In 1886, a lexicographer named Walter Skeat first used the phrase “ghost words” to describe words that he said had “no real existence.” In other words, ghost words are words that weren’t real to begin with—they made it into the dictionary because of an error or misunderstanding.
Gravy
For example, it appears that “gravy” only became a word because a 14th-century translator misread a French cookbook. (1, 2) In Old French, the word was spelled with an N: “grane” (also sometimes spelled “grain”), and it was related to the word “grain,” which according to the Oxford English Dictionary meant “anything used in cooking” at the time.
But English cookbooks translated from French in the 14th century and later nearly always have a V or a U instead of the N, leading to the word “gravy” that sounds so right to us today. Researchers believe it was simply a transcription error. If the word had been transcribed properly, we’d be having “grany” on our mashed potatoes at Thanksgiving.
Syllabus
In the 15th century, a misprint gave us another ghost word: “syllabus.” The Roman philosopher Cicero died in 43 BC, but his work has been read ever since. Two of his “Letters to Atticus” (one, two) have the word “sittybas” (or possibly “sittubas”—sources disagree). Either way, it was a Greek word meaning “a label for a book or parchment” or “title-slip,” but one printing of this work mistakenly spelled the word as “syllabus.” (3, 4)
People apparently thought “syllabus” was Latin, and the spelling stuck so well that “syllabus” took on its new meaning in the mid-1600s and now even has a fake Latin plural: “syllabi” (although “syllabuses” is also listed as an option in all the dictionaries I checked.)
Tweed
Here’s a more recent misunderstanding that gave us a new word. We got the word “tweed”—a type of wool—from a misunderstanding of the Scottish word “tweel,” which was how the Scots said “twill.” That mistake may have happened because there’s a Tweed river in Scotland, so when people heard or saw “tweel,” they thought of the Tweed River; but regardless of how it happened, “tweed” became an established word for the cloth in London in the mid-1800s. (3, 5, 6)
Dord
Here’s an even more recent ghost word you may not have heard of, but it has a quirky origin: “dord.” The story goes that the original dictionary entry was “D or d” (capital “d” or lowercase “d”)—as an abbreviation for “density in physics or chemistry”—but someone who worked on the entry misread it as a word spelled d-o-r-d instead of “D or d,” and thus, the word “dord” was born in the 1934 edition of Webster’s New International Dictionary.
I like to imagine a bleary-eyed employee looking at it and thinking, “‘Dord.’ Sounds like a word to me!” but actually, when people working on entries typed out the spelling of a word, it was standard to leave a space between each letter, so it wasn’t so far fetched to think that whoever typed “D or d” had meant “D o r d” and simply forgot to put a space between the O and the R. (7)
“Dord” isn’t in dictionaries anymore though. A Merriam-Webster editor discovered the mistake, and the entry was corrected 13 years later, in 1947.
Esquivalience
Not every non-word that ends up in a dictionary gets there by accident though. Some are intentional, such as the one that was invented by an editor at the New Oxford American Dictionary and was included in the 2001 edition to help the company track copyright violators who were lifting entries from the dictionary. If the made-up word Oxford had created appeared in another dictionary, it would be clear that it had been copied from them.
The word was “esquivalience,” which they defined as “the willful avoidance of one’s official responsibilities.” They even gave it a made-up etymology, saying it arose in the late 19th century, perhaps from the French word “esquiver” meaning “dodge” or “slink away.” (8)
Names for Intentionally Deceptive Words
Some people don’t believe that words created on purpose are true ghost words. The Merriam-Webster online dictionary definition for “ghost word” would include such words, but the entry link would not.
And there are, in fact, two other words that language geeks use to describe intentionally deceptive non-words: “mountweazel” and “nihilartikel.”
Mountweazel
Some encyclopedias also include fake entries to catch copyright infringers, and Henry Alford, the author of a 2005 New Yorker article about “esquivalience,” chose the entry for Lillian Virginia Mountweazel in the 1975 New Columbia Encyclopedia to coin a term intentional fakes. Amusingly, the encyclopedia described the fake Ms. Mountweazel as
“a fountain designer turned photographer who was celebrated for a collection of photographs of rural American mailboxes titled ‘Flags Up!’” She was said to have been born in Bangs, Ohio, in 1942, only to die “at 31 in an explosion while on assignment for ‘Combustibles’ magazine.” (8)
You can see why Alford chose “mounteweazel” as the term to use for a fake entry.
And yes, I checked: Henry Alford is the name of a real modern writer. I wondered if that was a ruse because Henry Alford is also the name of a well-known language writer from the 1800s.
Nihilartikel
The German term “nihilartikel” seems to predate “mountweazel” by at least a year or two though as a term to describe an intentionally fake dictionary entry. It’s combination of the Latin “nihil,” meaning “nothing” and the German “Artikel,” meaning “article. I say it predates “mountweazel” by at least a year or two because its origin is a bit in dispute.
Wikipedia and Wiktionary both say it is itself a fake, citing the origin as “a fictitious March 2004 English-language Wikipedia article.” However, the site World Wide Words, which I trust more than Wikipedia says that the word has had “half a dozen appearances in German sources since 2000, more than in English,” suggesting “that it is a real, [but] rare, native German word.”
As an aside, dictionaries and encyclopedias aren’t the only reference works that include fake entries. Maps also sometimes include made-up streets or even towns that publishers can use to track copyright infringement, and these also have multiple names such as “paper towns,” “phantom settlements,” and “trap streets” since they are used to trap plagiarists.
Phantomnation
I’ll finish with one more ghost-y word that arose from a misunderstanding: “phantomnation.”
Reading about these intentionally deceptive words and entries can start to make you paranoid, so when I saw the word “phantomnation” in a Visit for reference video about ghost words, and I’d never seen it before, I started to wonder if I was being duped. But I wasn’t, and the story behind it is far more odd than it being merely a transcription error or an intentional ruse.
While trying to confirm that the Visit for reference video was on the level, I found the word in Google Books from 1891 in a book titled “The Compounding of English Words: When and Why Joining or Separation is Preferable.” Yes, someone wrote a 223-page book about compound words. The author, Francis Horace Teall, seems to have strong feelings about compound words and believes that dictionaries are doing it wrong: they should be much more consistent in how they form compounds.
In one section of the book, Teall describes another author, a Mr. Jodrell, who thought that all compound words should be one continuous word, for example Jodrell wrote “marriagesettlement,” “stagegesture,” and “tapestryhanging” each as one word. No space between “tapestry” and “hanging” for example. It’s all squished together as one word: “tapestryhangning.”
Teall complains that Jodrell even did this when quoting other authors, and this is where we get to “phantomnation.” When Jodrell was quoting a line from Alexander Pope’s translation of “The Odyssey”—“all the phantom nations of the dead”—he followed his “craze for solidifying” as Jodell called it and wrote “phantomnations” as one word instead of two.
Joseph Worcester’s 1860 Dictionary of the English Language then included the word, defining it as “illusion,” and Webster’s included the same entry in 1864. Teall says that even though it meant “a multitude of spectres” in the quotation, the dictionaries interpreted it as the word “phantom” with the suffix “-ation” and an N added in to make it sound better, instead of realizing that it just came from Jodrell’s odd habit of slamming two words together,
The Oxford English Dictionary still includes “phantomnation” today but without a definition. The entry simply notes that it is a misinterpretation of “phantom nation” and includes the citations from Jodrell, Worcester, and Webster’s.
Web Bonus
Other words that arose from errors:
abacot. A misprint of “bycoket,” a kind of cap or head-dress. It appeared in reference books for approximately 300 years before the error was discovered by James Murray, editor of the Oxford English Dictionary. (9)
cherry. According to Merriam-Webster, “The Old North French word for ‘cherry’ is ‘cherise’. English speakers heard the ‘se’ at the end of the word and assumed it was plural, ‘cherries’, and that the singular form of the word must then be ‘cherry.'”
derring-do. Chaucer wrote “in durring don that longeth to a knight” meaning “in daring to do what is proper for a knight.” The phrase was misprinted in a later work by John Lydgate as “derrynge do,” and then taken by Edmund Spenser to mean “brave actions” or “manhood and chevalrie.” Sir Walter Scott used it in Ivanhoe in the manner of Spencer, using the spelling we use today, writing, “if there be two who can do a deed of such derring-do!” (10, 11, 12)
foupe. Multiple sources say that Samuel Johnson’s 1755 dictionary had the word “foupe” when it should have been “soupe” (another word for “swoop”) because the archaic long “s” so closely resembled the letter “f.”
Imogene. The name of the character in Shakespeare’s Cymbeline is hypothesized to be a misspelling of the name Innogen.
Sane (Middle English). In Middle English, “sane” was a verb that meant “to cure” or “to heal.” A work titled Middle English Word Studies: A Word and Author Index lists a 1986 paper by Lister Matheson, and summarizes it as hypothesizing that “sane” was a misreading of the verb “save” (also spelled “saue”) that came from the Latin “sanare,” which meant “to cure” or “to heal.” (13)
- Burridge, K. Weeds in the Garden of Words. Cambridge University Press. 2005. Visit for reference (accessed October 17, 2019).
- gravy. Oxford English Dictionary, Online Version. September 2012. Visit for reference
- Trask, R.L. (ed.) Dictionary of Historical and Comparative Linguistics. Edinburgh University Press. 2000. this link (accessed October 17, 2019).
- syllabus. Oxford English Dictionary, Online Version. Visit for reference
- tweed. Oxford English Dictionary, Online Version. Visit for reference
- Harper, D. “tweed.” Online Etymology Dictionary. Visit for reference (accessed October 17, 2019).
- Alford, H. “Not a Word.” New Yorker. August 29, 2005. Visit for reference (accessed October 17, 2019).
- Quinion, M. “Abacot.” World Wide Words. Visit for reference
- Martin, G. “derring-do.” The Phrase Finder. Visit for reference
- derring-do. Oxford English Dictionary, Online Version. September 2012. Visit for reference.
- Bloomfield, L. Language. Motilal Banarsidass: India. 2005. Visit for reference
- Sylvester, L. and Roberts, J. Middle English Word Studies: A Word and Author Index. D. S. Brewer: Cambridge. 2000. Visit for reference
Ghost image courtesy of Shutterstock.
Best friend cardi image, AnaKika at Flickr. CC BY 2.0.
Mignon Fogarty is Grammar Girl and the founder of Quick and Dirty Tips. Check out her New York Times bestseller, “Grammar Girl’s Quick and Dirty Tips for Better Writing.”