Class field theory (Q5901222): Difference between revisions

Class field theory studies abelian extensions of number fields and their completions. It has nowadays become standard to develop global class field theory by proving the two fundamental inequalities, and then derive the main theorems from Artin's reciprocity law. The main differences mainly stem from different approaches to the inequalities (analytic vs. arithmetic proofs) and from the choice of language (ideals vs. idèles). In this book, the first inequality (called the universal norm index inequality) is proved using analytic means, and flows very naturally from Weber's attempts of generalizing Dirichlet's theorem on primes in arithmetic progressions. Next, idèles are introduced, and the second inequality (called the global cyclic norm index inequality) is proved using a minimum of group cohomology (Herbrand quotients). The next chapters discuss Artin's reciprocity law, the existence theorem, and Hilbert class fields. The last chapter is dedicated to local class field theory; it constructs local class fields using formal groups, and proves the local version of Artin's reciprocity law without using the global theory. The book starts off at approximately the place where \textit{D. A. Marcus}' excellent [Number Fields. Universitext. New York etc.: Springer-Verlag (1977; Zbl 0383.12001)] ends; in particular, notions such as the decomposition and inertia subgroups, or results like Dirichlet's unit theorem, the finiteness of the class number, and even the calculation of the residue of Dedekind's zeta function \(\zeta_K(s)\) are taken for granted. Actually Thm.~4.2. states that \(\zeta_K(s)\) can be extended analytically to all of \(\mathbb C \setminus \{1\}\), which is true, but quite deep (and not, as claimed on p. 37, due to Dedekind); perhaps it would have been fair to remark that only the continuation to, say, \(\text{Re}\,s > 1 - 1:(K:\mathbb Q)\) is needed subsequently. The philosophy behind the author's approach to class field theory is close to the one chosen by \textit{S. Lang} in his book [ Algebraic Number Theory. 2nd ed. Graduate Texts in Mathematics. 110. New York: Springer-Verlag (1994; Zbl 0811.11001)], which is perhaps best described in Lang's own words: ``I have used the oldest proof for the universal norm inequality (Weber's proof) both because it is more natural than the ``purely algebraic'' proof of Chevalley [\(\ldots\)], and also because it motivates the whole approach to class field theory through the density of primes in arithmetic progressions, and the splitting laws.'' Like the author, Lang has ``intermingled the ideal and idelic approaches without prejudice for either''. There are of course a lot of topics that cannot be covered in such a short book; examples are the principal ideal theorem, infinite class field towers, higher reciprocity laws, or Artin's L-functions. Given the analytic approach of this text it is, however, unfortunate that the calculation of the residue of Dedekind's zeta function is omitted, that the Frobenius density theorem does not appear, and that Chebotarev's density theorem is only given as an exercise. These topics can be found in Lang's text, however, to which the present one may serve as an introduction. At various places, the author states that Artin reduced the principal ideal theorem to a purely group theoretic statement only after Furtwängler had given his proof; this is not correct (for the development of class field theory in the 1920s, and the first proofs of the principal ideal theorem in particular, see [\textit{G. Frei} and \textit{P. Roquette}, Emil Artin und Helmut Hasse. Their correspondence 1923--1934. Göttingen: Universitätsverlag Göttingen (2008; Zbl 1161.01002)]). The clarity of the exposition and the many exercises ranging from routine to quite challenging problems make the book perfect for a first introduction to class field theory.