Juggler of Information
His master’s thesis turned Boolean logic into circuitry; his 1948 paper made information measurable. He also built a maze-solving mouse and juggling machines.
Mathematician & Information Theorist
Claude Shannon found two unusually durable abstractions: Boolean algebra for relay circuits and information as a measurable engineering quantity. His career joined exact mathematics to a workshop culture of mazes, chess machines and juggling devices.
Shannon grew up in Gaylord, Michigan, making radios and other devices, then studied both electrical engineering and mathematics at the University of Michigan. One discipline dealt in physical circuits, the other in symbolic rules; his graduate work brought them into the same room.
At MIT he worked as a research assistant on Vannevar Bush’s differential analyzer. Its wheels and shafts calculated with continuously changing quantities, while relays helped govern the machine. The problems Shannon saw in those switching arrangements became the subject of his master’s thesis.
A Symbolic Analysis of Relay and Switching Circuits gives each relay contact a symbol and translates arrangements of contacts into an algebra of open and closed states. An expression can then be simplified before a circuit is wired. The thesis also works in the other direction: start with the behaviour wanted, derive an expression, and design a switching network that realises it.
The distinction matters. Shannon did not invent Boolean logic or the relay; he showed engineers how to use one to analyse and synthesise the other. In the thesis, an abstract identity becomes a way to reduce contacts in an actual machine. Read a few of its circuit diagrams alongside the algebra to see the method at work.
A Symbolic Analysis of Relay and Switching CircuitsBoolean algebra becomes a way to reason about electrical switching.Read the thesis PDFShannon joined Bell Labs in 1941. Wartime work included fire-control and a classified study of cryptography, published in revised form in 1949. Communication now meant more than making a circuit switch correctly: a message had to pass through a channel with limited capacity and sometimes with noise.
His 1948 paper A Mathematical Theory of Communication begins by separating the engineering task of reproducing a selected message from the message’s meaning. The source chooses among possibilities; the transmitter encodes that choice; the channel carries a signal to a receiver. Entropy measures the uncertainty in the source, while channel capacity sets a limit on reliable transmission. The paper asks how coding can use the statistical structure of messages and cope with noise.
These are conditions on communicating symbols, not measures of truth, wisdom or human significance. The 1949 book places Shannon’s paper beside Warren Weaver’s introductory discussion, giving readers a less technical route in while preserving the distinction between their contributions.
Shannon kept turning formal questions into visible mechanisms. His chess work set out ways a program could examine possible moves and evaluate positions; it was a proposal for programming, not a claim that he had made a champion computer player. Theseus, the electromechanical mouse, explored a maze and used what it had learned to take a direct route on a later run.
The playful machines did experimental work too. A route learned by trial and error or a chess choice made by a rule lets an observer ask what “learning” or “strategy” means in a particular mechanism. The MIT Museum preserves Theseus and other devices, including Shannon’s Nim-playing and juggling machines.
Shannon joined MIT’s faculty in 1956 while retaining his Bell Labs connection. He died in 2001. Begin with the 1948 paper for the communication problem, then turn back to the 1937 thesis and the machines: each makes a different relationship between an idea and a working device visible.
1948 · Claude Shannon · Corrected reprint
Start with the introduction, where Shannon deliberately brackets meaning to define a tractable engineering problem. That distinction prevents “information” here from being mistaken for knowledge or significance.Read the PDF1937 · Claude Shannon · MIT master’s thesis
Boolean algebra becomes a way to reason about electrical switching.Read the thesis PDF1949 · Claude Shannon · Publisher record
Shannon models secrecy systems with the same probabilistic discipline used for noisy channels, establishing foundational results for modern cryptography.Read the work1950 · Claude Shannon · Original paper at CHM
Before chess programs were practical, Shannon separates search, evaluation and strategy into an implementable research programme.Read the work1951 · Claude Shannon · Publisher record
Experiments with human guessing estimate the redundancy of English, connecting an abstract entropy measure to language sequences.Read the work1949 · Claude Shannon · Publisher record
Shannon relates bandwidth, sampling and noise to the recovery of continuous signals, complementing the discrete framework of the 1948 paper.Read the work1956 · Claude Shannon · Original essay
Shannon warns that information theory’s success invites vague application outside its tested domain—a compact antidote to treating “information” as a universal metaphor.Read PDF1982 · Interview with Robert Price · IEEE archive
Shannon recalls Bush’s differential analyzer, Bell Labs and the development of information theory in his own restrained voice.Read the interview1956 · Claude Shannon · Automata Studies
A small-machine construction asks how little internal structure is required for universal computation.Read the work1949 · Claude Shannon · Bell System Technical Journal
A technical extension of Shannon’s switching work concerned with constructing economical relay networks from specified functions.Read the workClaude Shannon · Paper reprinted in Collected Papers
Shannon analyses what differential analyzers can compute, linking his MIT machine work to questions of general computation.Read the work
Vannevar BushThe MIT machine on which Shannon worked as a student
Douglas EngelbartFrom transmitting information to helping people work with it
_by_Claude_Shannon_-_MIT_Museum,_early_1980s_-_Cambridge,_MA_-_DSC09099.jpg){kind=link}