0
$\begingroup$

I read in Munkres's topology about Seifert Van-Kampen , classical verson . Assume the hypotheses of the preceding theorem ( Seifert Van-Kampen ) . Let $j : \pi_{1}(U,x_{0}) * \pi_{1}(V , x_{0}) \to \pi_{1}(X,x_{0})$ . But i'm not sure $\pi(U)*\pi(V)$ is well-defined . Anybody explain it for me ? Thanks .

  • 0
    It is well defined. You can take the free product of any collection of groups. May be you are thinking the map is not well defined or it is not one-one?2017-01-23
  • 0
    We can define free product if any pair in family group intersect at 1-point and the representation of $1$ by empty word is unique ? Maybe I confuse ?2017-01-23
  • 0
    Are you sure that there are no more conditions on $U$ and $V$? They intersect at at least one point, which is $x_0$, so the 1-point intersection is done over with. I do not understand the representation of $1$ part.2017-01-23
  • 0
    $U,V,U \cap V$ are path-connected and $U,V$ are open in $X$ . The representation of $1$ is empty word . You can define free product of a family group $G_{\alpha}$ if reduce word of identity element is unique .2017-01-23
  • 0
    You can assume the reduced word of identity element is unique and proceed.2017-01-23
  • 0
    It's matter . I just know that the map induced by inclusion map from fundamental group of $U,V$ to $X$ is generated $\pi(X)$ . But I'm not sure any alternating production of some paths is unique .2017-01-23
  • 0
    Yes, but this page : https://en.wikipedia.org/wiki/Normal_form_for_free_groups_and_free_product_of_groups says that every word, including identity word, has unique reduced form.2017-01-23
  • 0
    I think we have a problem such that Munkres's definition is intrinsic definition but by the usual way we can always define free products . Ok , I understood and have no question more .2017-01-23
  • 0
    Ok great. Always happy to help.2017-01-23

0 Answers 0