Why Do People Say ‘A-Whole-Nother’?
Syelle Graves investigates why people say “a-whole-nother.” It’s the same reason we say “an apron” instead of “a napron.”
Sponsor: This episode is brought to you by NatureBox. Discover smarter snacking with a new NatureBox each month. Get 50% off your first box when you go to NatureBox.com/grammar.
Last week, in Part I, we learned about cool linguistics stuff such as morphemes and affixes, and decided that the way whole is splitting up another in a-whole-nother looks a lot like loose definitions of tmesis or infixing, but in fact, it’s not those things. It turns out that what matters here is the history and status of the word another.
Determiners Introduce Nouns
Merriam-Webster says another is an adjective, but linguists make a finer distinction than traditional grammar does, in that some adjectives, the ones that aren’t very “content-full” and are instead more “functional,” are given the more specific label determiner.
Determiners introduce nouns, and include articles like the, and possessive adjectives like my, but they do not describe nouns the way true, content adjectives do (think green or nice). (Nouns are content words too, rich with semantic meaning, such as apple or sunset, while prepositions are function words, in that prepositions like of are extremely difficult to define.) So, another is certainly a determiner; it introduces nouns, as in, another apple, but doesn’t modify them, like the true adjective ripe in ripe apple.
Next, what do we use another to express? We could summarize it as indicating “a second one” or “an additional one.” Notice that any context calling for That’s a-whole-nother thing! requires some original “thing” being contrasted with the new “thing.”
Whole Doesn’t Really Get “Infixed”
So, why can’t we just say that people have been pseudo-infixing the word whole into the word another? Let’s start with linguist James McMillan’s brief assessment of a-whole-nother as being metanalysis—not infixation—back in 1970, in which he cited examples dating back to 1958. Metanalysis, also spelled—incidentally—meta-analysis, or called rebracketing, is the idea that word boundaries—where one word ends and the next begins—can change or shift over time; it’s also a concept that we described in the article “How A Napron Became An Apron.” McMillan’s argument in favor of a-whole-nother not being infixation is that with that whole there, the n in another is “moveable.”
Another argument in favor of calling the a-whole–nother phenomenon rebracketing, instead of infixing, is the fact that it is an isolated case, not a pattern that can be applied to many words, and that is just as most cases of rebracketing are. We don’t insert whole into any other host words! Infixing, on the other hand, is usually what linguists refer to as a productive process. Notice how freely we can “infix” expletives, as in Ala-frickin’-bama, as long as we follow the syllable rule we talked about last week, into all sorts of host words, and remember how -iz- inserting works the same way.
These individual cases of rebracketing can also occur as temporary errors. The funniest ones are usually produced by children who can’t read yet and haven’t learned to spell, and some of the more well-known ones include thinking that Pulitzer Prize is pullet surprise, or replying to the order “Behave!” with “I am being -have!”
Another Used To Be Two Words: An Other
So, aside from a-whole-nother being an isolated case, how else do we know that another is getting rebracketed? Well, in a more general sense, although another appears to be a single word, it also looks suspiciously like a fused form that may at one point have been two words: an other. So, we hypothesize that, at some point in the past, writing conventions were to spell another as two words: an other. (Spelling has become pretty standardized now, which is a good thing, but it was not so long ago that spelling and punctuation were in flux, in the United States and elsewhere. Here is one example: As recently as 1930, in an original-print hardcover of a novel in the Nancy Drew series, tomorrow is spelled to-morrow! Remember when books even older than that used expressions like on the morrow?)
Over time, re-analysis of the word boundary, probably because speakers seldom put anything between an and other, led to spelling conventions changing. Intuitively, we all agree that an and other are both real words that we use elsewhere all the time. Also, another can only be used to introduce singular nouns (e.g., we can’t say another girls), which means that there is an “indefinite-article feel” to the an in another. Another way of saying this is that the an inside another still has the meaning and function of separate indefinite article an.
And, guess what? The Oxford English Dictionary confirms this suspicion that another was technically two words in the recent past, in its etymological listing for another. They report that, originally, the an was separate from the other, and at different times, it was found both with the a separate from the nother, and even (less often) as two words with no “n” at all, like a other. They also confirm that another can be summarized as meaning a second (one), a remaining (one), or a different (one).
[Added 6/20/2014: See line 15 in this old manuscript for an example of a nother. Hat tip to Matt Gordon.]
This means that although we spell another as a single determiner, the string another friend [determiner + noun] could theoretically and accurately be reanalyzed—in terms of its meaning—as the three-word sequence [determiner + adjective + noun], parallel to the three-word sequence an intellectual friend.
In Speech, Breaks Between Words Are an Illusion
How do these word-boundary changes happen? Well, word boundaries are fascinating, because acoustically, human speech is a continuous stream of sound, and the spaces that we “hear” between words are an illusion. For most of human language history, there were no writing systems, which partly explains why children learn to speak and understand language at a very early age, quickly and effortlessly, by mere exposure, but yet must laboriously learn to read and write in school over a period of many years. The point here is that word boundaries are not always as natural, obvious, or easy to define as you might imagine. For example, as we add new words to the lexicon, especially compound words like website, we often start out spelling them with a space (web site), and then over a short time we “reanalyze” them as a single, compound word with no space. But, underlyingly, is it one word or two? How do we know? Really, we don’t; we just go by the convention.
In fact, studies have shown that very young children, who can’t yet read or spell, often analyze word chunks like an apple or the monkey as one single word. Why? Well, primarily because articles like an and the are determiners, which are those “function” words, and they come before apple and monkey, which are those “content” words. If asked to provide a word, any word, you are not likely to think of saying a function word like “the.” So, these function words—like the and an—that do not provide true meaning, but do serve a grammatical function, and, usually appear right before content words, are likely to be involved in the shifting of word boundaries, and/or spelling conventions, over time.
Another fascinating aspect is the fact that nearly all cases of nother are spoken aloud, right? Can you imagine typing out nother, even in an informal e-mail? The spelling convention would probably prevent you. So, the fact that we almost exclusively hear this phenomenon makes a lot of sense, because that is how it has managed to override our spelling conventions, just like the way those young children who haven’t yet learned to read sometimes analyze behave as [verb + adjective], be and “-hayve”, instead of as just a [verb], and, like the way any adults who think it’s “pullet surprise” (and there are a few such adults) have certainly simply never seen the name of the award written out before.
(Well, mostly, except in cases of intentional mockery, like this old cigarette ad. Notice how they used an apostrophe to indicate the break in the word!)
Why Do People Say A-Whole-Nother?
Now we understand the phenomenon behind this expression, but we should also know how exactly it occurred in this particular case. Why not just say a whole other thing? Well, probably because speakers wish to express a specific nuance: A speaker references something, and then references an additional thing with the expression “another thing,” but wants to add the notion of “whole,” or “separate” to it, to emphasize the complete separateness of “thing 2.” On the blog English Stack Exchange, SmyWord comments, “I want to emphasize another so I add whole into the phrase.” He wants to emphasize the “addition” feel of it, and getting rid of the n would not allow such emphasis. (Think of that Oxford English Dictionary definition.)
Why re-analyze the an to a? SmyWord continues, “Whole, starting with a consonant, takes the article a, leaving over nother to complete the phrase.” SmyWord is absolutely correct: Whole begins with consonant “h,” which makes speakers need to put whole after the “a,” not after the “n,” yet not feel right saying a whole other thing as much, because losing the “n” loses the emphasis on an additional one or a different one that is encoded in the word another. [See Note 1.]
(For more on using a before words that start with a consonant, and an before words that start with a vowel, see “Why Do We Have Both A and An?” .)
So, it looks like speakers don’t quite “infix” whole into another; they “feel” like they are in fact uttering three separate words to introduce “thing.” (And, in a way, they are!) Because whole is serving as an adverb meaning “entirely,” a whole nother thing is a bit like the word sequence a very different dress: [determiner + adverb + adjective + noun]
Why Don’t People Say Another-Whole-Thing?
Last but not least, why don’t people feel inclined to put the whole somewhere else, like another whole thing? The answer is a difference in semantic scope. Another whole pie means a second pie in its entirety, right? Whole is an adjective modifying pie. But, a-whole-nother pie or the more-correct-yet-less-emphatically-satisfying a whole other pie means “an entirely different one” (in type, or flavor, perhaps), and whole is an adverb modifying adjective other. More important, when we use a-whole-nother to introduce a non-count noun, as in, a-whole-nother idea, we really cannot say another whole idea, because another would apply to whole idea, which speakers don’t want. Whole, again when used as an adverb meaning “entirely,” cannot modify the noun idea. What the speaker wants is for whole to apply to the another and to the idea, as an adverb. As you can see, another-whole-thing just doesn’t have the exact same meaning as a-whole-nother thing.
It’s still true that a whole lotta people dislike a-whole-nother, and many avoid it, but, we hope that if you’re one of those people, you now at least have an answer to the question of why people do it (and maybe even find it fascinating).
Syelle Graves has a master’s degree in linguistics, and is a linguistics index editor for the International Bibliography of the Modern Language Association.
http://syellegraves.ws.gc.cuny.edu/
References
another. Merriam-Webster Online Dictionary. http://www.merriam-webster.com/dictionary/another (accessed May 10, 2014).
another. Oxford English Dictionary (accessed May 11, 2014).
“A whole nother” way of looking at things. English Language and Usage. http://english.stackexchange.com/users/48/jsb (accessed May 9, 2014).
Curzan, Anne, and Adams, Michael. (2012). How English Works (3rd. ed.). Boston, MA: Longman.
Fromkin, Victoria, et al. (2011). An Introduction to Language (9th ed.) Boston, MA: Wadsworth.
infix. Wikipedia. http://en.wikipedia.org/wiki/Infix (accessed May 8, 2014).
Keene, Carolyn. (1930). Nancy Drew Mystery Stories: The Secret of the Old Clock. New York: Grosset & Dunlap Publishers.
Kirk, Carol. (1979). Patterns of word segmentation in Preschool Children. Child Study Journal, 9(3), 37–49.
Lindsay, Mark. American English iz‐infixation: Interaction of phonology, metrics, and rhyme.
McMillan, James B. (1980). Infixing and interposing in English. American Speech, 55(3), 163–183.
Myler, Neil. (2014). personal communication.
O’Grady, William, et al. (2005). Contemporary Linguistics (5th ed.). Boston, MA: Bedford/St. Martin.
whole. Merriam-Webster Online Dictionary. http://www.merriam-webster.com/dictionary/whole (accessed June 12, 2014).
tmesis. Wikipedia. http://en.wikipedia.org/wiki/Tmesis (accessed May 10, 2014).
Note 1: Some linguists may point out that there is an additional phonological (unconscious English sound rule) reason: We see that keeping the “n” is important for expressing that “additional” property, and, the word whole also winds up before the stressed syllable of another (just like expletive infixing—a common English stress pattern). Since, as we saw, an other has been largely re-analyzed as single word another, added linguistic material can’t go after the “n,” because doing so would split the stressed syllable in half. However, this syllable integrity rule would indicate that speakers think of the another in a-whole-nother as being a single word, while the clear metanalysis indicates that speakers are putting the “whole” between “two words.” The psychological reality of where speakers perceive the word boundaries in another when spontaneously producing a-whole-nother is difficult to ascertain anecdotally. It is likely an interface of phonological and semantic rules, and likely that the “moveable ‘n’” shows that both bracketed forms of another may optionally exist for speakers.
You May Also Like…