Today, there are dozens of dictionaries available for English – unabridged dictionaries, college dictionaries, children’s dictionaries, specialized dictionaries of music or architecture, an official Scrabble dictionary, not to mention on-line dictionaries in many varieties. Each one of these is edited by a team of individuals who make the judgment call whether xyz deserves to be in the dictionary. The decision is made on a number of grounds:

• the size of the dictionary, which determines the number of words it can hold;

• the intended audience of the dictionary (adults, children, language learners, etc.);

• whether a word has a sufficiently broad base of usage;

• whether it’s likely to last;

• whether it’s too specialized or technical for the intended audience;

• for a word borrowed from another language, whether it’s assimilated enough to be considered part of English.

With respect to size, the number of words and the depth of entries (whether etymologies and illustrative quotes are included, for example) in print dictionaries are determined by the number of pages and the font size of the print used. On-line dictionaries do not have the sort of space constraints that print dictionaries do. As for audience, a dictionary intended for college-age adults will probably have more learned and technical words than a dictionary for children. On the other hand, words that a native speaker is unlikely to need defined might be more of a focus in a dictionary for English language learners; the meanings of prepositions and their idiomatic uses come to mind here. The type of dictionary also determines how broad a base of usage a word needs to have in order to be included. Dictionaries of slang, dialect, or of specialized fields obviously contain more narrowly used words than general dictionaries do (although you might be surprised at how much slang and technical terminology can be found in general dictionaries).

Perhaps the trickiest issue is how long a word has to have been around to merit inclusion in the dictionary. These days, words can appear in dictionaries fairly quickly, especially in on-line dictionaries. The OED already lists google as a verb, with its first illustrative quotation dated 1999. The word bouncebackability – allegedly coined by British sportscaster Iain Dowie in 2004 (Hohenhaus 2006) – already had a draft OED entry by June, 2006 (although interestingly, the OED has traced the word as far back as 1961!).

And there is more to consider in deciding whether a word goes into the dictionary. Take, for example, words formed with various prefixes and suffixes. If happy is in the dictionary (as it certainly would be), do we need to have an entry as well for happiness? Similarly, if sad has an entry, do we need sadness? If our audience is a learner of English, perhaps yes, but for native speakers who know intuitively how the suffix -ness is used, is there any need for these extra entries? Interestingly, dictionaries are often quite inconsistent on how many and which derivatives with particular suffixes get entries. The on-line OED has entries for redness, blueness, pinkness, greenness, and yellowness, but not orangeness. The word purpleness is used in the definition of the word purplely, but does not have its own entry. And not surprisingly, there is no entry for mauveness or beigeness. What is more surprising is that there are so many entries for color words with the suffix -ness attached.

Certainly, if a word derived with a prefix or suffix takes on an idiosyncratic or lexicalized meaning, the dictionary needs to include it. Take, for example, the word transmission, which can have the transparent meaning ‘the act of transmitting’ but probably more often is used to denote a part of a car. This second meaning probably deserves to be in the dictionary. But is it necessary to include all derived words whose meanings are perfectly clear from the meaning of the base plus the meaning of the affix? Probably not.

Until the last decade of the twentieth century lexicographers made their decisions by reading materials of all sorts, and in more recent decades by listening to radio, TV, and talk in general. Potential entries would be recorded with their context on small slips of paper. These slips would then be filed, and when a critical mass of usages accumulated for a word, it might be considered for entry in the dictionary. These days lexicographers are aided by corpora (singular corpus), large computerized databases that can be searched for words in the context of their use, and by the internet, which might be viewed as a vast corpus. Indeed the rise (and sometimes fall) of a new word can be traced by searching for its use on the internet.

Perhaps the most interesting recent development in lexicography is the rise of Wiktionary – an on-line collaborative dictionary created not by professional lexicographers, but by users themselves. In the instructions for submitting entries, Wiktionary asks that words be attested, by which it means they must be in widespread use, available in well-known works or refereed publications, used at least three times in at least three sources over more than a year. It does, however, have a category of what it calls ‘protologisms’ for “terms defined in the hopes that they will be used, but which are not actually in wide use.”

One final note about the vagaries of dictionaries. Lest you think that lexicographers are humorless (“harmless drudges” as Johnson calls them in his dictionary), let’s consider the issue of mountweazels mentioned briefly above. As Henry Alford reveals in the August 29, 2005 issue of The New Yorker, the editors of the New Oxford American Dictionary (2001) planted the non-existent word esquivalience (defined as “the willful avoidance of one’s official responsibilities . . .”) among the entries for the letter “e” to catch potential dictionary pirates. Such false words are called ‘mountweazels’, from the false entry for Lillian Virginia Mountweazel in the New Columbia Encyclopedia. 3 What is most interesting for our purposes is that once these fake words have been coined, they take on lives of their own. As of December 2006, there were 55,300 hits for esquivalience and 22,700 for mountweazel on Google, leading me to wonder whether these fakes have now become real words.