126.96.36.199:57: BACK ON TRACK
William Rozycki included Manchu jugūn 'road' in his Mongol Elements in Manchu (1983: 171) - and rejected it as a Mongol element "on semantic grounds" since the corresponding Mongolian word jüg meant 'direction', not 'road'.
I would add two phonological arguments:
First, if Mongolian jüg (or some similar ancestor) were borrowed into Proto-Tungusic (or inherited from Proto-Altaic*), it should have had *i instead of u in languages that shifted *ü to i**: e.g., Evenki *jiɣun instead of juɣu 'direction'. I agree with Rozycki (1983: 171): Tungusic juɣ-words meaning 'direction' "are recent borrowings from Mo[ngolian]". They cannot predate the *ü to i shift and are not cognate to Manchu jugūn 'road'.
Second, the near-high (but not truly high) ū [ʊ] of Manchu jugūn [tʂuʁʊn] and the mid o of its Sibe cognate ǰoxon [tʂoʁon] 'id.' (Kim et al. 2008: 80) indicate that the word 'road' had nonhigh vowels, not high vowels like Mongolian jüg. I would expect Mongolian jüg to correspond to Manchu *jugun [tʂuɣun] and Sibe *ǰuxun [tʂuxun] with high vowels.
Their Jin Jurchen 'ancestor'***
'road' (Jin 1984: 30-31, 95, 167)
may have been pronounced [tʂʊɢʊ] and could be romanized as jūgū. *ʊ raised to high after Manchu j but not g and lowered to o in Sibe****.
188.8.131.52:46: The Ming Jurchen word ju transcribed as 住 [tʂu] in the Sino-Jurchen vocabulary of the Bureau of Interpreters (Kane 1989: 162, #133) may have been [tʂʊ(ː)] or [tʂu(ː)]. There was no way to distinguish between [ʊ] and [u] or indicate long vowels in Ming Chinese transcription.
It is not clear whether earlier [ɢʊ] fused with the preceding vowel, possibly lengthening it, or was a suffix (as proposed by Jin 1984: 31) that was either lost or always absent in that dialect.
I am skeptical that [ɢʊ] was a suffix since I know of no Manchu suffix -gū -[ʁʊ].
Moreover, the transcription 住兀 [tʂuu] for
in the Bureau of Translator vocabulary (Grube 1896, #57) may point to a long vowel but it is possible that the Bureau of Interpreters dialect lost it. However, 住兀 [tʂuu] may have represented a Ming Jurchen [tʂʊʁʊ] with a [ʁ] without an exact equivalent in Ming Chinese.
*184.108.40.206:59: I do not think a Proto-Altaic language existed.
**220.127.116.11:01: According to a 1996 class handout by Sasha Vovin, Proto-Tungusic *ü merged with *i in some languages like Evenki and merged with *u in others like Manchu.
***18.104.22.168:04: Jin Jurchen may not be directly ancestral to Written Manchu and Sibe, but it is certainly very closely related to their ancestors.
****22.214.171.124:31: At a glance, it seems that Written Manchu ū corresponds to Sibe o which is therefore the regular reflex of pre-Sibe [ʊ] (Kim et al. 2008: 52, 74, 107):
WMa akū [aqʰʊ] : S ako [aqo] 'not exist' (is the lack of aspiration a typo?)
WMa gūsin [qʊsin] : S gošun [koɻun] 'thirty'
WMa hūcin [χʊtɕʰin] : S xocin [χotɕʰin] 'well'
I am not entirely certain, as I have not yet investigated all Sibe cognates of Written Manchu ū [ʊ], and I found one exception (Kim et al. 2008: 79):
WM jakūn [tʂaqʰʊn] : S ǰakun [tʂaqʰun] 'eight' (instead of *jakon [tʂaqʰon])
I initially thought
WM juwen [tʂuwən] 'loan' : S ǰomjə [tʂomjə] 'to borrow' (instead of *ǰumjə [tʂumjə]; Kim et al. 2008: 80)
was an exception, but its o is a fusion of [uwə] also found in Sibe ǰwə [tʂwə] ~ [tʂo] 'two' (Kim et al. 2008: 81) corresponding to Written Manchu juwe [tʂuwə] and Ming Jurchen juwe [tʂuwə] transcribed as 拙 [tʂwə] (Kane 1989: 362, #1110).