The death of a mathematical theory. A study in the sociology of knowledge (Q2538106)

To explain the title: (1) the mathematical theory which is commemorated is the theory of invariants; (2) there is no sociology involved in the present paper; according to the author it is reserved for a forthcoming paper. The paper contains a few details which may be new even for expert readers. In many essential points, however, it betrays a lack of understanding of the subject matter as well as a lack of historical understanding. There is no trace of any source study. The paper seems to be based on second hand material such as reviews and oral appreciations. The presentation of the material and the style are such that sometimes entire passages are hardly intelligible (e. g., p. 151, 5 f. b. to 152, 13; or p. 153, 3 f. b. - 154, 6). The author never explains whether he is concerned with the theory of invariants in general, or with the theory of algebraic invariants. As long as he follows Franz Meyer's report (up to 1890) it is clear that he means the narrow interpretation but some quotations with respect to the more recent history can only be understood in the more general sense of the theory of invariants. It would be interesting to know whether his statistics do include or not the theory of differential invariants. It is a still more serious drawback that the author never compares the fate of the theory of invariants with that of other mathematical theories. To understand what the author means by the death of the theory of invariants one should know whether this theory is as dead as, or more or less dead than, or dead according to the same principles, as are Calculus such as taught 70 years ago, or variational calculus, or analytical number theory, or acoustics, or qualitative statistics. Clearly a case as old Calculus which has become unsatisfactory, is vastly different from variational calculus which has been completely integrated into functional analysis, or from analytical number theory which at the moment has exhausted its methods. The author does not make clear the special point of the theory of algebraic invariants. Essential parts have been integrated into other disciplines, and it would not simply be possible today to formulate the questions and to view the problems of this theory in a 19th century terminology. To grasp what was the particular fate of invariants, one has to acquaint himself with the particular shape of the theory of invariants before Hilbert. If the author ever did so, he would have been struck by the marvelous ``symbolic method'' which in fact is mentioned by him in a single sentence as an accessory fact. The author's lack of knowledge about the symbolic method is the most serious shortcoming of his paper, because if in the 90th of the last century there was anything that died in the sense that it was forgotten and never integrated into another discipline, if there was anything that was regretted because of its splendor and beauty by the confused contemporaries of Hilbert's discoveries, then it was just the symbolic machinery -- the reviewer admits it would have died anyhow because of its too restricted use, but he can imagine that some day a genius will succeed in adapting it to new uses. Anyhow, there is no understanding of the history of invariants as long as the symbolic method is not properly understood. This is the main point of the reviewer's criticism but he should still mention a few others. The characterization of \textit{R. Weitzenböck}'s book [Invariantentheorie. Groningen: P. Noordhoff (1923; JFM 49.0064.01)] can only have been written by somebody, who excerpted some review on this book which in turn was based upon the first page of its preface. The book which is said to have radically changed the appearance of Invariant Theory is rather an old fashioned presentation which added to the traditional weaknesses of the symbolic method new ones. Vectors and tensors are introduced in Weitzenböck's book for the needs of ``differential'' invariant but not at all integrated into the algebraic theory as the author falsely asserts. The author has registered the influence of Invariant Theory in what he calls two groups: algebraic geometry and (of all disciplines!) continuum mechanics (the latter group ``spearheaded by half a dozen men'' who published ``at least twenty papers''). Probably the author interviewed two people on Invariant Theory, and by chance one was an algebraic geometrician who knew nothing about Invariant Theory and the other, a physicist who knew Weitzenböck's book. The report on these interviews is difficult to assess; in no sentence it is clear whether it reflects the opinion of the expert or of the interviewer. It is a pity that obviously the author did not meet any modern mathematician who could tell him about syzygies and cohomological algebra. He would have wondered how strong a dead discipline still lives after 60 years.