����logo

In 2025, the ����logo Information Theory Society celebrated 75 years of Information Theory Symposia: London 1950 - Ann Arbor 2025!��At the Ann Arbor ISIT, conference attendees could have their pictures taken (alone or with other people). These pictures, together with collected pictures from the past, where then added to the following mosaic!

Some historic context about the Information Theory Symposia can be found at the bottom of this page.

��

Information Theory Symposia in the 1950’s

London Symposia

The 1950 London Symposium on Information Theory was held in the Lecture Theatre of the Royal Society, Burlington House, September 26-29, 1950. This was the first symposium or conference known to have the name Information Theory in its title.

According to the foreword of the proceedings of the Symposium, written by Dr. Willis Jackson from Imperial College London, the Symposium aimed to provide an opportunity for discussion of the nature and potentialities of recent works of a few “theoretically minded communication engineers” on a commodity called “information” among interested scientists and engineers.

The proceedings of the Symposium included papers associated with each presentation and recorded follow-up discussions from attendees. Claude E. Shannon gave talks on three subjects: “Communication Theory - Exposition of Fundamentals”, “General Treatment of Problem of Coding”, and “The Lattice Theory of Information”. These are captured in three technical papers included in the , which featured approximately twenty papers in total. There is also a discussion paper on his talks, to which he contributed.

Apart from one delegate from the United States, most attendees were from British academies, government or industry. Notably, Alan Turning was among the attendees and contributed to a discussion paper on machines and learning. There were also 23 participants from Western European countries.

The Second London Symposium on Information Theory was held in September 1952 at the Institute of Electrical Engineers. Of the 38 papers presented, eight dealt with transmission and analysis of speech.

The Third London Symposium on Information Theory was held between 12 and 16 September 1955 in the auditorium of the Royal Institution at 21 Albemarle Street, between Berkeley Square and Piccadilly Circus, London. There were about 250 participants, of which half were British, and the rest hailed from USA, Western European countries, USSR and Israel. However, Shannon did not attend the 1955 London Symposium.

The first day of the Symposium was devoted to information theory and its measures. The second day was on coding and featured, among many contributions, a presentation by Peter Elias on codes for the Binary Symmetric Channels and Binary Erasures Channels, and a talk by David Huffman on “The Synthesis of Linear Sequential Coding Networks”. The third to fifth day of the Symposium were devoted to language analysis and mechanical translation; meaning and the human senses; and behaviour and its mechanism, respectively.

See an and reflections about the Third London Symposium by Nelson M. Blachman, which was published by the Institute of Radio Engineers (IRE) Transactions on Information Theory in 1956.

��

IRE-led Information Theory Symposia in the USA in the 1950’s

In his 1973 paper entitled “”, David Slepian recalls: on May 8, 1951, a document signed by 50 IRE members was directed to the IRE Board of Directors which petitioned for the formation of an “IRE Professional Group on Information Theory.”

The new group sponsored the Symposium on Statistical Methods in Communication Engineering in Berkeley, California in 1953. The 1954 Symposium on Information Theory was organized by the group at MIT. There were two more IRE Information Theory Symposia in the USA in the 1950’s. One was held in 1956, again at MIT, and the next one was held in Los Angeles in 1959.

According to David Slepian US Symposia in the fifties covered subject areas like those treated in the IT Transactions. Whereas the London meetings were of a much wider ranging scope: “Information theory was construed in the widest possible sense.”

��

Publication of the Proceedings

The first publication of the IRE Professional Group on Information Theory was () in February 1953, which was a reprint of the papers and discussions of the first London Symposium on Information Theory, with the cooperation of the British Ministry of Supply. The papers of this proceedings can now be found on ����logoXplore as the first issue of the Transactions of the IRE Professional Group on Information Theory.

The Butterworths Scientific Press published the 1952 and 1955 London Symposium proceedings. They appeared in their formal form rather than as presented and the remarks made during the discussion periods were also included.

��