Biography · Claude Shannon

With a master's thesis he reshaped the electrical circuit; with a single paper he created an age — the Information Age.

A Master's Thesis

Cambridge, Massachusetts, 1937. The Massachusetts Institute of Technology. A twenty-one-year-old is writing his master's thesis. His name is Claude Elwood Shannon, from the small town of Gaylord, Michigan, son of a judge and of a high school principal. At the University of Michigan he had taken two undergraduate degrees — electrical engineering and mathematics. The combination, he later said, decided the course of his life.

That year, Shannon worked as an operator of MIT's Differential Analyzer, a behemoth of relays and mechanical gears built under the direction of Vannevar Bush — one of the most advanced analog computers in the world. Staring into its dense lattice of switching circuits, he saw something: those "open" and "closed" states were the same thing as the two-valued algebra invented in the nineteenth century by the English mathematician George Boole.

He wrote it down as "A Symbolic Analysis of Relay and Switching Circuits" (submitted in 1937, published in 1938). The core insight was almost embarrassingly simple: any circuit of relays and switches could be described in Boolean algebra; conversely, any Boolean expression could be realised as a circuit. AND was series, OR was parallel, NOT was a relay run backward. The analysis and simplification of complex circuits no longer depended on the engineer's intuition — they were now algebra problems.

This was no formal trifle. Before Shannon, an engineer faced with a network of relays could only proceed by trial and error: could the lights be made to flash in the desired order? How many parts would it take? Shannon's algebra made it all derivable. Designs could be proved. Errors could be located. Scale could be extended.

The Harvard scholar Howard Gardner later called this "the most important — and possibly the most famous — master's thesis of the twentieth century." It turned digital circuit design from an art into a science. Every chip in production today, with its billions of AND, OR, and NOT gates, still obeys the rules a twenty-one-year-old wrote down.

Bell Labs and the Codes of War

In 1940 Shannon took his PhD in mathematics at MIT, his dissertation on an algebraic treatment of Mendelian genetics — another crossing. The next year he joined Bell Labs, where he would work for fifteen years at one of the great industrial laboratories of the twentieth century.

During the war Bell Labs was a central node of American cryptography. Shannon's official charge was fire control and encrypted communications. He worked on the security analysis of SIGSALY — the Allies' highest-grade voice encryption system, the same one Churchill and Roosevelt used to speak to one another. In early 1943 Alan Turing, on a brief cryptographic mission to the United States, sat several times across from Shannon in the Bell Labs cafeteria. The two were forbidden to discuss their respective cryptanalytic work, so instead they spoke of machines and minds, of the nature of computation. Many years later, recalling these conversations, Shannon said what he remembered most clearly was Turing's repeated, insistent questioning of whether machines might one day think.

The war left Shannon something more than encounters with the great minds of his time: a classified internal memorandum titled "A Mathematical Theory of Cryptography." Completed in 1945 and declassified in 1949 as "Communication Theory of Secrecy Systems," it laid down for the first time a mathematical foundation for modern cryptography — defining "perfect secrecy" as statistical independence between ciphertext and plaintext, and proving that only the one-time pad attains it. From that moment, cryptography ceased to be a craft and became a provable, quantifiable science.

One Paper Made an Age

After the war Shannon returned to a quieter research life. He was thinking about a single question. What is information? What does communication actually transmit?

In July and October 1948 the Bell System Technical Journal serialised a paper titled "A Mathematical Theory of Communication." The single author was Shannon. The paper did several things that had never been done.

It defined "information." Before Shannon, information was a vague everyday word. Shannon said: information is the reduction of uncertainty. The more unexpected an event, the more information it carries. Borrowing from thermodynamics, he named the quantity entropy and gave it a precise expression: $H = -\sum p_i \log p_i$.

It defined the bit. The word, proposed by his colleague John W. Tukey and given canonical form by Shannon in the paper, became the basic unit of the digital world. One binary choice, one bit. The data humanity now generates each year is measured in zettabytes — and zetta- is just a string of thousand-fold multiples piled on top of bit.

It gave the channel capacity theorem. Every noisy channel has an upper bound C; so long as the transmission rate stays below C, information can in principle be sent with arbitrarily small error. The result was almost counter-intuitive — engineers had always assumed errors must accumulate with volume. Shannon proved they need not, provided one encodes cleverly enough. 5G, Wi-Fi, deep-space telemetry, SSD error-correcting codes — all stand on this theorem.

A year later, the paper, with an introduction by Warren Weaver, was bound as a book: The Mathematical Theory of Communication (1949). Shannon was the father of information theory. He was thirty-two.

A Chess-Playing Machine and a Maze-Walking Mouse

Shannon was not an armchair theorist. His house was a small mechanical menagerie — unicycles, juggling balls, a flame-throwing trumpet, a Roman-numeral calculator named THROBAC that read its digits backward. His MIT colleagues remember seeing him after work, riding a unicycle down the hallway while juggling three balls.

In 1950 he published "Programming a Computer for Playing Chess" in the Philosophical Magazine — the founding paper of computer chess. He distinguished two search strategies — Type A, which exhausts every move, and Type B, which examines only a few promising lines — and gave a basic evaluation function. Every computer chess program of the next thirty years, up to and including IBM's Deep Blue, was a refinement of these two ideas.

In the same year, Shannon built a mechanical mouse he called Theseus. The magnetic copper mouse traversed a metal maze, and on its second pass remembered the route, going straight to the cheese. The "memory" was held in the bank of electromagnetic relays beneath the maze — Shannon's familiar switching circuits again.

Theseus is among the earliest documented physical demonstrations of machine learning. In the surviving footage Shannon wears the grin of an irrepressible boy. He believed an important idea ought to be performable by a mouse.

A Quiet Old Age

Shannon returned to MIT as a professor in 1956 and retired in 1978. He took part in the early stirrings of artificial intelligence — among the four chief organisers of the 1956 Dartmouth Workshop he was the senior figure, perhaps a decade older than John McCarthy or Marvin Minsky.

But he kept his distance from the field's louder pronouncements. McCarthy once tried to enlist him fully into AI; Shannon refused, on the grounds that the field "had no mathematics yet."

What interested him more was using mathematics to explain whatever could be explained — including how to bet in the stock market (he studied the Kelly criterion and is reputed to have done very well as an investor), and how to balance objects while juggling (in the 1980s he wrote "Scientific Aspects of Juggling").

In his last years Shannon developed Alzheimer's disease. The information age he had personally founded was rushing in around him — the internet, mobile communications, deep learning — and his own memory was leaving him bit by bit. He died on 24 February 2001 in a Massachusetts care home, eighty-four years old. His wife Betty Shannon said afterwards: he knew he had created something important, but never thought he should be worshipped for it.

Selected Works

Year	Work	Significance
1937	"A Symbolic Analysis of Relay and Switching Circuits" (master's thesis)	Founded digital circuit design on Boolean algebra
1948	"A Mathematical Theory of Communication," Bell System Tech. J.	Founded information theory; defined bit, entropy, channel capacity
1949	"Communication Theory of Secrecy Systems," Bell System Tech. J.	Mathematical foundation of modern cryptography
1950	"Programming a Computer for Playing Chess," Phil. Mag.	Founding paper of computer game playing
1950s	Theseus (mechanical mouse)	One of the earliest physical demonstrations of machine learning

Historian's Note

Historian's Note

At twenty-one, with a single master's thesis, Shannon made the electrical circuit into algebra. At thirty-two, with a single paper on communication, he made information into mathematics. Today every text, every image, every sound, every code that travels through the cloud is measured in bits — a word he and Tukey settled on, almost in passing, one Bell Labs afternoon. The ancients spoke of the immortality of words; what Shannon set down was the medium against which all words must now be measured. Compared with Turing's ordeal and Wiener's loneliness, Shannon was fortunate — he lived long, lived freely, lived to see the world he had quietly engineered come crashing in. And yet there was sorrow too: a man who spent his life in the company of information had his own information erased, bit by bit, by oblivion. As if the universe itself wished to know how many bits it takes for a name to fade entirely from a single mind.

Eyewitness Accounts

Call for contributions

If you knew Claude Shannon personally or have firsthand sources or recollections, please contribute on GitHub.

References

1. Shannon, C. E. (1938). "A Symbolic Analysis of Relay and Switching Circuits." Transactions of the AIEE, 57(12), 713–723.
2. Shannon, C. E. (1948). "A Mathematical Theory of Communication." Bell System Technical Journal, 27(3), 379–423; 27(4), 623–656.
3. Shannon, C. E. (1949). "Communication Theory of Secrecy Systems." Bell System Technical Journal, 28(4), 656–715.
4. Shannon, C. E. (1950). "Programming a Computer for Playing Chess." Philosophical Magazine, 41(314), 256–275.
5. Shannon, C. E., & Weaver, W. (1949). The Mathematical Theory of Communication. Urbana: University of Illinois Press.
6. Soni, Jimmy & Goodman, Rob (2017). A Mind at Play: How Claude Shannon Invented the Information Age. New York: Simon & Schuster.
7. Gleick, James (2011). The Information: A History, a Theory, a Flood. New York: Pantheon.
8. Hodges, Andrew (1983). Alan Turing: The Enigma. London: Burnett Books. (Includes accounts of Turing's 1943 visit to the U.S. and his meetings with Shannon.)