Fiddle-Faddle! The Delight of Reduplication

Years ago, when the furniture in the Language Lounge was still spick-and-span, I wrote a column about reduplication. Not a day has passed since then that I did not use, hear, and delight in one or more reduplicative words; they constitute a reliable source of infotainment in English, and no speaker's lexicon can or should be without a ready supply.

By way of review: reduplications come in three basic flavors in English, namely:

1. reduplications narrowly defined, that is, where an identical syllable or pair of syllables is repeated to form a word. This can be called the yo-yo type.
2. rhyming reduplications, in which parts one and two of a word differ only by having a different initial consonant sound (that is to say, in which rhyming words or syllables fuse to form a single term). This can be called the claptrap type.
3. vowel-shift reduplications (my personal favorite), in which identical consonant sounds or clusters festoon two different vowels in successive parts of a word. This is the flimflam type.

Type 3 is perhaps the most open-ended category since there seems to be no requirement that either component of the word have standalone status as a word, and in cases where one component is a credentialed lexeme, speakers have license to make the other component fanciful; as seems to be the case fiddle-faddle, criss-cross, and perhaps tittle-tattle.

What I've been thinking about lately, and that I hope readers can share their insights and knowledge about, is this: just how much license actually exists in category 3? Of the many reduplications of this type, a striking number show a tell-tale pattern: the vowel in the first component is "short i," or what phonologists and IPA-aficionados call /ɪ/. The contenders I noted in the old Lounge piece fitting this category include willy-nilly, dilly-dally, flim-flam, whim-wham, flip-flap, clitter-clatter, tittle-tattle, criss-cross, mishmash, pitter-patter, and  riffraff. The last four in this list are all from the 15th century, and among the earliest noted reduplications in English.

And these, of course, are the tip-top of the iceberg. There's also clip-clop, jingle-jangle, King Kong, singsong, ding-dong, Ping-Pong, jimjams. Splish splash, I was takin' a bath. The strong preference for this pattern in reduplication looks to be such that, if no lexeme exists containing the preferred starter vowel /ɪ/, speakers will make the first component of the reduplication, rather than the second one, the fanciful part. This seems to be the case in clitter-clatter, criss-cross, mishmash, pitter-patter, and splish splash.

This starts to look not like license but rule implementation. So the question is, why do so many vowel-shift reduplications exist in which the first vowel is /ɪ/? I'm not the first person to have noted the pattern. Indeed, personages far more lofty than I have noted it, and those personages, bearing advanced degrees, have advanced theories about the phenom. To understand the theories, a short primer on vowel formation is in order.

Phonologists slice and dice the qualities of vowels in a number of arcane ways but for our nutshell purposes here, the important distinctions are: long and short (a distinction familiar to most speakers), high and low (a distinction based on the position of the tongue in relation to the roof of the mouth when the vowel is articulated) and front and back (a distinction based on the position of the tongue relative to the front of the oral cavity when the vowel is articulated).

One theory about the prevalence of  /ɪ/-/*/ reduplications comes from Steven Pinker, in his well-known book The Language Instinct (1994). He claims that:

Words that connote me-here-now tend to have higher and fronter vowels than verbs [sic] that connote distance from 'me' . . . Words that connote me-here-now tend to come before words that connote literal or metaphorical distance form 'me' . . .The syllogism seems to be: 'me' = high front vowel; me first; therefore high front vowel first.

Another theory is put forth by Israeli linguist Reuven Tsur in his 1992 book, What Makes Sound Patterns Expressive? Tsur's idea is based on his perception that back vowels are more cognitively demanding than front vowels (because they are more difficult to distinguish phonetically), and his hypothesis that "our cognitive economy tends to relegate to the end of the phrase (or clause) anything that requires relatively great processing effort." Because of this, he advances the explanation that it is a natural tendency to prefer high and front vowels at the beginning of reduplications, while pushing low back vowels to the end.

Pinker's theory, which we might think of as representing the "me first" school, cannot be expected to explain the phenomenon in languages other than English, unless it is a language in which the first personal pronoun contains a high front vowel. Tsur's theory, from the cognitive load-shifting school, ought to apply universally if it is valid.

Both of these theories are respectfully examined and then sent packing in an interesting paper by Maria Beldon, "Order Preference for Reduplicated Words with Differing Vowels." Ms. Beldon designed an experiment to test whether native speakers showed a preference for vowel ordering in the way that Pinker's and Tsur's theories would predict. Her data does not support the thrust of either thinker's theory, though her results "suggest that English speakers prefer the vowel [ɪ] before the vowels [ɛ], [æ], and [ɑ] in reduplicated words," without an explanation or hypothesis about why this might be.

The question that remains then is this: why does English exhibit such a rich collection of similarly-patterned reduplications, in the absence of a viable theory that explains them? There is no evidence to suggest that speakers are constrained to propagate reduplications that instantiate only a preferred vowel ordering, but there is also a dearth of viable reduplications that show a contrary pattern. We spoke with the Lounge's consultant phonetician Will Styler, who had some interesting observations about the phenomenon, including a supposition that the existing store of similar reduplications is effectively a pattern that speakers use, no doubt unconsciously, to create new reduplications:

This is prime territory for a construction of sorts to emerge by analogy, even without a rule. I, as a native speaker, have the word "ring" and want to make a homey expression, say, for the din of hundreds of phones ringing, say, in a news office. Clearly, then, it's "ring-rang", following the same pattern as all the others.

Is he constrained by a rule in his coinage? He says, "we should never attribute to constraint what is more easily attributed to laziness."

So I ask you, native English speakers: do you prefer [ɪ]-first reduplications, and if so, why? You can say that "they just sound right," but you will not be contributing to the advancement of science if that's all you have — everyone concurs on this. Is it possible that this preference arises because so many irregular verb paradigms show this same kind of change in the infinitive > past tense form (such as ring-rang)? For speakers of other languages: do vowel-shift reduplications in your language show this, or any such discernable pattern?