Various notes on Powhatan

[jamie]

I want to preface this by saying that I am neither an Algonquianist nor an educated linguist by any means, and I’m bound to get some stuff wrong here—but I figured it couldn’t hurt to put a bunch of ideas out for people who may be more familiar with general linguistics or Algonquian languages! I personally prefer using “Powhatan” over “Virginia Algonquian” for the name of the language, but they mean the same thing, and are used interchangeably by many authors.

There’s a few documents that are helpful for anyone who wants to take a further look into Powhatan, and I've added them below:

1. J. P. Harrington’s facsimiles (https://repository.si.edu/bitstream/han ... 89-202.pdf) are incredibly helpful as this pdf is the only (?) available document containing the original manuscript of Powhatan vocabulary from Strachey himself, though his supposed transcriptions of them aren’t great.

2. Wright and Freund’s transcription (https://www.dropbox.com/s/chhmy0m3c2p51 ... ).pdf?dl=0) is much more accurate, though it does not show <ſ, n̄, m̄>.

3. Siebert’s 1975 paper reconstructing Powhatan (https://archive.org/details/siebert-res ... -dead-1975), which also has a few errors, notably opposing the merger of Proto-Algonquian *θ and *r to /r/, and his interpretation of the vowels as /a e i o a· e· i· o·/, instead of Goddard’s (and my) interpretation as /a ə ɑ e i o/, phonetically probably s.t. like [ă ə̆~ɪ̆ ɑˑ~ɒˑ ɛˑ iˑ oˑ~uˑ]. Conceivably /i/ could’ve ranged as low as [e]—cf. Strachey’s ⟨Wawirak⟩ “the hornes of a deare” ← PA *wi·wi·θaki “(his) horns,” phonetically [ˈweˑwiɾăk]?

There was also marginal [(w)ɔ] (~[(w)ɒ]?) from certain instances of PA *wa·, which seems especially difficult for the English to spell:

/wɑhkʷ/ :: ⟨Woock⟩ “Caviare or the Ro of a Surgeon [sturgeon]” ← PA *wa·hkwa “roe, spawn”
/wɑr/ :: ⟨Woor⟩ “a hole” ← PA *wa·ši “hole, trench, burrow” (the root was *wa·θ-, with mutation of *θ → *š before *i, leveled out in Pw.)
/apətew (a)pwɑn/ :: ⟨Apetawh poan⟩ “to Broyle or toaſt bread” ← PA *apwete·wi “it is hot, roasting” & *apwa·ni “(corn)bread, roast”
/apwɑn/ :: ⟨Appoans⟩ “bread” ← ditto, + inanimate plural /-r/ [-ɾ̥] or diminutive /-hs/ “piece of bread”
/mata wəhpwɑkanəw/ :: ⟨Mata vppoannonvwh⟩ “I have no tobacco” ← PA *mata “but, alternatively; no, neg.” & *wexpwa·kaniwa “it (an.) is a tobacco pipe”—I don’t have a good explanation for the ⟨nnon⟩ though, unless it is an error for *⟨Mata vppoacconvwh⟩, which would be more consistent with the PA term.

My thoughts here are summed up well in a passage from an email correspondence I had with Ozaawaa Bineshiinh (I don’t know his name to credit him properly, but his blog can be found here: https://miidashgeget.wordpress.com/):

“(As for the outcome of *wa·, I wonder if Strachey’s ⟨oo⟩ here is actually an attempt to spell something like phonetic [wɔ:] – with the first ⟨o⟩ = [w] and the second = [ɔ:]? Maybe that was the outcome in careful speech, while in faster speech that could be simplified to [ɔ:], or Strachey just had a harder time hearing the [w], so it’s usually spelled accordingly? Siebert has a much more complicated list of outcomes for *wa· which seems dubious, although I haven’t double-checked it.)”

Which makes sense, since PA *wa· was (one) of the sources for Unami /ɔ:/, and was retained in Carolina Algonquian (cf. ⟨Vppówoc⟩ from the same root).

Finally I want to touch on Powhatan /r/, and its relationship with *θ. Sometimes it can be written as supposed /t/ and /r/ with the same PA etyma (⟨Mattoume⟩ “... groweth as our bents”, ⟨Pausarowmena⟩ “a rare dish ... boyled with beanes”, PA *maLo·mini “wild rice” & *pa·ʔsaLo·minahi “grains of dried wild rice (obv.)”), and there was occasional use of ⟨t⟩ for PA *r, like ⟨Outacan⟩ “a diſh” from PA *wera·kana “pan, dish, plate”. I think this pretty clearly shows that they were one segment.

To quote Goddard, both in a personal correspondence and in his 1980 paper Eastern Algonquian as a Genetic Subgrouping:

“PEA *r seems to have several somewhat mysterious reflexes that observers had trouble deciding how to spell. I think Va. Alg. ⟨r⟩, ⟨t⟩, ⟨ht⟩ could have been a simple apical tap.”

“These Powhatan data on their face show that PA *θ and *l were both reflected by segments sometimes recorded as “r” and sometimes as “t,” as well as “s” word-finally; they give no basis for the claim that *θ and *l were kept distinct. Plausibly, the Powhatan reflex of PEA *r (← PA *θ and *l) was a type of flap /r/ (and hence most commonly written as “r”), but subject to optional devoicing that led English-speaking recorders to write it sometimes as “t” (“tt”) or “ht” word-medially and as “s” word-finally (and after /h/).”

That paper (which can be found here: https://ojs.library.carleton.ca/index.p ... ew/782/673) gives many examples of this, but I won’t just repeat the whole thing since it’s available right there.

That’s it for the introduction. I guess I just wanted to get some feedback on ideas / get ideas from other people through discussion? I hope this wasn’t worded horribly, haha.

[Moose-tache]

The notion that a phonemic merger can be hidden by incompetent transcription is interesting, and I'm surprised I don't see more of that in conlangs. This is a very cool project, and I look forward to seeing more.