The Half-Life of Facts
Arbesman explores patterns in how knowledge grows, becomes outdated, and changes across scientific and professional fields.
Samuel Arbesman asks readers to consider knowledge as something that changes over time, rather than a collection of statements acquired once and retained indefinitely.
An old map may be inaccurate because roads were built after it was printed, not because the original survey was careless. A revised medical claim may instead reflect stronger evidence; both require updating, but for different reasons.
His title borrows a scientific metaphor for the way accepted information can become obsolete, corrected, or less useful as inquiry and circumstances develop.
Different domains change at different speeds. A historical event, a medical recommendation, and a figure describing a city's population do not become outdated for identical reasons.
Some information changes because the world changes; other information changes because measurement improves or an earlier explanation proves inadequate. Keeping these processes distinct prevents unnecessary confusion.
Arbesman looks for regularities in the growth and turnover of knowledge, including patterns in publications and scientific activity that may be easier to see collectively than individually.
Knowing that a field changes rapidly can justify checking current guidance before an important decision. It does not justify treating a well-tested claim and a fashionable unsupported assertion as equally uncertain.
Such patterns do not supply an expiration date for every fact. A statistical description of a field cannot tell you exactly which particular belief will be revised next.
The book also considers how knowledge travels unevenly. A discovery can exist in a specialist setting long before it changes textbooks, professional routines, or public assumptions.
This makes updating an institutional problem as well as a personal habit. Continuing education and reliable channels for correction matter when outdated information can cause practical harm.
Recognizing change should not produce the conclusion that nothing is trustworthy. Some findings are highly stable, and revision often reflects improved methods rather than the failure of inquiry itself.
Arbesman's account encourages a durable relationship with learning: hold claims with appropriate confidence, notice their context and age, and build routines that let credible corrections reach you.
Some findings are highly stable, and revision often reflects improved methods rather than the failure of inquiry itself.
Which knowledge in your work or daily life needs regular updating, and how would you distinguish a credible revision from a merely novel claim?