The encyclopaedia on the shelf and the edition that stayed wrong

Both approaches to the household encyclopaedia work. What differs is what they cost you, and the cost is what this sets out.

The set was a snapshot with a date on it

A printed encyclopaedia was fixed at the moment it went to press, so a household's general reference stayed at the state of knowledge in the year it was bought. Publishers sold annual yearbooks specifically to patch it, which is an unusually honest admission that the main product decays. Children looked up countries that had changed names and scientific consensus that had moved, and were often not aware of it.

The version you had was visible on the spine, which is more than can be said for most reference now.

Bounded subjects had visible edges

An article occupied a defined number of columns, so you could see how much there was to know at that level of detail. Cross-references pointed to a finite set of related entries, which made a subject explorable and, importantly, finishable. Search results have no boundary, so a reader cannot tell whether they have covered a topic or scratched it.

The sense of completing something is a real cognitive service that unlimited sources cannot provide.

The selling was aggressive and the buying was aspirational

Encyclopaedias were frequently sold door to door on instalment terms to households investing in their children's education. The sets were expensive relative to income, and their presence in a front room was a deliberate statement.

Look closer and that commercial model collapsed when the same content became free, which happened faster than the publishers expected. The object survives in second-hand shops in enormous numbers and is worth almost nothing.

Continuous editing changed what an error is

An online reference article can be corrected within minutes, which is straightforwardly better than waiting for the next edition. It also means the text a reader saw last year may no longer exist, and citing it accurately requires citing a version and a date. Edit histories preserve this on some platforms and not on others, which is a large difference in accountability.

Printed error was permanent and inspectable; digital correction is instant and, without history, invisible.

Authority was concentrated and named

Signed articles by identified specialists were a selling point, and the editorial hierarchy was explicit. Collaborative reference distributes that authority and is generally found, in comparison studies, to be broadly comparable in accuracy on many topics while varying more between articles.

The variation between entries is the important part, and it is easy to forget when every page looks alike. A uniform layout implies a uniform reliability that no reference work has ever had.

What replaced browsing

People read encyclopaedias sideways, following cross-references and stopping at whatever caught their eye on the way to something else. Hyperlinks reproduce that perfectly, and arguably improved it, which is a rare case of the replacement being better at the incidental use.

Look closer and what went instead was the physical accident of the neighbouring entry, since alphabetical adjacency puts unrelated things together. Randomness by alphabet is a weaker recommender than a link and produces stranger discoveries.

Side by side

The takeaway

One knowledge went out of date on a shelf where you could see it. The other goes out of date silently and then denies it ever said otherwise.

What replaced it works. That is not the same as it being nothing.