Paul Otlet and Emanuel Goldberg
Abstract
Long before computers, two men tried to make the world’s information searchable. In Brussels the Belgian lawyer Paul Otlet (1868–1944) built, from 1895, a universal catalogue of everything published, on index cards filed by a decimal classification, and imagined people consulting it from their homes by telephone and television. In Dresden the chemist and camera executive Emanuel Goldberg (1881–1970) patented in 1927 and demonstrated in 1931 a “statistical machine” that searched microfilm with a photocell. Both were forgotten after the Second World War and rediscovered by library scientists as precursors of search engines and of Vannevar Bush’s Memex.
The Universal Catalogue
Library catalogues on index cards, 12.5 by 7.5 centimetres, spread through libraries in the late nineteenth century, and so did systems for ordering knowledge. In 1876 the American librarian Melvil Dewey published his decimal classification, in which numbers denote classes and subclasses: if 5 stands for the natural sciences and 52 for astronomy, 523 is descriptive astronomy.
Paul Otlet, born in 1868 to a father who had grown rich on tramway companies, studied law in Leuven and Brussels and from 1890 worked for the lawyer and writer Edmond Picard, without much enthusiasm. From 1891 he worked with an older colleague, Henri La Fontaine, on bibliographies of the social sciences; by 1894 they had produced about 100,000 index cards. When Otlet encountered the Dewey system in 1895 they set themselves a larger goal. That year they founded the International Institute of Bibliography in Brussels, with state funding, to compile the Répertoire bibliographique universel (RBU), a catalogue of the whole of the world’s published knowledge, filed by their extension of Dewey’s scheme, the Universal Decimal Classification. In practice it consisted of thousands of drawers of index cards, with further card files for images and documents.
The years before the First World War were the project’s best. La Fontaine received the Nobel Peace Prize in 1913, mainly for his work for international pacifism, and the prize money helped. The collections moved into a building in the Parc du Cinquantenaire, which from the 1920s was called the Mundaneum (the HNF gives 1924; other accounts date the move and the name to 1920). Otlet dreamed of a whole city of knowledge, and Le Corbusier drew plans for it. By 1934 the RBU held more than 15 million cards (some sources give 12 million). That year the Belgian government withdrew its funding and the Mundaneum closed. Otlet published his main work, Traité de documentation, the same year, and in his last years sketched information networks using telephone and television and a workstation for the information seeker, the Mondothèque. He had already proposed microfilm catalogues in 1906. He died in Brussels in 1944.
The Australian library scientist W. Boyd Rayward rediscovered him, publishing a biography in 1975 and an English selection of his essays in 1990. Since 1998 a new Mundaneum in Mons has kept Otlet’s papers and some of the original card cabinets.
The Statistical Machine
Emanuel Goldberg was born in Moscow on 31 August 1881, the son of a senior military doctor. He studied chemistry in Russia and at German universities, took his doctorate in Leipzig in 1906 and stayed in Germany; from 1907 he was a professor at the Royal Academy of Graphic Arts in Leipzig, working on photochemistry, printing and light measurement. In 1917 he joined the board of the Dresden camera maker Ica, which in 1921 brought out his handheld camera for 35 mm film; in 1925 he invented a process for microdots, extremely reduced photographs later favoured by spy services. When Ica merged with three other optical firms under Zeiss’s control in 1926, he became general director of the new Zeiss Ikon, by 1931 the largest camera maker in the world.
On 13 April 1927 Zeiss Ikon and Goldberg applied for a patent on a “device for selecting statistical and accounting data.” Each frame of a microfilm held a record along with identifying codes. A small glass plate carried the codes being searched for; when a frame with matching codes passed, a photocell was darkened and triggered a camera that photographed the record. A second application, of 2 August 1931, replaced the glass plate with a pattern of dots and small lamps. Three days later, on 5 August 1931, Goldberg demonstrated a model of the machine at the Eighth International Congress of Photography at the Technische Hochschule Dresden, in a lecture on “new ways of photographic recording.”
In 1933 the Nazis forced Goldberg out of his post. He worked for Zeiss Ikon in Paris for some years, emigrated to British Mandate Palestine in October 1937 and from 1 March 1938 ran his own firm in Tel Aviv, Professor Goldberg’s Laboratory for Precision Instruments. He received the Israel Prize in 1968 and died in Tel Aviv on 13 September 1970. The two prototypes of his statistical machine were probably destroyed in the bombing of Dresden in 1945.
The Memex Connection
In the 1930s Vannevar Bush developed a very similar device, the Rapid Selector, which searched microfilm with photocells; a working version built for libraries after the war did not prove itself in practice. Bush carried elements of it into the Memex, the desk-sized microfilm system with associative trails he described in 1945 (see Vannevar Bush and the Memex). IBM’s Hans Peter Luhn took the problem of automatic document retrieval into the computer age in the 1950s.
The library scientist Michael Buckland, born in England and at Berkeley since 1976, has studied Goldberg since 1988; his biography appeared in German in 2010. In the winter semester of 2013/14 students at the Hamburg University of Applied Sciences under Dirk Lewandowski built a working model of the statistical machine, and in 2017 the Technische Sammlungen Dresden, housed in Zeiss Ikon’s former headquarters and holders of Goldberg’s papers, devoted a large exhibition to him. Both men are now commonly listed among the precursors of the search engine (see The Search Engine Wars).