I have a friend, a native English speaker from Boston, MA, USA (I believe he is mostly Irish American), who is absolutely adamant that the first sound in "thank you" is voiced, rather than voiceless. Moreover, one of his college English professors (also a native English speaker, but not sure where from) agrees.
My friend acknowledges that all the dictionaries seem to say that the sound is voiceless; moreover, I was able to produce numerous examples from YouTube where a native speaker clearly uses the voiceless "th" (e.g., Matthew McConaughey and Viola Davis).
And still he---and his professor---insist that, in their daily lives, they have never heard the "th" in "thank you" pronounced as anything other than voiced.
My question is: is the voiced "th" in "thank you" something that is well-known and documented as a regional variant? Or is there something else going on here?
(I also posed this question on the Linguist List, here... no answers yet.)
Alex B. pointed out that there indeed exists some published research: The Survey of English Dialects (Orton et al. 1962-71) recorded "a number of instances of the voicing of all four fricatives in word-initial (Anlaut) position in the South and the South-West Midlands" (Fisiak 1988); see Voitl 1988 (also here). Related research includes that by Martyn Wakelin, esp. Wakelin and Barry (1968).
From the comments by musicallinguist, it seems that sometimes even among siblings raised in the same household there can be a difference in the voicing of the initial "th," which persists into adulthood.
I have discovered that, several months prior to my original posting of the present question, essentially the same question had been posted to the English stackexchange. The original asker was motivated by the observation described in 2. above, in her own family (though it is unclear what are the ages of her siblings who pronounce the "th" as voiced). One contributor there stated that the voiced "th" in "thank" can be found "[i]n parts of New England," but, as of yet, no one has provided published references for this.
In light of this, let me refine my question a bit.
A. Is there any published research on the situations where even among native speakers who were born and raised (and live) in the same area, or even the same household, there are some who pronounce the "th" as voiceless, and some who pronounce it as voiced? Are those with the minority pronunciation usually aware that they are in the minority, or are they usually not really aware of it?
B. Also helpful would be examples of (research on) other kinds of within-group linguistic differences---those that are in some relevant sense 'similar' to the difference in the voicing. As far as what exactly should count as 'similar,' well, that is itself kind of a part of the question. Here is one proposal for what it may mean: the differences in some linguistic feature such that 1. one would not expect this feature to vary within such a (linguistically close) group of people; 2. the group members exhibiting this feature are clearly in the minority, but, at the same time, they are also clearly a group and not just a couple of isolated cases; 3. there is no relevant pathology (e.g. impaired hearing or speech processing or speech production) that could explain the existence of the subgroup with the minority feature.
C. Further references on regional variation with respect to the voicing of the initial "th" would also be very much appreciated.