Being a lexicographer compiling dictionaries in the internet age can be viewed as both exciting, because of all the new words that can quickly gain currency, or a nightmare, because one has to decide whether to include some new word or not and what the word might even mean, knowing that whatever you decide will be hotly contested by some.

In a review of the book Unabridged: The Thrill of (and Threat to) the Modern Dictionaryby Stefan Fatsis, Louis Menand looks at the history of the modern dictionary.

Samuel Johnson’s Dictionary of the English Language, published in London in 1755, carved out a role for the dictionary: to establish what would become known as Standard English. Johnson himself was aware that language is a living thing, always in flux. But his dictionary, with its conclusiveness, was a huge publishing success. It was considered authoritative well into the nineteenth century. In England, it would be replaced by the Oxford English Dictionary. But, in the United States, its role was usurped by Noah Webster’s American Dictionary of the English Language, which made its début in 1828.

Webster deliberately set out to supersede Johnson. His ambition was to create not a dialect of British English but an identifiably American language. Johnson’s dictionary had about forty-two thousand words; Webster’s had seventy thousand. Webster added New World words including “skunk,” “boost,” and “roundabout”; words with Native American origins, such as “canoe” and “moose”; words derived from Mexican Spanish, like “coyote.” Most dramatically, he Americanized spelling, a project started in an earlier work of his, a schoolbook speller called “A Grammatical Institute of the English Language,” published in 1783. It is because of Webster that we write “defense” and “center” rather than “defence” and “centre,” “public” and not “publick.” He changed the language.

Webster’s New International Dictionary, Second Edition, announced as “unabridged,” appeared in 1934. Web. II was a doorstop—six hundred thousand entries, thirty-five thousand geographical names, and, in the appendix, thirteen thousand biographical names. 

That dictionary was prescriptivist, laying out what words are acceptable as ‘proper’ and how they should be used. But that attitude changed dramatically with the Third Edition.

Webster’s Third New International Dictionary, published in 1961, flipped the script. Fatsis says that it “changed lexicography.” Web. III had an open-door policy. It was descriptivist. The editors did not abandon the concept of Standard English, which they defined as English “well established by usage in the formal and informal speech and writing of the educated,” and they indicated when a word was considered nonstandard. But they eliminated the label “colloquial” and reduced the number of words labelled as slang. The spirit was nonjudgmental.

This switch in policy was brutally attacked by prescriptivists, usually professional writers who were particularly incensed by the inclusion of the word ‘ain’t’, the use of which had been seen (and perhaps still is) as a marker or poor language skills. It was essentially an indicatorof social class.

The rapidity with which new words are coined is a huge challenge for lexicographers.

The last print edition of Web. III (which was the basis for Merriam-Webster’s subscription website) is two thousand seven hundred and eighty-three pages and has four hundred and sixty-five thousand entries. You need a book stand to use it. But it probably contains less than half of the words in the English lexicon. According to one study of digitized library books, there were about six hundred thousand words in the language in 1950, and more than a million by 2000. The same study concluded that fifty-two per cent of English words found in printed books are “lexical dark matter,” not represented in any standard reference work. 
…

The second and, as it turned out, final print edition of the O.E.D., published in 1989, comes in twenty volumes, weighs a hundred and thirty-eight pounds, and has a little under three hundred thousand entries. The online O.E.D. weighs nothing (so there’s less risk of user injury) and has eight hundred and fifty thousand entries. The dictionary is updating or adding new words at the rate of fifteen thousand a year. With the internet, the O.E.D. can expand forever, but it will never come close to recording every meaning of every word used by English speakers—of which there are, according to the International Center for Language Studies, 1.52 billion. Even the most unabridged of unabridged dictionaries is a highly curated sample of the language.

And that does not even take into account generational slang which is pretty much impossible to keep up with, given the rapidity of its rise and fall. Some slang terms that were coined in my generation, such as ‘cool’, seem to have become permanent. But the status of others is doubtful. I have no idea if young people use the term ‘hip’ anymore. One sign that a current slang term is going out of favor is when oldsters like me know what it means. And if we actually start using it, that is the kiss of death.

Menand argues that focusing on words may be problematic

Dictionaries rely on the belief that the word is the basic unit of linguistic meaning. It is not. The basic unit of meaning is the sentence, or, sometimes, especially in speech, the phrase. You can memorize vocabulary, but if you can’t make a phrase you can’t speak the language. This is not simply a matter of grammar, of syntax and morphology. It’s ultimately a matter of cultural literacy. The dictionary is like the periodic table: it can tell you what the elements are, but not how to combine them. Words take a lot of their semantic coloration from the words around them.

While that may well be true, describing words and their meanings is hard enough. Dictionaries do give examples of words used in different phrases and sentences to show their contextuality.

I consult dictionaries all the time, sometimes for meaning, sometimes for etymology, and sometimes for pronunciation. But I do it all online. I cannot remember the last time I consulted a book version. I cannot remember when I last even saw a paper one. So as far as I am concerned, the print versions of dictionaries are obsolete but digital versions still remain very much a necessity.