Why isn’t it a gumsballs machine?
Have you ever looked at one of those gigantic gumball machines that hold thousands of gumballs, and wondered why it’s called a gumball machine instead of a gumballs machine? Or why people who manage projects in the corporate world are called project managers instead of projects managers? You may have figured it was just one of the peculiarities of English that the first noun in a compound noun has to be singular. But then what about singles bars, systems analysts, and the rewards cards that so many stores are forcing us to get these days?
Let’s start with some vocabulary we’ll need to talk about compound nouns. In a compound noun such as “gumball,” the second noun, “ball,” is called the head noun. The head noun tells you what kind of thing the whole compound is referring to: A gumball is a kind of ball. The first noun, “gum,” which modifies the head noun, is called an attributive noun or a qualifying noun.
Singular or Plural?
Linguists have been trying to figure out what’s going on with singulars and plurals inside English compound nouns for at least 30 years. It turns out that there are some discernible patterns and tendencies, but unfortunately, they’re subject to a lot of variation and exceptions, which makes them practically useless if you’re just trying to figure out whether to go with a singular or a plural. We’ll look at these tendencies, and then give our Quick and Dirty Tip for dealing with attributive nouns.
David Crystal, a linguist who has written many books on English, notes that when there’s a meaning difference between a compound with a singular attributive noun and one with a plural, people usually keep the plural (1). For example, if I talked about a reward card instead of a rewards card, that could sound like a card that was given to me as a reward. I also found out that system analysis is a field of electrical engineering, so it makes sense that the business-related term “systems analysis” uses the plural noun “systems” to distinguish itself. For yet another example, “sale” often refers to an event in which normal prices are reduced. “Sales,” however, refers to all purchases that a store’s customers make, whether during a special-event sale or not. So to refer to an employee who deals with everyday sales, we say “sales clerk” and not “sale clerk.”
Words That Are Only Plural When Alone Stay Plural
Crystal also points out that if a noun is a pluralia tantum—that is, a noun that exists only in the plural—it will probably stay that way as an attributive noun. So we have “economics textbooks,” “clothes hampers,” and of course, “Thanksgiving.” Of course, there are exceptions, in which even nouns that supposedly don’t have singular forms get one just for use in compounds. For example, what do you call the legs on a pair of pants? In the Corpus of Contemporary American English, which I’ll talk more about in a minute, “pant leg” beats “pants leg” two to one.
A Noun Made Up of Different Kinds of Things Is Usually Plural in Compounds
But what about when you’re dealing with an ordinary noun that has regular singular and plural forms, and when you put it into a compound noun, you can’t tell if it makes a difference whether you use the singular or the plural? There are a couple of tendencies at work in these cases. First, linguists have observed that plural attributive nouns are more likely when you’re thinking about different kinds of that noun. So for example, an enemies list contains specific people you need to keep an eye on, and a publications catalog lists different publications available to order (2). Notice that it’s not enough that there be lots of the thing the attributive noun refers to; there have to be significantly different kinds. That’s why someone who keeps bees is still a beekeeper and not a bees-keeper, even though the hives may contain thousands of bees.
For the bad news, this reasoning would also predict that we would say "projects manager" instead of "project manager," since they manage many kinds of project. Again, we’re facing exceptions.
Abstract Nouns Are Usually Plural in Compounds
The second tendency is that plurals are also more likely if the attributive noun is abstract. For example, the abstract noun “admission” is usually in the plural, in compounds such as “admissions committee,” “admissions policies,” and “admissions department.” Conversely, the concrete noun “rock” stays singular in the compound “rock pile (3).”
A Corpus Helps
With all these exceptions and all this variation, how can you actually decide whether to go with singular or plural for an attributive noun? These days you can search through multimillion-word collections of digital text, known as corpora, and find out exactly which possibility is preferred. Mark Davies of Brigham Young University has created several useful corpora, including the Corpus of Contemporary American English and the Google Books Corpus, and they’re easy to use. I’ll put a link on the transcript (http://corpus.byu.edu).
Corpora Versus Google
You may be asking, “Why bother with these corpora, when I can just use Google?” For one thing, nobody knows exactly how Google counts the number of hits it finds, and their algorithm is continually changing. On the first page of results, Google may claim to have several million hits, but by the time you get to the last page, the number will have shrunk to maybe a few hundred thousand. So you can’t get a good comparison of, say, “benefit cuts” and “benefits cuts.” Also, many of the results are duplicates. If you want a more accurate picture of how people are using the language, a corpus is the way to go, and that’s our Quick and Dirty Tip for attributive nouns.
This article was written by Neal Whitman, who has a PhD in linguistics and blogs at literalminded.wordpress.com.
1. Crystal, David. May 18, 2020. “On plural adjectives.” Blog post on DCblog. Accessed July 21, 2011. http://david-crystal.blogspot.com/2010/05/on-plural-adjectives.html
2. Pinker, Steven. 1999. Words and Rules. New York: HarperCollins. 185.
3. Sneed, Elisa. 2002. “The acceptability of regular plurals in compounds.” Proceedings of the 38th Meeting of the Chicago Linguistics Society, 617-631.