When can two expressions with square roots be equal? 8 and 22 are, but can 2 be a rational multiple of 3, or can 1+2 equal 3−5? §14 answers such questions, and the answer has a remarkable consequence: in any identity built from rationals and q, you may replace q by −q throughout and the identity stays true. That symmetry is the seed of Galois theory, and in this chapter it explains why the “conjugate” a−b is always a root alongside a+b.
The idea
Every number made from rationals and one surd q by +, −, × can be written as P+Qq with P, Q rational: the rule (q)2=q lets you collapse any power. For example
(1+2)2=3+22,(1+2)3=7+52.
The calculation used only (2)2=2, which is equally true of −2. So the same calculation, run with −2, gives (1−2)3=7−52. For this to be a theorem rather than a trick, the pair (P,Q) must be determined by the number P+Qq; that is what Hardy proves.
ExplorePowers and their conjugates
Try this.With p=1, q=2, compare the columns P+Qq and P−Qq with (1+2)m and (1−2)m. Look at the last column. Then try p=2, q=3, and finally make q a perfect square such as 4.
(1+2)m=P+Q2 and (1−2)m=P−Q2
m
P
Q
P+Qq
P−Qq
P2−qQ2
0
1
0
1
1
1
1
1
1
2.414214
−0.4142136
−1
2
3
2
5.828427
0.1715729
1
3
7
5
14.07107
−0.07106781
−1
4
17
12
33.97056
0.02943725
1
5
41
29
82.01219
−0.01219331
−1
6
99
70
197.9949
0.005050634
1
Each row comes from the one above by (P+Q2)(1+2)=(P+2Q)+(P+Q)2, a rule that only uses (2)2=2. That rule is just as true for −2, so the same P and Q give the conjugate powers. Check, for m = 6: (1+2)6≈197.9949 and (1−2)6≈0.005050634, matching the last row. The last column is (12−2)m=(−1)m, the product of the two conjugates. Since ∣1−2∣<∣1+2∣, the ratio P/Q approaches 2: here P/Q=1.41428571 against 2=1.41421356.
What to notice
The same integers P, Q serve both (p+q)m and (p−q)m. The last column is (p2−q)m, the product of the two: for p=1, q=2 it alternates ±1, so P/Q are the approximations 11,23,57,1217,… to 2 of Examples III. When q is a square, q is rational and the whole argument collapses.
Precise statements
DefinitionSimilar and dissimilar surdsHardy: §14
Two pure quadratic surds are similar if they are rational multiples of the same surd, and dissimilar otherwise. For example 8=22 and 25/2=252 are similar.
TheoremCoprime non-squares give dissimilar surdsHardy: §14
If M and N are integers with no common factor, neither of which is a perfect square, then M and N are dissimilar.
TheoremComparing A+B and C+DHardy: Theorem, §14
If A, B, C, D are rational and A+B=C+D, then either A=C and B=D, or B and D are both squares of rational numbers.
CorollaryConjugatesHardy: Corollary, §14
If A+B=C+D, then A−B=C−D, unless B and D are both rational.
Proofs
ProofCoprime non-squares give dissimilar surds
Strategy
If M were a rational multiple of N, then MN would be rational, hence an integer, so MN would be a perfect square. With no common factor, the prime factors of that square must split cleanly between M and N, making each a square.
MN would be rational
Suppose M=(a/b)N with a, b integers. Multiplying by N: MN=(a/b)N, a rational number.
So MN is a perfect square
An integer that is the square of a rational is the square of an integer (§3). So MN=P2 for an integer P.
Split the prime powers
Write P=aαbβcγ⋯ with distinct primes a,b,c,…. Then a2α divides MN. Since M and N share no factor, the prime a divides only one of them, and that one is divisible by all of a2α. The same holds for each prime of P.
Both are squares, a contradiction
So M is the product of some of the factors a2α,b2β,… and N of the rest: both are perfect squares, contrary to hypothesis.
∎
ProofIf A+B=C+D then A=C, B=D, or both roots are rational
Strategy
Assume B=D. Then B−D is a non-zero rational, and dividing the rational B−D by it shows that B+D is rational too. Sum and difference rational means each root is rational.
If B=D, then A=C
Subtract B=D from both sides. So assume B=D from now on.
B−D is a non-zero rational
It equals C−A, which is rational. It is not 0, since B=D would give B=D on squaring.
B+D is rational
B−D=(B−D)(B+D), so B+D=(B−D)/(B−D), a quotient of rationals.
Both roots are rational
B=21{(B+D)+(B−D)} and D=21{(B+D)−(B−D)} are rational, so B and D are squares of rationals.
∎
The corollary is immediate: unless both roots are rational, A=C and B=D, and then of course A−B=C−D. Read differently, the theorem says that P and Q in P+Qq are uniquely determined (when q is irrational): if P+Qq=P′+Q′q with Q=Q′, then q=(P−P′)/(Q′−Q) would be rational. That uniqueness is what lets us “switch the sign of q” in the widget.
Worked examples and exercises
Exercise●○○Hardy: Examples VIII, 1
Prove from first principles that 2 and 3 are not similar surds.
Hint
If 3=(a/b)2 then 3b2=2a2. Look at the power of 2 (or of 3) on each side.
Solution
Suppose 3=(a/b)2 with positive integers a,b. Squaring, 3b2=2a2. In the prime factorisation, the power of 3 dividing 3b2 is odd (one more than twice the power in b), while the power of 3 dividing 2a2 is even (twice the power in a). Unique factorisation makes this impossible. (Alternatively: 6=23=(a/b)⋅2 would be rational, but 6 is not a square.)
Exercise●●○Hardy: Examples VIII, 3
If a and b are rational, show that a+b cannot be rational unless a and b are both rational; and the same for a−b, unless a=b.
Hint
Use the identity a−b=(a−b)(a+b), as in the proof of the theorem.
Solution
Suppose s=a+b is rational. If s=0 then a=b=0 (both are ≥0), which are rational. Otherwise a−b=(a−b)/s is rational too, and so are a=21(s+(a−b)/s) and b. Now suppose d=a−b is rational and a=b. Then d=0 and a+b=(a−b)/d is rational, and we conclude as before. (If a=b, the difference is 0 whatever a is.)
Show that every polynomial in x=p+q with rational coefficients can be written P+Qq with P, Q rational. Deduce that if a+b (b not a square) is a root of a polynomial equation with rational coefficients, so is a−b.
Hint
Induction on the power: (P+Qq)(p+q)=(pP+qQ)+(P+pQ)q.
Hint
For the second part, compute f(a+b)=P+Qb and f(a−b) by the same steps.
Solution
By the recurrence in the hint, each power xm is Pm+Qmq with rational Pm, Qm, and a rational linear combination of such expressions has the same form.
Let f have rational coefficients. The computation of f(a+b)=P+Qb uses only rational arithmetic and the rule (b)2=b; running the identical computation with −b (which also squares to b) gives f(a−b)=P−Qb. If f(a+b)=0, then P+Qb=0+0b, so by the theorem (uniqueness of P, Q, as b is irrational) P=Q=0. Hence f(a−b)=0 too.
ExerciseRationalising a denominator●○○Hardy: Examples VIII, 8
Write 1/(p+q) in the form P+Qq, and compute 1/(3+2).
Hint
Multiply top and bottom by the conjugate p−q.
Solution
p+q1=(p+q)(p−q)p−q=p2−qp−p2−q1q,
valid when p2=q (always the case if q is irrational and p rational). For example 1/(3+2)=(3−2)/7=73−712. Combined with the previous exercise, every quotient of polynomials in x=p+q can be written P+Qq (Examples VIII, 9).
ExerciseDenesting a square root●●○Hardy: Examples VIII, 10
If p, q and p2−q are positive, show that p+q=x+y with x=21{p+p2−q} and y=21{p−p2−q}. Use it to simplify 3+8.
Hint
Square x+y and compute x+y and xy.
Solution
We have x+y=p and xy=41{p2−(p2−q)}=41q, with x>0 and y≥0 (since p>p2−q). Then
(x+y)2=x+y+2xy=p+2⋅21q=p+q,
and since both x+y and p+q are positive, they are equal. For p=3, q=8: p2−q=1, so x=2, y=1, and 3+8=2+1. The formula gives something simpler than the original only when p2−q is a rational square, which is Examples VIII, 11.
Check yourself
Check yourselfTrue or false?
12 and 27 are similar surds.
Why
True: 12=23 and 27=33, both rational multiples of 3. The theorem on dissimilar surds needs M, N with no common factor and neither a square; 12 and 27 share the factor 3.
Check yourselfChoose one
Given (2+3)3=26+153, what is (2−3)3?
Why
Replace 3 by −3 throughout: the computation used only (3)2=3. Check: 26−153≈26−25.981=0.019, and (2−3)3≈0.2683≈0.019.
Check yourselfEnter a number
With (2+3)3=26+153, compute P2−3Q2 for P=26, Q=15.
Why
676−3⋅225=676−675=1. In general P2−qQ2=(P+Qq)(P−Qq)=(2+3)3(2−3)3=(4−3)3=1.
Check yourselfChoose one
Why does the theorem allow the exception ”B and D are both squares of rationals”?
Why
When the square roots are rational, A+B is just a rational number, and a rational number can be split as “rational plus square root of a square” in many ways: 1+4=2+1=3.
Hardy’s language
Summary
Summary
Surds are similar when they are rational multiples of one surd; M, N with coprime non-square M, N are dissimilar.
If A+B=C+D with A,B,C,D rational, then A=C and B=D, unless both square roots are rational.
Equivalently, P and Q in P+Qq are unique, so q may be replaced by −q in any rational identity.
Hence conjugate surds a±b are roots of the same rational polynomials, and denominators p+q can be rationalised.