Some practice with induction: a proof of the sum of exponents law for integer exponents
24 April 2022 · 937 words · Joseph Liu
The sum of exponents law, bx⋅by=bx+y, is well-known; indeed it was one of the first
properties that I learnt after being introduced to the concept of exponentiation. It’s easy to
reason intuitively as to why this property holds for integral exponents:
23+2=25=232⋅2⋅2⋅222⋅2,
However, I wanted to see how difficult it’d be to formalize this argument with induction. Let’s see!
Sum of Exponents Law for Integer Exponents. For positive real b, the equation
bx⋅>by=bx+y
holds for all integer x, y.
This law holds for x, y real, but we concern ourselves only with the easiest integer case in
this blog. Perhaps someday there will be a follow-up post…
Lemma 1.1 covers the case of two positive exponents (and, indirectly, the case of two negative
exponents); we still have no way to deal with exponents of different signs. The following lemma
addresses this gap:
Lemma 1.2. For all x,y∈Z+ where x≥y, we have bx−y=bx⋅b−y.
Equivalently, this lemma claims that bx+y=bx⋅by where y is negative and ∣x∣≥∣y∣;
however, the original presentation in terms of subtracting positive exponents is more amenable to induction.
Proof. We induct on x.
Base case. If x=1, then the condition x≥y forces y=1. Substituting, we have b1−1=b0=1
which is equal to b1⋅b−1=1.
Induction step. Let x∈Z+ be arbitrary and suppose bx−y=bx⋅b−y for all y∈Z+ such that x≥y. Then
We are now equipped to tackle the full proof where x and y are unrestricted aside from being
integers. Broadly speaking, Lemma 1.1 covers the case where x and y are both positive or both
negative, and Lemma 1.2 covers the case where one is negative and the other positive; all that
remains is to carefully invoke them and handle the straightforward case where one exponent is zero.
To recall, the statement we aim to prove is:
Sum of Exponents Law for Integer Exponents. For positive real b, the equation
bx⋅>by=bx+y
holds for all integer x, y.
Proof. We consider the following cases.
Case 1.Both x and y are positive. Lemma 1.1 discusses precisely this case.
Case 2.Both x and y are negative. Then −x and −y are both positive, so taking the
reciprocal allows us to apply Lemma 1.1 as follows: