A few weeks ago, a staggeringly cute YouTube video made the rounds of the usual web sites. It shows a three-year-old girl describing Star Wars. Here it is, for those of you who somehow missed it:
If you're a Star Wars geek like me, you thought this was very cool and made you want to have kids like right now. (My favorite comment on the MetaFilter thread about the video was "made me ovulate".) If you're a linguistics geek like me, though, I'll bet you had a different reaction: "Hmm, what's going on with her postalveolar consonants?"
Before we can form any theories, we need some data. (That's The Tensor's First Law of Half-assed Internet Phonology.) Below, I've compiled a corpus of the postalveolar consonant tokens we have to work with. For each, I've included the context in which it occurred (with the consonant in question in red), the standard pronunciation, the three-year-old's pronunciation, and whether or not her lips were rounded when she pronounced it (a feature of [ʃ] in standard English).
example standard 3-year-old rounded (1) ...sandpeople capture robots... [tʃ] [ts] yes (2) ...a garage sale... [ʒ] or [dʒ] [z]* no (3) ...like [a] garage sale... [ʒ] or [dʒ] [z] or [ʒ]* no (4) ...and the shiny guy... [ʃ] [s] no (5) ...the shiny guy always... [ʃ] [s] no (6) ...kind of a teacher... [tʃ] [tʃ] yes (7) ...he's teaching Luke... [tʃ] [ts] no (8) ...got out [of] jail... [dʒ] [dʒ] no (9) ...in the spaceship... [ʃ] [s] no (10) ...he'll get ya' [tʃ] [tʃ] yes
* The final consonant in both occurrences of garage is extremely hard to make out. The first is almost deleted, and may be nothing more than a lengthening of the initial [s] in the following sale. The second is clearly realized as some voiced consonant, but I can't make out if it's postalveolar or not; again, the following [s] makes it hard to decide.
Notice that when pronouncing most of the tokens whose standard pronunciation contains [ʃ], she doesn't round her lips as most adults speakers of English would. The exceptions are tokens (1), (6), and (10). In the case of (1), the lip rounding can be accounted for as assimilation from a following [u] vowel. In fact, I suspect the same thing is happening with (6); that is, she's treating the final syllable of both words as underlyingly containing a /u/. This lip rounding causes (1) and (6) to sound different from, and somewhat more standard than, the other tokens. At first I thought that they might involve an alveolo-palatal fricative [ɕ], but I think I've convinced myself that I'm just hearing the lip-rounding.
The general pattern, then, seems to be that she's pronouncing all the postalveolar consonants using the corresponding alveolar consonants instead—that is, she uses [s] for [ʃ] and also (maybe) [z] for [ʒ]—with a couple of interesting exceptions. First, she uses the standard [dʒ] pronunciation for the first consonant in jail in (8), and she uses [tʃ] for the second consonant in get ya' in (10). In addition to (8) and (10), she may also be using a postalveolar consonant in garage in (3), and maybe even in (2) as well. Of these four exceptions, three—(2), (3), and (8)—are voiced, which may imply that she's only doing the alveolar substitution for voiceless consonants. It depends, of course, on what you think she's actually pronouncing for (2) and (3), which as I mentioned I can't clearly make out.
The true exception in (10), in which she clearly pronounces a voiceless [tʃ]. I can think of two ways this might come about. First, she may simply have acquired an exceptional lexical item for git ya' with the pronunciation [gitʃʌ]. Second, and more interestingly, she may have acquired the phonological rule that produces the postalveolar consonant in adult speakers, namely the rule that alveolar consonants become postalveolar before the reduced, cliticized [jʌ] form of the pronoun you [update: maybe optionally, and maybe before the unreduced form as well]. That would imply that she's got both a rule turning (voiceless) postalveolar consonants into alveolars and a rule, applying later, that produces surface postalveolars before ya'. Unfortunately, we don't have enough data to decide between these possibilities—we'd need to hear how she pronounced sentences like, "He'll hit ya'" and "I'll miss ya'". (I consider it unlikely, by the way, that she simply has underlying alveolar consonants, since perception usually precedes production, but again, we'd need more data to be sure.)
What do you think? Have I missed something important? This phon-stuff is outside my usual sub-field, so it's possible I've embarrassed myself. If so, please feel free to leave detailed and derisive comments explaining how. I look forward to the HTML-formatted OT tableaux in your more modern analyses. :)
Recent Comments