From reader EG:
I am writing you because I encountered the perplexing singular y'all while watching trailers for Disney's newest film, The Princess and the Frog. Now, not being a Southerner I can't attest to my own usage of "y'all," but my linguistic intuition is in accord with your Language Log posting "Out of the y'all zone" (9/18/2005), namely that y'all is generally not used to address singular individuals, but plural and occasionally implied plurals. […] In the cited trailer, Tiana uses singular y'all three times. Addressing the frog with evident dismay, she says "So what now? I reckon y'all want a kiss." at 0:32. And then again, at 2:14, when the frog is dismayed that she will not kiss him after her apparent offer, she retorts "I didn't expect y'all to answer!" In the intervening time, she does refer to him (using apparently less careless or emotionally influenced wording) as standard second person singular "you." Finally, "Y'all don't look that much different… but how'd you get way up there?" 3:13. This last example is perhaps the most perplexing of all, as it contains both forms.
Both the Wikipedia article for the movie and the IMDb page give screenwriting credit to Ron Clements (born in Sioux City, Iowa), John Musker (from Chicago, Illinois), and Rob Edwards (origins unclear). The character of Tiana is acted by Anika Noni Rose, who "was born in Bloomfield, Connecticut to Claudia and John Rose, Jr., a corporate counsel for the city of Hartford". Thus it's not clear whether anyone associated with writing or acting that scene has native intuitions about the likely distribution of y'all in the speech of a young African-American woman from New Orleans.
So it's a reasonable guess that the sprinkling of y'alls in Tiana's speech is a bit of southern spice added by northern chefs. However, it's worth quoting Jan Tillery and Guy Bailey, "Yall in Oklahoma", American Speech 73(3) 1998;
In spite of the large body of writing on yall, we know very little about the form. For example, we know almost nothing about its social and spatial distribution (i.e., about precisely who uses the form and exactly where it is used) and very little about its origins or even its precise meaning. This paradox is largely a consequence of the peculiar research strategy that underlies a great deal of the literature on yall. Rather than basing their conclusions on surveys of usage or ethnographic studies or even attestations in literary dialect, most of those who have written on yall rely on what is best termed the personal testimony of true believers. Especially in response to skeptics who cite apparent singular uses of you-all or yall that they may have overheard by chance, true believers simply give their personal testimony that these forms never occur as singulars in the South. They often do so with zeal, as in Axley's assertion that, in a lifetime of observation, he had "never heard any person of any degree of education or station in life use the expression you all as a singular" (1927, 343). Even Atwood (1962), in an otherwise excellent dialect survey, relies on the strategy of personal testimony. Although he surveyed the use of yall as a plural in Texas and Oklahoma, Atwood did not investigate its possible occurrence as a singular; he merely asserted that the form could not be used as a singular, adding that "if anything is likely to lead to another Civil War, it is the Northerner's accusation that Southerners use you all to refer to only one person" (1962, 69). In fact, only one study (Richardson 1984) provides anything like systematic evidence on the possible use of yall as a singular (she argues that the form is used only as a plural and that apparent singular occurrences usually reflect Southerners' exaggeration of their dialect for social effect); few studies provide any data at all on the social and spatial distribution of the form, either singular or plural. A century of fervent scholarship on you-all and yall, then, has produced mostly fervor.
Of course, a lot of linguistics, both prescriptive and descriptive, has been based on the "personal testimony of true believers". And I'll repeat that several southerners whose intuitions I trust are categorically certain that they could never use y'all with singular reference — and my observations of their usage seem to bear this out. They also assert that the same applies to everyone in their native speech community — but this seems to me to be less trustworthy testimony.
Tillery and Bailey's survey results
… suggest that the debate over yall singular is not just an academic enterprise but one in which the general public participates as well, and they provide some insight into the motives for using yall as a singular. For example, an 83-year-old woman from Bryan County says that when yall is used, "it needs to be at least two" people, while a 51-year-old school teacher from Major County indicates that "in Oklahoma [yall] can be used for anything." A 61-year-old resident of Cherokee County, though, suggests that it is in Tennessee that anything goes: "for me, [yall is for] more than one; in Tennessee it isn't. I was raised in Tennessee." (This respondent, of course, was coded as a user of yall plural only.) A 73-year-old native of Rogers County says that she uses yall (actually you-all) as a singular and indicates when she would do so: "if I'm real friendly with someone and know them real well, I'll say how are you-all?"
And as they point out, individual instances of usage are almost always ambiguous:
[S]tatements about the use of yall as a singular or an associative plural are as much statements about the investigator's interpretation of what speakers meant when they used the form as they are statements about what speakers actually said. An example will illustrate this problem. Several years ago Bailey called the research office at The University of Memphis and asked a grants officer (a native Southerner) if she had a particular form that he needed. The grants officer said "yes," and Bailey indicated that he would be over to pick up the form in a few minutes. The grants officer then said, "I'll put it in a pick-up tray for yall." This token of yall might seem like a straightforward singular use, but it is possible to give the token an associative reading. It may be that the grants officer, knowing that Bailey was a department chair, assumed that even though Bailey said he would pick up the form himself, he would actually send a secretary over to get it. In this case, yall might refer to Bailey and whomever he might send to pick up the form, even though Bailey was the only addressee. Then again, it might not. The problem here is that we are trying to classify the token as a singular or an associative plural based on what we think the grants officer meant when she used it. Situations like these surely account for many of the disagreements about whether apparent singular uses of yall are true singulars or associative plurals. The only way to resolve such disagreements, of course, is to ask users what they mean when they use yall. Can they use yall for just one person, or does it always have to refer to more than one? This is precisely what our surveys ask.
This ambiguity doesn't seem to apply to Tiana's lines: "I reckon y'all want a kiss"; "I didn't expect y'all to answer!"; "Y'all don't look that much different". As far as I can see, there's no possibility of implicit plurality there, associative or not.
Of course, the whole question here is whether those lines are a Yankee interpolation. But T & B also list (pp. 272-275) a baker's dozen examples of apparently unequivocal singular yall from recorded sources, starting with this conversation between Bob Wills and Tommy Duncan from a 1937 recording ("Oozlin' Daddy Blues", Vocalion 03693):
Bob: Lord, Lord, Tommy.
Tommy: What's the matter?
Bob: You got your dictionary with you?
Tommy: What yall want with a dictionary?
Bob: I want to look up this new word, this here ((word …))
Tommy: What word's that?
Bob: That word oozlin', boy.
Returning to the surveys,
The testimony of many Southern linguists notwithstanding, nearly a third of the Oklahomans surveyed acknowledged using yall as a singular. We might be tempted to dismiss these results as spurious, but the convergence of results from two surveys with different methodologies (one a telephone survey making use of a random sample, the other a field survey relying on a purposive sample) suggest strongly that the results are an accurate depiction of yall in Oklahoma. The data from SFP [myl: Southern Focus Poll, a south-wide survey — see the paper for details] provides additional confirmation and suggests that in regard to the use of yall, both plural and singular, Oklahoma is much like the South as a whole. […] Thus given the convergence of results from the two SOD [Survey of Oklahoma Dialects] surveys and SFP, we cannot simply dismiss yall-singular as just a devious Yankee rumor.
There's some indication that singular y'all, though perhaps not a devious Yankee rumor, is an informal usage in a way that plural y'all isn't:
As part of [a Memphis business communication survey], Tillery asked business professionals if they used yall in their oral business communications and if they ever used yall for just one person in those communications. This survey differs from SOD and SFP in that it explores language in a more formal context, so we might expect lower rates of acknowledgment. Even in this formal context, however, some 15% of the respondents (33 of 220) acknowledge using yall for just one person.
Looking over the correlations with demographic variables, they conclude that
[Singular yall] is more likely to be used by better educated Oklahomans than by less educated ones, by urban residents than by rural ones, by middle-aged adults than by older or younger ones, and by men than by women. The data from SFP confirms all of these except for the rurality constraint. …
And they suggest as a hypothesis that
[Singular yall] is used by native Southerners, especially those who live in areas with large numbers of non-Southerners or who are in contact with non-Southerners in their work, as a badge of local identity, that is, as a way of affirming local values in the face of widespread migration into the area by outsiders who (often unwittingly) pose a threat to local values.
New Orleans is not Oklahoma, and Tiana is neither middle-aged nor male. On the other hand, Naveen (whether as frog or as prince) is certainly an outsider. And maybe African-American usage of singular yall is more advanced — Tillery and Bailey don't give a statistical evaluation of that factor, though they cite a number of singular yall examples from African-American speakers.
[Update — I should also have included this second hypothesis from Tillery and Bailey:
Thomas Nunnally (1994) has offered a second hypothesis for the emergence of yall as a singular. He suggests that it may well be expanding to fill the role of a polite singular, just as you did several centuries ago. He points out that many of the citations of yall-singular show the form occurring at the edges of discourse-in greetings, partings, and so forth. The following citation, provided to us by Robin Sabino (1994), certainly fulfills this function. Sabino overheard an African-American waitress in an Opelika, Alabama, restaurant say to a customer eating alone, "How are you-all's grits?"
See also Jan Tillery, Tom Wickle, and Guy Bailey, "The Nationalization of a Southernism", Journal of English Linguistics, 2000. And for a contrary opinion on the grammaticality of singular y'all, see Ron Butters, "Data Concerning Putative Singular y'all", American Speech 76(3):335-336 2001.