2

I need help for the following task. Unfortunately I don't have any idea how to start.

a) Prove: A field $K$ is an ordered $\leftrightarrow$ Subset $P \subseteq K$ exists with following features:

  1. $\forall x,y \in P$ : $x+y \in P$

  2. $\forall x,y \in P : x\cdot y \in P$

  3. $\forall x \in K$ applies exactly one of these relationships:

$x = 0$, $x\in -P$, $x \in P$

Edit: A field is called ordered if it satisfies all these properties:

  1. $\forall x,y \in \mathbb {R}$ applies exactly one of these relationships: $x=y, x<y$, or $x>y$ (trichotomy)

  2. $\forall x,y,z \in \mathbb{R}$ applies $x<y,y<z \rightarrow x<z$ (transitive relation)

  3. $\forall x,y,z \in \mathbb{R}$ applies: $x<y \rightarrow x+z<y+z$

  4. $\forall x,y,z \in \mathbb{R}$ applies: $x<y,0<z \rightarrow xz<yz$ (I don't know the english names for these properties)

b) Let K be a field of rational expressions with real coefficients, that means expressions of the form $\frac{p}{q}$ where $p$ and $q\neq0$ are polynoms with real coefficients. Show that $K$ is ordered. (Hint: Use a) with $P$={$\frac{a_0 +a_1x+...+a_{n-1}x^{n-1}+a_nx^n}{b_0+b_1x+...+b_{n-1}x^{m-1}+b_mx^m}$ $\in K$ : $a_n \cdot b_m > 0, m, n \in \mathbb{N}$

My approach for b: https://i.stack.imgur.com/tafOz.png

I am thankful for any advice.

  • @OliverHouse Yeah exactly. We have to prove that K is an ordered field implicates the right side and the right side implicates the left side. – Analysis_Mark Feb 11 '22 at 01:32
  • @OliverHouse The task is definetly true since it is from an exam from my professor. Maybe we could talk about, how to prove that the left direction implicates the right direction? Or you could point out your prove of the contradiction by showing that $P \subseteq \mathbb{Z}$ satisfies all 3 properties. It would help me to understand your thought processes. – Analysis_Mark Feb 11 '22 at 01:53
  • What is an ordered field to you? – azif00 Feb 11 '22 at 02:25
  • @Analysis_Mark If $K$ is an ordered field, then the strictly positive elements satisfy all 3 properties; this can be proved from the basic properties of an ordered field. On the other hand $\mathbb{Z}$ is not an ordered field, yet the positive integers satisfy all three properties. However, if $K$ is a field, then I think it works. It would help if you add the precise definition of an ordered field to your question, since it is presented differently in different texts – user829347 Feb 11 '22 at 03:01
  • @OliverHouse azif00 I updated the question. – Analysis_Mark Feb 11 '22 at 08:26
  • I think you definitely need an initial statement "Let $K$ be a field, then ...". And then it's surely trivial: for one way define $P={x\mid x>0}$, for the other define $x>y$ to mean $x-y\in P$. – ancient mathematician Feb 11 '22 at 10:22
  • @Analysis_Mark Your definition of an ordered field included in your edit is unfortunately quite flawed. Real numbers don't come into it at all (although $\mathbb{R}$ is one of the most important examples of an ordered field) – user829347 Feb 11 '22 at 14:50

1 Answers1

1

The question has essentially been answered in the comments, but I shall write it out explicitly for the sake of completeness.

It seems very likely that your professor wanted you to demonstrate that there are two different ways to understand ordered fields: firstly by means of a total order compatible with the field structure, and secondly by means of a 'positive cone' $P$ that tells you what the 'positive' elements should be. It looks like you (or your professor) missed out a crucial detail. It doesn't really make much sense to say that '$K$ is an ordered field if and only if' it satisfies some given properties, since there is no indication of what exactly $K$ is to start with.

It's a bit like saying that a 'thing' is an ordered field if it has said properties, without defining what this 'thing' is and what it what it means for it to have or not have these properties. There always has to be a clear, well-defined starting point for such questions, and in this case your starting point is assuming that $K$ is a field (which rules out $\mathbb{Z}$, as I mentioned in the comments).

Here is the 'field order' definition (field order is my own terminology and not in standard use so far as I am aware):

A field order on a field $K$ is a total order $\leq$ on $K$ satisfying

(i) If $x,y,z\in K$ and $x\leq y$, then $x+z\leq y+z$

(ii) If $x,y,z\in K$ and $0\leq z$, then $xz\leq yz$.

An ordered field is a field $K$ equipped with a field order.

Here is the 'positive cone' definition:

A positive cone for a field $K$ is a subset $P\subseteq K$ satisfying

(i) $x,y\in P\implies x+y\in P$

(ii) $x,y\in P\implies xy\in P$

(iii) If $x\in P$ and $-x\in P$, then $x=0$.

An ordered field is a field $K$ equipped with a positive cone $P$.

It is then fairly routine to check that there is a one-to-one correspondence between field orders on $K$ and positive cones of $K$:

The field order $\leq$ corresponds to the positive cone $\{x\in K:0\leq x\}$. The positive cone $P$ corresponds to the field order defined by declaring that $x\leq y$ if and only if $y-x\in P$.


EDIT: If $K$ is an ordered field (by the first definition), then we show that $P:=\{x\in K:0\leq x\}$ is closed under addition as follows:

Let $x,y\in P$. Then $0\leq y$ and using (i), we have that $0\leq x\implies 0+y=y\leq x+y$. So $0\leq x+y$ by transitivity and then $x+y\in P$.


See $\S$ 23.1 of Stewart, I., Galois Theory (Fourth Edition), CRC Press, 2015 and the Wikipedia article Ordered Field.

user829347
  • 3,402
  • Yeah you are right. The task was a field K is ordered if and only.... Thanks for your answer. Unfortunately I am still unsure how to prove that. For example Why does an ordered field implicates that $\forall x,y \in P$ $x+y \in P$?

    I also updated the question with another task b, since it is connected with the first task. Could you maybe take a look at it?

    – Analysis_Mark Feb 12 '22 at 18:25
  • I've added a proof that $P$ is closed under addition to my answer. I'd suggest posting a new question for part (b) with a link to this one explaining how they are related (the rules are a little rigid here sometimes) I'll take a look at it if I see it (which is likely) – user829347 Feb 12 '22 at 19:43
  • All right. To a). Why can we say that K is a field of positive elements. We don't know that. The problem is also, that our professor did never teach us anything about positive cone. Thats why I believe that there must be another way. But we learned about this lemma: Let $M$ be a field with $(+,\cdot)$ and $N$ a subset of M. If
    1. $N$is closed under addition and multiplication
    2. 0 $\in N$ and 1 $\in N$
    3. $N$ is closedin terms ofthe inverseelement: $a \in N$ $\rightarrow$ $-a \in N$; $a$ $\in$ $N{0}$ $\rightarrow$ $a^{-1} \in N$ Then N is a field with +,$\cdot$. We call N a subfield of M.
    – Analysis_Mark Feb 12 '22 at 20:27
  • @Analysis_Mark You've misunderstood. An ordered field $K$ is not a 'field of positive elements', it simply contains positive elements. In any ordered field, we have the total order $\leq$ and the additive identity $0_K$. The positive elements are those $x\in K$ for which $x\geq 0$ (equivalently, $0\leq x$). Using the properties of the field order given in my answer, one can check that the positive elements form a positive cone (as defined in my answer). Your 'lemma' is what most people I think would say is the definition of an ordered field, rather than a property to derive from it. – user829347 Feb 12 '22 at 20:34
  • But why are we allowed to say that P contains the positive elements of K? Also here is the link for part b: https://math.stackexchange.com/questions/4380584/show-that-a-field-of-rational-expressions-with-real-coefficients-is-ordered – Analysis_Mark Feb 12 '22 at 20:42
  • In my edit, I defined $P$ to be precisely the positive elements of $K$. I used $P$ just so I don't have to write the full thing out over and over again. Those $x\in K$ for which $x\geq 0$ form a positive cone for $K$ – user829347 Feb 12 '22 at 21:26
  • I think I made the proof from left to right correct. A small question: We can't define $P:={x \in K: 0\geq x}$ right? We have to say that x>0 is that correct? Because we have to show that if $x\in K$, then its either $x=0$, or $x \in P$ but not both. If x=0 would be in P, then x=0 would be zero and in P. With regard to the proof from right to left, I still have problems. Could you show me aa approach why for example $x,y \in P$ with $x+y \in P$ should implicate, that K is ordered. I mean yeah we show that $y\geq y$ implicates $y+x\geq 0$. But why should this be also valid $for x,y \in K$ – Analysis_Mark Feb 13 '22 at 09:30
  • @Analysis_Mark Your professor might be using a strict order, which is a subtly different approach. I'd suggest asking him/her precisely what definition of an ordered field should be used in your course and then including it here because it is very difficult to know how to help you when there are various different approaches. Your current definition of an ordered field in your question (as I have said before) is wrong, it does not involve $\mathbb{R}$. The 'ordered field' wiki is likely the definition your professor is using, but the proof is not much different either way. – user829347 Feb 13 '22 at 16:21
  • The definition of an ordered field that I wrote down is from my professor's script. But yeah I am sure, the proof isn't different either way. The problem is, I just can't see how to show that if $x,y \in P$ with $x+y \in P$ could implicate that K is an ordered field.

    Also I updated the post with a linked image of my approach to b. Could you check that?

    – Analysis_Mark Feb 13 '22 at 19:18