1953
Does information carry meaning?
The cuts
Yehoshua Bar-Hillel 1953
Semantic information can be measured too — by the possibilities a statement excludes.
The first serious attempt to put meaning back into Shannon's framework. Still incomplete.
An Outline of a Theory of Semantic Information1952
Rudolf Carnap 1952
Give meaning a logical form — state descriptions, inductive probability — and content becomes as measurable as Shannon's bits.
Supplied the formal semantics the information measure stood on — meaning made rigorous, if not yet whole.
An Outline of a Theory of Semantic Information1952
There is live information and dead information. DNA in a living cell is a code with causal power — Bateson's 'a difference that makes a difference'; the same molecule in a dead cell is only Shannon bits. Information means something only while something is computing with it.
Tied meaning to process — the difference between Shannon information and the richer, algorithmic kind is whether anything is currently acting on it.
We Merged With Machines a Long Time Ago (with Carlo Rovelli)2026The tensions
Shannon set meaning aside on purpose; Bar-Hillel and Carnap tried to put it back within five years, and the attempt remains unfinished. Every LLM is a standing wager on this question: that meaning — or something indistinguishable from it — can be recovered from statistics alone.
