1: \begin{abstract}

2:   %

3:   %

4:   %

5:   %

6:   %

7:   %

8:   %

9:   Weak convergence of maxima of dependent sequences of identically distributed

10:   continuous random variables is studied under normalizing sequences arising as

11:   subsequences of the normalizing sequences from an associated

12:   iid sequence. This general framework allows one to derive several

13:   generalizations of the well-known Fisher--Tippett--Gnedenko theorem under

14:   conditions on the univariate marginal distribution and the dependence

15:   structure of the sequence. The limiting distributions are shown to be

16:   compositions of a generalized extreme value distribution and a

17:   distortion function which reflects the limiting behavior of the diagonal of

18:   the underlying copula. Uniform convergence rates for the weak convergence

19:   to the limiting distribution are also derived. Examples covering

20:   well-known dependence structures are provided. Several existing results, e.g.\

21:   for exchangeable sequences or stationary time series, are embedded in the

22:   proposed framework.

23: \end{abstract}

24: