# Iterated reduplication?

Some friends and I were discussing whether you could reduplicate an already reduplicated string. I came up with this rather contrived example in English to show what I'm talking about:

Person A: So, do you like her?
Person B: Of course I like her!
Person A: Yeah, but do you like her, or do you like like her?
Person B: I mean... okay, I like like her, but I don't LIKE like like like her.

(Paraphrase: I do like her as more than friends, but my feelings are not outstandingly passionate.)

However, while I think a lot of native speakers could extract the intended meaning from that sentence, I don't think people would naturally produce it very readily. So, my question: Are there any natural languages in which reduplication can apply iteratively like this?

• you probably can't sing well enough to give it a suitable intonation. Also it only works because of like as a modal. However, I woulxn't doubt that were the source source of the modal in the first place – vectory Nov 12 '19 at 22:44

I am not aware of any particular reduplication process that lends itself to reiteration. It is worth pointing out that jlawler's example contains two different morphological processes utilizing reduplication which can combine, but it is not the same reduplicative process reiterated again, which might be what you're after. Your example shows, I think, that reduplication for intensity/contrast could be iterative; but whether other types of reduplication can be so, I am not sure.

There might not be robust reduplication in English per se, but there are similar types of recursive structures which might interest you, which I've pointed out below.

One example is anti-anti-anti-missile-missile-missile. I.e. to defeat your enemy's missile you need anti-missile, then they'll need an anti-(anti-missile)-missile, at which point you'll have to build anti-(anti-(anti-missile)-missile)-missiles. Another is: The mouse the cat the dog chased bit ran. That is, the mouse (that) the cat (that) the dog chased bit ran.

There are processing limitations, which is why these center-embedded strings sound funny and are not common (if they are present at all) in natural speech. Ability to easily interpret them rapidly diminishes beyond one embedding; the same limit, it seems, for reduplication. Reduplication like center embedding strains our ability to keep track of structural relations between words, by forcing us to keep a word "active" in memory longer before we know its full context. So it's not surprising that in speech people rarely produce them. But, as you write, they can be processed with effort -- just like you can write out and do a complex math equation on paper that you couldn't solve in your head. And people will recognize them as grammatical after the parsing.

Here is a paper on this type of reduplication in English:

Contrastive Focus Reduplication in English (The Salad-Salad Paper) Jila Ghomeshi, Ray Jackendoff, Nicole Rosen and Kevin Russell Natural Language & Linguistic Theory Vol. 22, No. 2 (May, 2004), pp. 307-357 (article consists of 51 pages) Published by: Springer. (pdf)

(Update 1/15) Here's a relevant LL post on the subject: Ask Language Log: Raped-Raped-Raped. This post gives several more examples of the recursiveness of this process, as well as more links to articles, for those interested.

Oh, yes. But not in English; English barely has any morphology, let alone real reduplication. You have to look at the right language.

In Lushootseed, a polysynthetic Coast Salishan language on Puget Sound, there are seven or eight kinds of reduplication. Several of them can be used together. For instance, Augmentative/Plural is initial full root reduplication,

1) čʼƛʼačʼƛʼa ‘rocks’ < čʼƛʼa ‘rock’

while Diminutive/Attenuative is initial CV- root reduplication (occasionally with vowel change, as here).

2) čʼičʼƛʼa 'pebble'

And these can be used together.

3) čʼičʼƛʼačʼƛʼa 'pebbles'