Between 'Dorcopsis', a genus of small kangaroo, and 'doré' in Webster's New International Dictionary, second edition — the great 1934 'Webster's Second', then the most authoritative dictionary in America — sat a short, confident entry: dord, a noun of physics and chemistry, meaning density. It had a pronunciation, a part of speech and a definition. What it did not have was any existence outside the dictionary. Nobody had ever written or said 'dord'. The entry had been born in the office files: a slip reading 'D or d, cont./density' — noting that the abbreviation for density could be a capital or a lower-case D — was misread as a single word, 'Dord', and processed as one, acquiring its phantom pronunciation along the way.
The ghost was only spotted on 28 February 1939, when an editor noticed that 'dord' lacked an etymology and went digging in the files. Philip Babcock Gove, who joined the dictionary's staff some years later and would go on to edit the even more famous, and famously controversial, Webster's Third, told the whole story in a 1954 journal article. An order went out to strike the word, but printing plates have long memories, and dord lingered in printings for years before it was finally exorcised. Merriam-Webster's editors have admitted a certain fondness for it since: as errors go, it was at least a pronounceable one.
Lexicographers have a name for such spectres: ghost words. The term was coined by the Cambridge philologist Walter William Skeat in his presidential address to the Philological Society in 1886, for words that have no real existence — creatures of misprints, misreadings and scribal slips that then breed in the safe habitat of reference books. The classic example is 'abacot', born as a garbled form of 'bycoket', a type of medieval cap; once it slipped into a sixteenth-century chronicle it was copied from dictionary to dictionary for some three centuries, acquiring ever more confident definitions along the way. Another, 'phantomnation', arose when a nineteenth-century compiler ran together Alexander Pope's hyphenated 'phantom-nation' — the ghostly peoples of the Odyssey — into a single word glossed as 'illusion'. Ghost words spread because dictionaries trust one another; an error told once in a reference book compounds like interest.
Traps set on purpose
Not every fake entry is an accident. Reference publishers, whose products are expensive to compile and cheap to copy, have long planted them deliberately as copyright traps. The patron saint is Lillian Virginia Mountweazel, who according to the 1975 New Columbia Encyclopedia was born in Bangs, Ohio, abandoned fountain design for photography, and died at thirty-one in an explosion while on assignment for Combustibles magazine. She never existed — but her surname has become the accepted generic term for a planted fake: a mountweazel. The word gained currency in 2005, when the New Yorker investigated a rumour that the New Oxford American Dictionary contained a fictitious entry beginning with E. The hunt narrowed to 'esquivalience', defined as 'the wilful avoidance of one's official responsibilities', which had been planted in the 2001 first edition. The trap worked, too: esquivalience later surfaced on Dictionary.com, credited to a different dictionary, and was quietly removed.
Cartographers play the same game with trap streets and paper towns. In the 1930s the General Drafting Company placed the invented hamlet of Agloe — an anagram of its founders' initials — on its road map of New York State. Decades later Agloe appeared on a rival publisher's map, and the accusation of copying foundered partly because, in the meantime, someone had opened a real Agloe General Store on the spot: the fiction had quietly summoned the fact. In law, though, the traps have proved weaker than publishers hoped. The landmark 1991 US Supreme Court ruling in Feist v. Rural held that bare facts cannot be copyrighted, so catching a rival with your fake proves copying without necessarily proving infringement.
Dord survived five years; abacot survived three centuries. The moral is not that dictionaries are unreliable — it is that they are human documents, compiled from slips of paper by tired people, and that their errors, once fossilised in print, take on lives of their own. Every ghost word is a small experiment in how authority works: we believe a word is real because the dictionary says so, and the dictionary, occasionally, believes it because another dictionary said so first. Somewhere on the reference shelf, a few ghosts are surely still waiting to be noticed.
Quiz nuggets
- 'Dord', defined as a physics and chemistry term for density, entered Webster's Second in 1934 after a slip reading 'D or d' was misread as a word.
- An editor spotted dord in 1939; Philip Babcock Gove, who joined the staff later and went on to edit Webster's Third, recounted the story in 1954.
- The philologist Walter William Skeat coined the term 'ghost word' in an 1886 address to the Philological Society.
- Lillian Virginia Mountweazel, a fictitious photographer in the 1975 New Columbia Encyclopedia, gave planted fake entries their generic name.
- The invented word 'esquivalience' was planted in the New Oxford American Dictionary (2001) as a copyright trap — and later caught Dictionary.com.