second equation tells us that and are equal. They refer to the same mathematical object, the number six. Thus, the substitution is not substituting what is in the equation. Rather, it merely changes the label for the underlying mathematical object.
That is, on the semantic level there is no substitution happening, and a more accurate term would be “relabeling” or “renaming”. Indeed, the substitution procedure works precisely because there is no substitution happening on the semantic level, only on the syntactic level.
Thus, when applying the “solve-by-substitution” technique, you are not actually substituting anything other than some labels.
OK, maybe there is substitution. Namely, when doing approximations. Consider the case when . However, we may not care that isn’t exactly ; it may be close enough for our purposes. Then, we may say
That is, we explicitly express that, within the current context, substituting the number for is sufficient.
Perhaps a philosophical difficulty could arise if is some small away from or . As we take the limit , can we ever say we don’t substitute?