Behind the Dictionary
Lexicographers Talk About Language
A Troop of One
Today is Veterans Day in the United States, and linguist Neal Whitman has been thinking about a question of military usage: if "50,000 troops" refers to 50,000 people, then does "one troop" refer to one person?
The horrifying events at Fort Hood last week have shown us that even in friendly territory, members of the armed forces may lose their lives in service to the rest of us. Whether overseas in active war zones, or within the relative safety of our own borders, our troops willingly accept that doing their job may require the ultimate sacrifice, and their acceptance humbles and amazes me. Regardless of our views on where the United States should send our troops or for what purposes, we can agree that the men and women who serve have undertaken a dangerous but necessary job that would scare most of us, and for that we owe them our thanks. If you see a troop today, or if you know one, please express your appreciation to him or her. If you are a troop, thank you.
Many speakers might have had a problem with those last two sentences — not for the sentiment (I hope), but for the way I used the word troop. How can one person be a troop? Shouldn't I have said soldier, or some suitable word, perhaps servicemember, that would cover members of any branch of the armed forces?
To tell you the truth, I have a problem with that usage of troop myself. I remember reading in my U.S. history class in high school about how President Kennedy at some point had sent 11,300 troops to Vietnam. I wondered how many people that actually translated to. Were there 20 soldiers in a troop? Fifty? It wasn't until I was in college that I finally began to figure out that for sufficiently large values of X, X troops simply meant X people.
In his Political Dictionary published last year, William Safire had this to say on the issue:
Troops is a word in semantic trouble. In one sense, it means "soldiers"; does this exclude sailors and airmen (now grouped as "service personnel")? Troops means "a group of," but so does a troop; the extent of the number is fuzzy. ... A troop means both "one soldier" and "a group of soldiers," which is not what a word is supposed to do.
Before I go any further, I need to lay out some terminology. Troop in the sense of "a group of soldiers" is an example of a collective noun, like group, family, or collection. If you use the plural form troops to mean "more than one group of soldiers" — the meaning I insisted upon in high school — it's a plural collective noun. I will refer to the use of troops to mean "a group of soldiers" and troop to mean "one soldier" as noncollective troop(s).
Concern about noncollective troops seems to be a 21st-century phenomenon. The earliest complaint I've found is in a letter from one of columnist Barbara Wallraff's readers, quoted in her book Word Court (2000). The reader describes an experience much like mine: When history teachers mentioned statistics like 50,000 troops in Vietnam, they wondered if that translated to half a million soldiers. The reader's position is uncompromising: if 50,000 troops means 50,000 people, then one troop is one person, and that's wrong.
A similarly hard line comes from Susan Jacoby, in her book The Age of American Unreason (2008):
As every dictionary makes plain, the word "troop" is always a collective noun; the "s" is added when referring to a particularly large military force.
So for exactly how long has troops been used noncollectively? The Oxford English Dictionary's earliest citation of troop is from 1545, and Google Books turns up plenty of citations from the 1600s onward with troops following large, round numbers. A few of the hits contain enough math clues to make it clear that noncollective troops is intended. For example, in The History and Proceedings of the Second Session of the Third Parliament of King George II, 1742-1743, we find:
[I]f we take 16,000 into our Pay, fresh Troops must be raised for that Purpose, and, I hope, I may say, without any Derogation, that 16,000 Hanoverians newly raised, are not so good as 16,000 of the Veteran Troops of any other Potentate in Europe.
Clearly, the number 16,000 that keeps coming up refers to 16,000 people.
Moving down an order of magnitude, there are references to hundreds of troops from the same time period. In 1757, during the French and Indian War, the minutes of a colonial governors' meeting contain a breakdown of one army to be assembled, including "200 Provincial Troops from Pensilvania" and "200 Troops from North Carolina", which, together with other groups totaling 1600 soldiers, make a grand total of "2000 men."
So the use of noncollective troops with large, round numbers is nothing recent; it has been going on for at least 250 years, and probably longer. Wallraff's response to her reader recognizes this. After pointing out the need for a term to cover any member of the armed services, she writes:
We are to ignore the special qualities of troops when it appears in a context like "5,000 troops were sent overseas," in which it means a body of soldiers and the number is indicating the size of that body.
Patricia O'Conner, author of Woe is I, wrote something similar on her Grammarphobia blog in 2006:
"Troops" (plural), in the military sense, properly refers to a LARGE number of individuals (as in, "Five thousand troops were deployed.") When "troops" is used with a SMALL number, it properly refers not to individuals but to collections of people. "Three troops were attacked" would mean three units were attacked.
So the singular "troop" in reference to an individual soldier ("one troop was slightly injured") is considered a misuse, as is the plural "troops" to refer to a small number of individuals ("four troops were captured"). We do hear and read such irregular usages, though.
The earliest complaint I've been able to find about troops with these smaller numbers starts appears only a year after Wallraff's reader's complaint about troops with any number, but during that interval came the attacks of 9/11 and the beginning of the war on terror(ism). The news reports from Iraq and Afghanistan have told stories involving roadside bombs, rocket-propelled grenades, suicide bombers, and frequent, small numbers of deaths, both of civilians and military personnel. The complaint appeared on the alt.usage.english online forum in December 2001, and took issue with the New York Times's phrasing "four Arab al Qaeda troops."
Complaints for various numbers of troops less than 100 have appeared in one forum or other nearly every year since. In 2005, James Kilpatrick responded to some of his readers, who were complaining about 19 and 31 troops. In 2003, northern California columnist Debra DeAngelo ranted about "twelve troops." Patricia O'Conner's blog entry from 2006 was in response to a reader's complaint about "2 or 3 troops." In fact, "two troops" attracts quite a bit of attention. In April 2004, I myself blogged about a headline mentioning "two troops," and a year later, Geoff Pullum wrote on Language Log about hearing "two troops" on NPR. In November 2007, Ralph Harrington at the Grey Cat Blog complained about a BBC story that referred to "two Danish troops." "If one individual had been reported dead," he asked, "would the headline have referred to the killing of 'one Danish troop'?"
That's the question, isn't it? What happens when we take noncollective troops to its mathematical conclusion, and find ourselves with one troop corresponding to one person, in the most overt possible ambiguity with the collective noun troop? In her blog post, O'Conner rejects such a usage for the same reason as for rejecting troops with small numbers. (Just last week, O'Conner updated the post to allow for small numbers with troops, but is still silent on the singular.) Other grammarians, while grudgingly allowing troops with small numbers, draw the line at one troop. A post from November 2001 on a grammar blog associated with Capital Community College in Hartford, Connecticut states, "It seems awfully clumsy to refer to an individual as a 'troop,' though, and I don't think that happens (even though it would seem logical enough)."
In a 2007 essay on NPR, linguist John McWhorter states, "One cannot refer to a single soldier as a troop." Bryan Garner concurs in the third edition of his Modern American Usage (2009), which finally takes on the question of noncollective troops. Even while admitting as standard phrases such as three troops (for which he provides an 1853 citation), Garner writes that "a single soldier, sailor, or pilot would never be termed a troop."
But it had to happen sometime. On June 9, 2008, Ralph Harrington returned with another post, informing us that CBS News had broken a "stupidity barrier" when they put out the headline "Afghan violence claims 100th British troop."
So the "one troop" stupidity barrier was crossed just last year. Or was it? O'Conner mentioned having read or heard "such forms" prior to August 2006, and the Corpus of Contemporary American English contains this example from 2002: "We do know that there was one troop killed over the weekend." On November 24, 1990, during the Persian Gulf crisis, President George Bush stated, "As long as I have one American troop — one man, one woman — out there, I will work closely with all those who stand up against this aggression."
When members of the American Dialect Society's listserv talked about noncollective troop in April 2003, one recalled being addressed individually as "troop" while in the Army during the late 1960s. The Oxford English Dictionary has "Can you spare a bite for a front-line troop?" from 1947. The June 20, 1874 Sydney Mail reports, "One of the troops was severely handled, being beaten with clubs, and two others received shot wounds." And finally, the OED provides the earliest citation yet, from 1823: "The monkey stowed himself away..till the same marine passed.., and laid hold of him by the calf of the leg... As the wounded 'troop' was not much hurt, a sort of truce was proclaimed."
In short, although noncollective troop(s) with small numbers (including one) has been in existence for almost 200 years, language watchers noticed it only recently. Some reject it with any number; some allow it only with large numbers; some allow it with any number greater than one. And one writer has gone so far as to allow it with any number, including one. It was none other than the curmudgeonly James Kilpatrick who wrote in his 2005 column: "In today's nomenclature, a troop is also an individual soldier."