A Question Of Degree

If a a and b b are two algebraic numbers of degree 2 and 3, respectively, what is the least possible degree of a × b a\times b ?


Inspiration .

3 4 5 6 1 2

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Mark Hennings
Mar 6, 2016

We have intermediate fields Q Q ( a ) , Q ( b ) , Q ( a b ) Q ( a , b ) \mathbb{Q} \,\le\, \mathbb{Q}(a)\,,\,\mathbb{Q}(b)\,,\,\mathbb{Q}(ab) \,\le\, \mathbb{Q}(a,b) , with [ Q ( a ) : Q ] = 2 [\mathbb{Q}(a):\mathbb{Q}] = 2 and [ Q ( b ) : Q ] = 3 [\mathbb{Q}(b):\mathbb{Q}] = 3 . Since the minimum polynomial of a a over Q ( b ) \mathbb{Q}(b) must divide the minimum polynomial of a a over Q \mathbb{Q} , it follows that [ Q ( a , b ) : Q ( b ) ] [ Q ( a ) : Q ] = 2 [\mathbb{Q}(a,b):\mathbb{Q}(b)] \,\le\, [\mathbb{Q}(a):\mathbb{Q}] = 2 , and hence [ Q ( a , b ) : Q ] = [ Q ( a , b ) : Q ( b ) ] [ Q ( b ) : Q ] 6 , [\mathbb{Q}(a,b):\mathbb{Q}] \; = \; [\mathbb{Q}(a,b):\mathbb{Q}(b)] \,[\mathbb{Q}(b):\mathbb{Q}] \; \le \; 6 \;, while both 2 = [ Q ( a ) : Q ] 2 = [\mathbb{Q}(a):\mathbb{Q}] and 3 = [ Q ( b ) : Q ] 3 = [\mathbb{Q}(b):\mathbb{Q}] divide [ Q ( a , b ) : Q ] [\mathbb{Q}(a,b):\mathbb{Q}] . Thus [ Q ( a , b ) : Q ] = 6 [\mathbb{Q}(a,b):\mathbb{Q}] = 6 , and so it follows that [ Q ( a b ) : Q ] [\mathbb{Q}(ab):\mathbb{Q}] divides 6 6 .

If [ Q ( a b ) : Q ] 2 [\mathbb{Q}(ab):\mathbb{Q}] \le 2 then, since [ Q ( a , b ) : Q ( a b ) ] [ Q ( a ) : Q ] = 2 [\mathbb{Q}(a,b):\mathbb{Q}(ab)] \,\le\, [\mathbb{Q}(a):\mathbb{Q}] \,=\, 2 , we deduce that [ Q ( a , b ) : Q ] = [ Q ( a , b ) : Q ( a b ) ] [ Q ( a b ) : Q ] [\mathbb{Q}(a,b):\mathbb{Q}] = [\mathbb{Q}(a,b):\mathbb{Q}(ab)]\,[\mathbb{Q}(ab):\mathbb{Q}] would have to be one of 1 1 , 2 2 or 4 4 , which is not the case. Thus it follows that [ Q ( a b ) : Q ] 3 [\mathbb{Q}(ab):\mathbb{Q}] \,\ge\, 3 , and so [ Q ( a b ) : Q ] [\mathbb{Q}(ab):\mathbb{Q}] can be either 3 3 or 6 6 .

Cue the standard example of a = ω a = \omega , b = 2 3 b = \sqrt[3]{2} to see that a degree of 3 3 is possible.

Nice detailed solution! (+1)

Otto Bretscher - 5 years, 3 months ago
Otto Bretscher
Mar 5, 2016

First we give an example where the degree of a b ab is 3. Let a = e 2 π i / 3 a=e^{2\pi i/3} and b = 2 3 b=\sqrt[3]{2} , with minimal polynomials x 2 + x + 1 x^2+x+1 and x 3 2 x^3-2 , respectively. Then a b = e 2 π i / 3 2 3 ab=e^{2\pi i/3}\sqrt[3]{2} is another root of x 3 2 x^3-2 .

Now [ Q ( a , b ) : Q ( b ) ] = 2 [\mathbb{Q}(a,b):\mathbb{Q}(b)]=2 so [ Q ( a , b ) : Q ] = 6 [\mathbb{Q}(a,b):\mathbb{Q}]=6 so [ Q ( a , b ) : Q ( a ) ] = [ Q ( a b , a ) : Q ( a ) ] = 3 [\mathbb{Q}(a,b):\mathbb{Q}(a)]= [\mathbb{Q}(ab,a):\mathbb{Q}(a)]= 3 so [ Q ( a b ) : Q ] 3 [\mathbb{Q}(ab):\mathbb{Q}]\geq 3 since the minimal polynomial of a b ab over Q \mathbb{Q} may factor over Q ( a ) \mathbb{Q}(a) .

Thus the least possible degree of a b ab is 3 \boxed{3}

As far as I know, an algebraic number must be a real number. Maybe clarify that you allow complex numbers as well?

EDIT: Apparently an algebraic number can be complex (from Wikipedia, at least). The more you know.

Ivan Koswara - 5 years, 3 months ago

Log in to reply

Yes, exactly ;) I have never seen the definition of the algebraic numbers restricted to the reals. It really would not make any sense; you need an algebraically closed field to do good number theory (for splitting fields etc.).

Otto Bretscher - 5 years, 3 months ago

Log in to reply

I suppose I thought of that definition because usually I see algebraic numbers being a subset of the real numbers, hence why.

Ivan Koswara - 5 years, 3 months ago

Log in to reply

@Ivan Koswara Have you ever seen a definition in print where the algebraic numbers were given as a subset of the reals?

Otto Bretscher - 5 years, 3 months ago

Log in to reply

@Otto Bretscher Not really in print, but I think I recall something like "we extend naturals to integers to have solutions for x + 2 = 1 x+2 = 1 ; we extend integers to rationals to have solutions for x 2 = 1 x \cdot 2 = 1 ; if we want to extend the rationals to have solutions for x x = 2 x \cdot x = 2 , we only need to extend it to algebraics, not reals" explained in a lecture, hence why.

Ivan Koswara - 5 years, 3 months ago

It's impressive

Duy Anh Tran Le - 5 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...