Understanding the proof a metrizable connected space has uncountable elements if $|X|geqslant 2$












2












$begingroup$



Exercise: Using the fact $d$ is a metric hence continuous. Prove that if a metrizable space $(X,tau)$ is connected and $X$ has at least 2 points, then $X$ has uncountable number of points.




I found the following answer on Mathstackexchange by Mariano Suárez-Álvarez




Let your points be $a$ and $b$.



Let $lambdain(0,1)$ Suppose there is no point $x$ in your space such that $d(a,x)=lambda d(a,b)$. Then the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are two non-empty open sets which partition the space.



Since we are assuming connectedness, this is impossible.



Therefore, the image of the function $d(a,mathordcdot):Xtomathbb R$ contains the interval $(0,d(a,b))$, and this can only happen if $X$ is uncountable.




Questions:



1) Why does the author uses $lambda$? If instead the author used $d(a,b)$ and all other points were diferent from $b$ the distance would still be different, allowing the disjoint partition of the space.



2)
I know the metric is continuous function into $mathbb{R}$. How does the author deduces that the image contains $(0,d(a,b))$? How does that relate to the previous argument of the partition?



Thanks in advance!










share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
    $endgroup$
    – lulu
    Dec 28 '18 at 13:33








  • 1




    $begingroup$
    @lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
    $endgroup$
    – Pedro Gomes
    Dec 28 '18 at 13:43






  • 1




    $begingroup$
    Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 13:45






  • 1




    $begingroup$
    The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
    $endgroup$
    – metamorphy
    Dec 28 '18 at 13:45








  • 2




    $begingroup$
    The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
    $endgroup$
    – lulu
    Dec 28 '18 at 14:13
















2












$begingroup$



Exercise: Using the fact $d$ is a metric hence continuous. Prove that if a metrizable space $(X,tau)$ is connected and $X$ has at least 2 points, then $X$ has uncountable number of points.




I found the following answer on Mathstackexchange by Mariano Suárez-Álvarez




Let your points be $a$ and $b$.



Let $lambdain(0,1)$ Suppose there is no point $x$ in your space such that $d(a,x)=lambda d(a,b)$. Then the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are two non-empty open sets which partition the space.



Since we are assuming connectedness, this is impossible.



Therefore, the image of the function $d(a,mathordcdot):Xtomathbb R$ contains the interval $(0,d(a,b))$, and this can only happen if $X$ is uncountable.




Questions:



1) Why does the author uses $lambda$? If instead the author used $d(a,b)$ and all other points were diferent from $b$ the distance would still be different, allowing the disjoint partition of the space.



2)
I know the metric is continuous function into $mathbb{R}$. How does the author deduces that the image contains $(0,d(a,b))$? How does that relate to the previous argument of the partition?



Thanks in advance!










share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
    $endgroup$
    – lulu
    Dec 28 '18 at 13:33








  • 1




    $begingroup$
    @lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
    $endgroup$
    – Pedro Gomes
    Dec 28 '18 at 13:43






  • 1




    $begingroup$
    Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 13:45






  • 1




    $begingroup$
    The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
    $endgroup$
    – metamorphy
    Dec 28 '18 at 13:45








  • 2




    $begingroup$
    The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
    $endgroup$
    – lulu
    Dec 28 '18 at 14:13














2












2








2





$begingroup$



Exercise: Using the fact $d$ is a metric hence continuous. Prove that if a metrizable space $(X,tau)$ is connected and $X$ has at least 2 points, then $X$ has uncountable number of points.




I found the following answer on Mathstackexchange by Mariano Suárez-Álvarez




Let your points be $a$ and $b$.



Let $lambdain(0,1)$ Suppose there is no point $x$ in your space such that $d(a,x)=lambda d(a,b)$. Then the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are two non-empty open sets which partition the space.



Since we are assuming connectedness, this is impossible.



Therefore, the image of the function $d(a,mathordcdot):Xtomathbb R$ contains the interval $(0,d(a,b))$, and this can only happen if $X$ is uncountable.




Questions:



1) Why does the author uses $lambda$? If instead the author used $d(a,b)$ and all other points were diferent from $b$ the distance would still be different, allowing the disjoint partition of the space.



2)
I know the metric is continuous function into $mathbb{R}$. How does the author deduces that the image contains $(0,d(a,b))$? How does that relate to the previous argument of the partition?



Thanks in advance!










share|cite|improve this question











$endgroup$





Exercise: Using the fact $d$ is a metric hence continuous. Prove that if a metrizable space $(X,tau)$ is connected and $X$ has at least 2 points, then $X$ has uncountable number of points.




I found the following answer on Mathstackexchange by Mariano Suárez-Álvarez




Let your points be $a$ and $b$.



Let $lambdain(0,1)$ Suppose there is no point $x$ in your space such that $d(a,x)=lambda d(a,b)$. Then the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are two non-empty open sets which partition the space.



Since we are assuming connectedness, this is impossible.



Therefore, the image of the function $d(a,mathordcdot):Xtomathbb R$ contains the interval $(0,d(a,b))$, and this can only happen if $X$ is uncountable.




Questions:



1) Why does the author uses $lambda$? If instead the author used $d(a,b)$ and all other points were diferent from $b$ the distance would still be different, allowing the disjoint partition of the space.



2)
I know the metric is continuous function into $mathbb{R}$. How does the author deduces that the image contains $(0,d(a,b))$? How does that relate to the previous argument of the partition?



Thanks in advance!







general-topology metric-spaces proof-writing proof-explanation






share|cite|improve this question















share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Dec 28 '18 at 13:39







Pedro Gomes

















asked Dec 28 '18 at 13:21









Pedro GomesPedro Gomes

1,8612721




1,8612721








  • 2




    $begingroup$
    I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
    $endgroup$
    – lulu
    Dec 28 '18 at 13:33








  • 1




    $begingroup$
    @lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
    $endgroup$
    – Pedro Gomes
    Dec 28 '18 at 13:43






  • 1




    $begingroup$
    Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 13:45






  • 1




    $begingroup$
    The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
    $endgroup$
    – metamorphy
    Dec 28 '18 at 13:45








  • 2




    $begingroup$
    The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
    $endgroup$
    – lulu
    Dec 28 '18 at 14:13














  • 2




    $begingroup$
    I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
    $endgroup$
    – lulu
    Dec 28 '18 at 13:33








  • 1




    $begingroup$
    @lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
    $endgroup$
    – Pedro Gomes
    Dec 28 '18 at 13:43






  • 1




    $begingroup$
    Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 13:45






  • 1




    $begingroup$
    The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
    $endgroup$
    – metamorphy
    Dec 28 '18 at 13:45








  • 2




    $begingroup$
    The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
    $endgroup$
    – lulu
    Dec 28 '18 at 14:13








2




2




$begingroup$
I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
$endgroup$
– lulu
Dec 28 '18 at 13:33






$begingroup$
I don't understand your questions. The writer's $lambda$ argument appears to make sense, no? It's possible, of course, that you have an alternate argument, though I don't understand what you wrote. That wouldn't invalidate the given argument though. And I can't follow question $2$ at all.
$endgroup$
– lulu
Dec 28 '18 at 13:33






1




1




$begingroup$
@lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
$endgroup$
– Pedro Gomes
Dec 28 '18 at 13:43




$begingroup$
@lulu Thanks for reading! On point 1) I am writing about how would the proof work if $lambda$ was absent. On point 2) there was a flaw I forgot to write the interval. I do not understand how the author deduces the image of $d(a,.):Xtomathbb{R}$ contains the open interval (0,d(a,b)) on the real line. Hope I was a clearer.
$endgroup$
– Pedro Gomes
Dec 28 '18 at 13:43




1




1




$begingroup$
Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
$endgroup$
– Rob Arthan
Dec 28 '18 at 13:45




$begingroup$
Hint: the first paragraph proves that for any $rho in (0, d(a, b))$, there is a point $x$ such that $d(a, x) = rho$ (by taking $lambda = rho/d(a, b)$).
$endgroup$
– Rob Arthan
Dec 28 '18 at 13:45




1




1




$begingroup$
The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
$endgroup$
– metamorphy
Dec 28 '18 at 13:45






$begingroup$
The author takes $rin(0,d(a,b))$ and shows that there exists $x$ such that $d(a,x)=r$.
$endgroup$
– metamorphy
Dec 28 '18 at 13:45






2




2




$begingroup$
The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
$endgroup$
– lulu
Dec 28 '18 at 14:13




$begingroup$
The argument appears to be entirely straight forward. If $0<r<d(a,b)$ then let $lambda = frac r{d(a,b)}$. We note that $0<lambda <1$ so we can apply the writer's argument to deduce the existence of $x$ such that $d(a,x)=lambda times d(a,b)=r$.
$endgroup$
– lulu
Dec 28 '18 at 14:13










2 Answers
2






active

oldest

votes


















2












$begingroup$

I think your confusion arises because the proof you quote only uses continuity of the metric in a rather artificial way. Here are two proofs, the first doesn't really use the continuity and the second does. In both proofs, we assume $X$ is connected and metrizable and has at least two points, say $a neq b$.




  1. Let $lambda in (0, 1)$. I claim that $d(a, z) = lambda d(a, b)$ for some $z = z(lambda)in X$, for, if not, the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are non-empty, open, disjoint and partition $X$, leading to a contradiction because $X$ is connected. Since there are uncountably many $lambda in (0, 1)$, and $d(a, b) neq 0$ there must be uncountable many $z(lambda) in X$. (This is the proof you quote adapted to remove the artificial appeal to continuity of the metric.).

  2. Consider the function $f : X to Bbb{R}$ defined by $f(x) = d(a, x)$. $f$ is continuous and hence the image $I = f[X]$ is a connected subspace of $Bbb{R}$, because $X$ is connected and the continuous image of a connected set is connected. A non-empty connected subset of $Bbb{R}$ is either a single point or an interval. As $f(a) = d(a, a) = 0$ and $f(b) = d(a, b) neq 0$, $I$ is an interval and so contains uncountably many points. Thus the uncountable set $I$ is an image of $X$ and hence $X$ must be uncountable.






share|cite|improve this answer











$endgroup$













  • $begingroup$
    I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:22






  • 1




    $begingroup$
    I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:28










  • $begingroup$
    Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:29












  • $begingroup$
    I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:37










  • $begingroup$
    I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:40



















2












$begingroup$

If it helps, the idea of the argument is two parts.



1) If $f: X to Y$ is a continuous map and $X$ is connected, then so is $f(X)$. It suffices to assume $f$ is surjective (otherwise, replace $Y$ with $f(X)$ with the subspace topology).



If $f$ is surjective, then given any pair of open sets $U_1, U_2 subset Y$ with $U_1 cap U_2 = varnothing$ and $U_1 cup U_2 = Y$, we may take the inverse image to obtain $f^{-1}(U_1)$ and $f^{-1}(U_2)$. These are nonempty if and only of the corresponding $U_i$ is nonempty; furthermore their intersection is trivial and union is all of $X$ (this does not use surjectivity). Therefore, if $U_1 cup U_2 = Y$ is a disconnection of $Y$ (that is, neither $U_i$ is the empty set), then the same is true of $f^{-1}(U_1) cup f^{-1}(U_2) = X$, and so if $Y$ is disconnected so is $X$, as desired.



2) If $I$ is a connected subset of the real line, if $a, b in I$, then $[a, b] subset I$. For if not, choose $x in (a,b)$ not in $I$, and write $I = left((-infty, x) cap Iright) cup left(x, infty) cap Iright)$ to obtain a disconnection of $I$.



Now consider the conditions map $X to [0,infty)$ given by $x mapsto d(a,x)$. This has image containing both $d(a,a) = 0$ and $d(a,b) neq 0$. So therefore the image must contain all of $[0, d(a,b)]$ and hence must be uncountable; so the domain is as well.



Mariano rolled these two points into one proof: he took the open sets $U_1$ and $U_2$ in the first step to be $(0, x)$ and $(x, d(a,b))$ if $x$ is not in the image.






share|cite|improve this answer









$endgroup$













  • $begingroup$
    I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:15










  • $begingroup$
    @RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:19










  • $begingroup$
    I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:24










  • $begingroup$
    @RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:25











Your Answer





StackExchange.ifUsing("editor", function () {
return StackExchange.using("mathjaxEditing", function () {
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
});
});
}, "mathjax-editing");

StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "69"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});

function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});


}
});














draft saved

draft discarded


















StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3054881%2funderstanding-the-proof-a-metrizable-connected-space-has-uncountable-elements-if%23new-answer', 'question_page');
}
);

Post as a guest















Required, but never shown

























2 Answers
2






active

oldest

votes








2 Answers
2






active

oldest

votes









active

oldest

votes






active

oldest

votes









2












$begingroup$

I think your confusion arises because the proof you quote only uses continuity of the metric in a rather artificial way. Here are two proofs, the first doesn't really use the continuity and the second does. In both proofs, we assume $X$ is connected and metrizable and has at least two points, say $a neq b$.




  1. Let $lambda in (0, 1)$. I claim that $d(a, z) = lambda d(a, b)$ for some $z = z(lambda)in X$, for, if not, the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are non-empty, open, disjoint and partition $X$, leading to a contradiction because $X$ is connected. Since there are uncountably many $lambda in (0, 1)$, and $d(a, b) neq 0$ there must be uncountable many $z(lambda) in X$. (This is the proof you quote adapted to remove the artificial appeal to continuity of the metric.).

  2. Consider the function $f : X to Bbb{R}$ defined by $f(x) = d(a, x)$. $f$ is continuous and hence the image $I = f[X]$ is a connected subspace of $Bbb{R}$, because $X$ is connected and the continuous image of a connected set is connected. A non-empty connected subset of $Bbb{R}$ is either a single point or an interval. As $f(a) = d(a, a) = 0$ and $f(b) = d(a, b) neq 0$, $I$ is an interval and so contains uncountably many points. Thus the uncountable set $I$ is an image of $X$ and hence $X$ must be uncountable.






share|cite|improve this answer











$endgroup$













  • $begingroup$
    I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:22






  • 1




    $begingroup$
    I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:28










  • $begingroup$
    Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:29












  • $begingroup$
    I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:37










  • $begingroup$
    I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:40
















2












$begingroup$

I think your confusion arises because the proof you quote only uses continuity of the metric in a rather artificial way. Here are two proofs, the first doesn't really use the continuity and the second does. In both proofs, we assume $X$ is connected and metrizable and has at least two points, say $a neq b$.




  1. Let $lambda in (0, 1)$. I claim that $d(a, z) = lambda d(a, b)$ for some $z = z(lambda)in X$, for, if not, the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are non-empty, open, disjoint and partition $X$, leading to a contradiction because $X$ is connected. Since there are uncountably many $lambda in (0, 1)$, and $d(a, b) neq 0$ there must be uncountable many $z(lambda) in X$. (This is the proof you quote adapted to remove the artificial appeal to continuity of the metric.).

  2. Consider the function $f : X to Bbb{R}$ defined by $f(x) = d(a, x)$. $f$ is continuous and hence the image $I = f[X]$ is a connected subspace of $Bbb{R}$, because $X$ is connected and the continuous image of a connected set is connected. A non-empty connected subset of $Bbb{R}$ is either a single point or an interval. As $f(a) = d(a, a) = 0$ and $f(b) = d(a, b) neq 0$, $I$ is an interval and so contains uncountably many points. Thus the uncountable set $I$ is an image of $X$ and hence $X$ must be uncountable.






share|cite|improve this answer











$endgroup$













  • $begingroup$
    I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:22






  • 1




    $begingroup$
    I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:28










  • $begingroup$
    Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:29












  • $begingroup$
    I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:37










  • $begingroup$
    I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:40














2












2








2





$begingroup$

I think your confusion arises because the proof you quote only uses continuity of the metric in a rather artificial way. Here are two proofs, the first doesn't really use the continuity and the second does. In both proofs, we assume $X$ is connected and metrizable and has at least two points, say $a neq b$.




  1. Let $lambda in (0, 1)$. I claim that $d(a, z) = lambda d(a, b)$ for some $z = z(lambda)in X$, for, if not, the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are non-empty, open, disjoint and partition $X$, leading to a contradiction because $X$ is connected. Since there are uncountably many $lambda in (0, 1)$, and $d(a, b) neq 0$ there must be uncountable many $z(lambda) in X$. (This is the proof you quote adapted to remove the artificial appeal to continuity of the metric.).

  2. Consider the function $f : X to Bbb{R}$ defined by $f(x) = d(a, x)$. $f$ is continuous and hence the image $I = f[X]$ is a connected subspace of $Bbb{R}$, because $X$ is connected and the continuous image of a connected set is connected. A non-empty connected subset of $Bbb{R}$ is either a single point or an interval. As $f(a) = d(a, a) = 0$ and $f(b) = d(a, b) neq 0$, $I$ is an interval and so contains uncountably many points. Thus the uncountable set $I$ is an image of $X$ and hence $X$ must be uncountable.






share|cite|improve this answer











$endgroup$



I think your confusion arises because the proof you quote only uses continuity of the metric in a rather artificial way. Here are two proofs, the first doesn't really use the continuity and the second does. In both proofs, we assume $X$ is connected and metrizable and has at least two points, say $a neq b$.




  1. Let $lambda in (0, 1)$. I claim that $d(a, z) = lambda d(a, b)$ for some $z = z(lambda)in X$, for, if not, the sets ${z:d(a,z)<lambda d(a,b)}$ and ${z:d(a,z)>lambda d(a,b)}$ are non-empty, open, disjoint and partition $X$, leading to a contradiction because $X$ is connected. Since there are uncountably many $lambda in (0, 1)$, and $d(a, b) neq 0$ there must be uncountable many $z(lambda) in X$. (This is the proof you quote adapted to remove the artificial appeal to continuity of the metric.).

  2. Consider the function $f : X to Bbb{R}$ defined by $f(x) = d(a, x)$. $f$ is continuous and hence the image $I = f[X]$ is a connected subspace of $Bbb{R}$, because $X$ is connected and the continuous image of a connected set is connected. A non-empty connected subset of $Bbb{R}$ is either a single point or an interval. As $f(a) = d(a, a) = 0$ and $f(b) = d(a, b) neq 0$, $I$ is an interval and so contains uncountably many points. Thus the uncountable set $I$ is an image of $X$ and hence $X$ must be uncountable.







share|cite|improve this answer














share|cite|improve this answer



share|cite|improve this answer








edited Dec 28 '18 at 20:16

























answered Dec 28 '18 at 20:04









Rob ArthanRob Arthan

29.3k42966




29.3k42966












  • $begingroup$
    I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:22






  • 1




    $begingroup$
    I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:28










  • $begingroup$
    Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:29












  • $begingroup$
    I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:37










  • $begingroup$
    I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:40


















  • $begingroup$
    I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:22






  • 1




    $begingroup$
    I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:28










  • $begingroup$
    Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:29












  • $begingroup$
    I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:37










  • $begingroup$
    I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:40
















$begingroup$
I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:22




$begingroup$
I think that (2) is essentially just (1), because you need to prove that connected sets are intervals. The proof in (1) includes the fact that any interval which contains two points contains the entire interval. This is the proof I would use to check the stated property. Also, (1) absolutely needs continuity of the metric, otherwise there is no reason to believe that the inverse image of these open sets are open.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:22




1




1




$begingroup$
I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:28




$begingroup$
I've just written a comment on your answer disagreeing with the comment that you were writing at the same time $ddot{smile}$. That those two inverse images are open follows from first principles from the definition of open sets wrt a metric. I am very happy to agree to differ: the important thing is that the OP understands what's going on. Cheers!
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:28












$begingroup$
Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:29






$begingroup$
Ok, sure, I grant my second point :) But I think that we are just playing word games, since the definition of open set encodes continuity of the metric. And of course I agree this sort of disagreement is just for fun.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:29














$begingroup$
I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:37




$begingroup$
I am not sure it is entirely about word games. The statement that the metric is continuous presupposes an understanding of the product topology, whereas the problem and the proof here don't require that.
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:37












$begingroup$
I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
$endgroup$
– Mike Miller
Dec 28 '18 at 20:40




$begingroup$
I really only meant the function $x mapsto d(a, x)$, which is all either of us use. anyway, I'll let you spend your time on something more productive!
$endgroup$
– Mike Miller
Dec 28 '18 at 20:40











2












$begingroup$

If it helps, the idea of the argument is two parts.



1) If $f: X to Y$ is a continuous map and $X$ is connected, then so is $f(X)$. It suffices to assume $f$ is surjective (otherwise, replace $Y$ with $f(X)$ with the subspace topology).



If $f$ is surjective, then given any pair of open sets $U_1, U_2 subset Y$ with $U_1 cap U_2 = varnothing$ and $U_1 cup U_2 = Y$, we may take the inverse image to obtain $f^{-1}(U_1)$ and $f^{-1}(U_2)$. These are nonempty if and only of the corresponding $U_i$ is nonempty; furthermore their intersection is trivial and union is all of $X$ (this does not use surjectivity). Therefore, if $U_1 cup U_2 = Y$ is a disconnection of $Y$ (that is, neither $U_i$ is the empty set), then the same is true of $f^{-1}(U_1) cup f^{-1}(U_2) = X$, and so if $Y$ is disconnected so is $X$, as desired.



2) If $I$ is a connected subset of the real line, if $a, b in I$, then $[a, b] subset I$. For if not, choose $x in (a,b)$ not in $I$, and write $I = left((-infty, x) cap Iright) cup left(x, infty) cap Iright)$ to obtain a disconnection of $I$.



Now consider the conditions map $X to [0,infty)$ given by $x mapsto d(a,x)$. This has image containing both $d(a,a) = 0$ and $d(a,b) neq 0$. So therefore the image must contain all of $[0, d(a,b)]$ and hence must be uncountable; so the domain is as well.



Mariano rolled these two points into one proof: he took the open sets $U_1$ and $U_2$ in the first step to be $(0, x)$ and $(x, d(a,b))$ if $x$ is not in the image.






share|cite|improve this answer









$endgroup$













  • $begingroup$
    I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:15










  • $begingroup$
    @RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:19










  • $begingroup$
    I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:24










  • $begingroup$
    @RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:25
















2












$begingroup$

If it helps, the idea of the argument is two parts.



1) If $f: X to Y$ is a continuous map and $X$ is connected, then so is $f(X)$. It suffices to assume $f$ is surjective (otherwise, replace $Y$ with $f(X)$ with the subspace topology).



If $f$ is surjective, then given any pair of open sets $U_1, U_2 subset Y$ with $U_1 cap U_2 = varnothing$ and $U_1 cup U_2 = Y$, we may take the inverse image to obtain $f^{-1}(U_1)$ and $f^{-1}(U_2)$. These are nonempty if and only of the corresponding $U_i$ is nonempty; furthermore their intersection is trivial and union is all of $X$ (this does not use surjectivity). Therefore, if $U_1 cup U_2 = Y$ is a disconnection of $Y$ (that is, neither $U_i$ is the empty set), then the same is true of $f^{-1}(U_1) cup f^{-1}(U_2) = X$, and so if $Y$ is disconnected so is $X$, as desired.



2) If $I$ is a connected subset of the real line, if $a, b in I$, then $[a, b] subset I$. For if not, choose $x in (a,b)$ not in $I$, and write $I = left((-infty, x) cap Iright) cup left(x, infty) cap Iright)$ to obtain a disconnection of $I$.



Now consider the conditions map $X to [0,infty)$ given by $x mapsto d(a,x)$. This has image containing both $d(a,a) = 0$ and $d(a,b) neq 0$. So therefore the image must contain all of $[0, d(a,b)]$ and hence must be uncountable; so the domain is as well.



Mariano rolled these two points into one proof: he took the open sets $U_1$ and $U_2$ in the first step to be $(0, x)$ and $(x, d(a,b))$ if $x$ is not in the image.






share|cite|improve this answer









$endgroup$













  • $begingroup$
    I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:15










  • $begingroup$
    @RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:19










  • $begingroup$
    I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:24










  • $begingroup$
    @RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:25














2












2








2





$begingroup$

If it helps, the idea of the argument is two parts.



1) If $f: X to Y$ is a continuous map and $X$ is connected, then so is $f(X)$. It suffices to assume $f$ is surjective (otherwise, replace $Y$ with $f(X)$ with the subspace topology).



If $f$ is surjective, then given any pair of open sets $U_1, U_2 subset Y$ with $U_1 cap U_2 = varnothing$ and $U_1 cup U_2 = Y$, we may take the inverse image to obtain $f^{-1}(U_1)$ and $f^{-1}(U_2)$. These are nonempty if and only of the corresponding $U_i$ is nonempty; furthermore their intersection is trivial and union is all of $X$ (this does not use surjectivity). Therefore, if $U_1 cup U_2 = Y$ is a disconnection of $Y$ (that is, neither $U_i$ is the empty set), then the same is true of $f^{-1}(U_1) cup f^{-1}(U_2) = X$, and so if $Y$ is disconnected so is $X$, as desired.



2) If $I$ is a connected subset of the real line, if $a, b in I$, then $[a, b] subset I$. For if not, choose $x in (a,b)$ not in $I$, and write $I = left((-infty, x) cap Iright) cup left(x, infty) cap Iright)$ to obtain a disconnection of $I$.



Now consider the conditions map $X to [0,infty)$ given by $x mapsto d(a,x)$. This has image containing both $d(a,a) = 0$ and $d(a,b) neq 0$. So therefore the image must contain all of $[0, d(a,b)]$ and hence must be uncountable; so the domain is as well.



Mariano rolled these two points into one proof: he took the open sets $U_1$ and $U_2$ in the first step to be $(0, x)$ and $(x, d(a,b))$ if $x$ is not in the image.






share|cite|improve this answer









$endgroup$



If it helps, the idea of the argument is two parts.



1) If $f: X to Y$ is a continuous map and $X$ is connected, then so is $f(X)$. It suffices to assume $f$ is surjective (otherwise, replace $Y$ with $f(X)$ with the subspace topology).



If $f$ is surjective, then given any pair of open sets $U_1, U_2 subset Y$ with $U_1 cap U_2 = varnothing$ and $U_1 cup U_2 = Y$, we may take the inverse image to obtain $f^{-1}(U_1)$ and $f^{-1}(U_2)$. These are nonempty if and only of the corresponding $U_i$ is nonempty; furthermore their intersection is trivial and union is all of $X$ (this does not use surjectivity). Therefore, if $U_1 cup U_2 = Y$ is a disconnection of $Y$ (that is, neither $U_i$ is the empty set), then the same is true of $f^{-1}(U_1) cup f^{-1}(U_2) = X$, and so if $Y$ is disconnected so is $X$, as desired.



2) If $I$ is a connected subset of the real line, if $a, b in I$, then $[a, b] subset I$. For if not, choose $x in (a,b)$ not in $I$, and write $I = left((-infty, x) cap Iright) cup left(x, infty) cap Iright)$ to obtain a disconnection of $I$.



Now consider the conditions map $X to [0,infty)$ given by $x mapsto d(a,x)$. This has image containing both $d(a,a) = 0$ and $d(a,b) neq 0$. So therefore the image must contain all of $[0, d(a,b)]$ and hence must be uncountable; so the domain is as well.



Mariano rolled these two points into one proof: he took the open sets $U_1$ and $U_2$ in the first step to be $(0, x)$ and $(x, d(a,b))$ if $x$ is not in the image.







share|cite|improve this answer












share|cite|improve this answer



share|cite|improve this answer










answered Dec 28 '18 at 20:02









Mike MillerMike Miller

37.6k473140




37.6k473140












  • $begingroup$
    I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:15










  • $begingroup$
    @RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:19










  • $begingroup$
    I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:24










  • $begingroup$
    @RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:25


















  • $begingroup$
    I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:15










  • $begingroup$
    @RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:19










  • $begingroup$
    I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
    $endgroup$
    – Rob Arthan
    Dec 28 '18 at 20:24










  • $begingroup$
    @RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
    $endgroup$
    – Mike Miller
    Dec 28 '18 at 20:25
















$begingroup$
I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:15




$begingroup$
I find it hard to fit your explanation to the quoted proof, which contains no instance of the assertion that the continuous image of a connected set is connected (your part (1)).
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:15












$begingroup$
@RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:19




$begingroup$
@RobArthan not explicitly, but the entire point of the proof is to show that a continuous map $f: X to Bbb R$ from connected $X$ which contains $a$ and $b$ indeed contains $[a, b]$. To do that you are implicitly showing that $f(X)$ is path-connected (hence connected). Of course, Mariano's proof does both in one step. This is what my last line was supposed to indicate.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:19












$begingroup$
I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:24




$begingroup$
I don't agree. See my answer: my proof (1) is the real essence of Mariano's proof and does not involve what you claim is the "entire point of the proof". Personally, I prefer my proof (2), which is the same as the proof you give, but I can see why the OP might be confused given a (correct, but over-complicated) mixture of the two proofs.
$endgroup$
– Rob Arthan
Dec 28 '18 at 20:24












$begingroup$
@RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:25




$begingroup$
@RobArthan I disagree that there is any difference in the idea in the two proofs, and we will have to leave it at that.
$endgroup$
– Mike Miller
Dec 28 '18 at 20:25


















draft saved

draft discarded




















































Thanks for contributing an answer to Mathematics Stack Exchange!


  • Please be sure to answer the question. Provide details and share your research!

But avoid



  • Asking for help, clarification, or responding to other answers.

  • Making statements based on opinion; back them up with references or personal experience.


Use MathJax to format equations. MathJax reference.


To learn more, see our tips on writing great answers.




draft saved


draft discarded














StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3054881%2funderstanding-the-proof-a-metrizable-connected-space-has-uncountable-elements-if%23new-answer', 'question_page');
}
);

Post as a guest















Required, but never shown





















































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown

































Required, but never shown














Required, but never shown












Required, but never shown







Required, but never shown







Popular posts from this blog

Måne

Storängen

VLT Carioca