Gustav Herdan

Born on January 21, 1897 in Brno (Moravia; mother Anna, father Adolf, merchant); died on November 16, 1968 in Bournemouth. Jurist, statistician and linguist.

Attended the first German secondary school in Brno, school-living certificate 1915, high school diploma at the Staatsgymnasium Brno. Studies in jurisprudence in Wien und Prague at the German University, in between 2 years military service; PhD at the German University, Prague (at that time, law students did not write PhD theses. Subsequently, he worked at the county court Brno. From 1933 he studied in the first place Chinese in Berlin, London (diploma in classical Chinese), Prague, and Wien. He finished his studies with an PhD in sinology (east Asian languages) and English philology. In 1938, he emigrated to England; Studies in mathematics and statistics. From 1939 to 1045, he placed his knowledge in statistics in service for wartime economy. Later, he worked as a statistician in the industrial economy. From 1948, “Lecturer in Statistics” at the Faculty of Medicine of the University Bristol. Member of the American Statistical Society, Fellow of the Royal Statistical Society, member of the Linguistic Society of America.

Herdan’s importance for linguistics consists of the fact that he was apparently the first to present an overall view of quantitative linguistics. An essential aspect of his work is the development and the testing of mathematically formulated language laws (“statistical laws”). His conception of such laws is reflected in the citation “The masses of linguistic forms...are a part of the physical universe, and as such are subject to the laws which govern mass assemblies of any kind... This is how the need for statistical linguistics arises“ (Herdan 1960a: 3).

Resuming Saussure’s dichotomy langue-parole and using methods from information theory and cybernetics, he became, together with Pierre Guiraud and Charles Muller, mover of the upsurge of quantitative linguistics in the decade from 1950 to 1960 (Aichele 2005: 18). He dealt with a great variety of themes: authorship determination, stylometrics, language change and blending, applications of information theory, type-token relation, word length and frequency distributions, interrelations between text length and vocabulary size, and between stylistics and language typology. Another topic is the German language of the Nazis (Herdan 1960a: 263ff). He presents in his works numerous language laws, among others the Zipf and Zipf-Mandelbrot, the Poisson, and the lognormal distributions. Even if not every detail is viewed in the same way today, Herdan remains one of the pioneers of quantitative linguistics. In connection with many of his studies, he collected together with his students a plentitude of data, which can be used also from the point of view of modern theoretical considerations (cf. z.B. Best & Zhu 2001: 103ff.).

Herdan studied rather philology than linguistics and adhered — as did many linguists at that time — to the teachings of de Saussure und those of the Prague structuralism. This background opened up to him several perspectives but at the same time it barred him from new horizons. 40 years after his death and taking into account the development of modern quantitative linguistics; it is easy to detect his mistakes. Detractors reviewing his work from the linguistic point of view criticized in the first place his “non-linguistic view” on language phenomena and how he interpreted them, more often than his methods. Nevertheless, he produced quite a number of approaches whose continuation could open up new vistas in linguistics.

Herdan committed himself to struggle against the “qualitative” linguistics and attacked in particular the proponents of the school of generative grammar, which began to develop at that time, on any occasion. Diplomacy was not his strong point. At his time, he could not resolve the argument in his sense; today, the situation is considerably different. Unfortunately, he had a negative attitude also towards proponents of quantitative linguistics. He brusquely refused Zipf and his “principle of least effort” as well as Zipf’s law (which has found its place in at least 20 scientific disciplines meanwhile). Today, Zipf’s discoveries build a fundament of synergetic linguistics, and his principle, taken as an axiom, was split up in numerous special cases.

A wealth of ideas can be found in Herdan’s writings. The most valuable heritage is formed by the rich source of problems and questions he brought into discussion — rather than specific methods or approaches. Some of them may be more or less inadequately solved or even pseudo problems but they can serve to recognize perspectives. Herdan is still being cited frequently (cf. vgl. e.g. Köhler, Altmann & Piotrowski 2005; Nikitopoulos 1980). Perhaps he relied in his linguistic efforts too much on his own linguistic knowledge, and avoided any cooperation with linguists – as opposed to his activities in medicine, where he concentrated on statistics and cooperated with other scientists.

Source
English version of: Karl-Heinz Best, Gabriel Altmann: Glottometrics 15, 2007, 92-96

Other languages

 * German: Gustav Herdan