> The SAOB is a historical record of the Swedish language from 1521 to modern day.<p>Does anyone know why they would start in 1521? Usually it's some historical event. Perhaps the introduction of the printing press?<p>I wonder how 'comprehensive' historical dictionaries will manage the Internet era. When the Oxford English Dictionary (OED) and Swedish Academy Dictionary (SAOB) began in the later 19th century, think of the scope of their projects from their perspective:<p>For the OED, Early Modern English - the era of Shakespeare and the King James version of the bible - began (and Middle English ended) around 1500, with the arrival of the printing press. Printed output wasn't nearly at 20th century levels (and to the extent they want Middle English, pre-printing press the published record was much less). Publishing a book required resources, limiting the number of publishers. You might imagine that you could make a record of nearly every word that ever had currency, and trace it back to its origins. You just needed the books.<p>In the 20th century, books and other printed materials (magazines, etc.) exploded and so did word invention; it seems more challenging, but you still had a well-defined domain of data limited by the cost of publishing.<p>In the 21st century, publishing is effectively free and immediate, done by probably over a billion people. How do you track all those words, determine which are worth putting in the dictionary, trace their origins, define them? Automated tools could do a lot: Keep a db of new words, track their frequency and usages. But the entire Internet is a very large search domain.<p>How feasible is it to automate tracking all novel words on the Internet? You'd need some deduplication too for different spellings. The hardest part may be that the dictionaries want not only the words, but every meaning or 'sense' of it. If a new usage evolves for an existing word, how can automated tools detect that?