I have been scrolling through the pronunciation guide I mentioned in a comment about broadcasters’ rendering of Russian and other polysyllabic names. I guess I was thinking it would be a good source to check on how names are anglicized, since it provides “foh-NEH-tihk” pronunciations, including STRESS and a key for deciphering the transcriptions. Let me say here that I don’t consider it a fully reliable source for data.

A closer look on my part shows that it might need its own Cliff notes, since there are a few problems with symbol consistency. Hence the confound – we can’t evaluate whether something is awkwardly borrowed into English phonology because it might just be badly transcribed in the list. I’ve included the whole key to the symbology at the end of this post, annotated with comments.

Overall, there seem to be three main issues (among others): unstressed and central vowels, the letter “u”, and the voicing contrast between [s] and [z]. I couldn’t help but try to make sense of it, but I’ve only managed to have something substantial to say about the unstressed vowels.

Unstressed vowels are transcribed as lowercase ‘ih’ or ‘uh’. However, “IH” is also used for stressed [ɪ] and “UH” is used for [ʌ], as well as the nucleus for syllabic /r/, as in Berg –> BUHRG. (also, unfortunately, for [ʊ] in some cases). Nevertheless, MSWord’s Search-and-Replace can be used case-sensitively, so I searched for lower case ih and uh, to see how often each appeared: 606 uh vs. 255 ih (out of 1230 names).

I was thinking that a lot of those instances of uh would be word-final or involved as some way of notating syllabic [r] or [l]. In fact, there were 89 word-final uh and zero word-final ih. I then did some more restrictive searches, to remove strings like “uhr” and “uhl”, and got a new count of the two digraphs: 367 uh vs. 248 ih.

An unexpected bonus of the list is that syllables are separated by spaces – although whether they are done so accurately is another question. Nevertheless, the spaces allow for searching for strings like “uh_”, which uncovers the number of non-word final uh in open syllables. So of the 367 uh, 268 were in word-medial open syllables and 99 were in closed syllables, while of the 248 ih, 89 were in open syllables and 159 were in closed syllables.

I won’t run the chi-square, but there seems to be a hint of order to this, whereby uh tends to be transcribed for schwa in open syllables and ih tends to be used in closed syllables.

As for the transcription of [s] and [z] … check out the symbol key:

S for soft C (as in cease)
Z for hard S (as in decrease)

There’s actually no way in this guide to transcribe [z]. This is a serious shortcoming…in fact, both “s” and “z” are used for either [s] or [z]: Kozlov and Baines are transcribed improperly with [s], while Messier and Bowness are both improperly transcribed with [z].

Anyway, the rest of the symbol key is repeated below:

AY for long A (as in mate)
A for short A (as in cat)
AI for nasal A (as in air)

Something like this is necessary since the name list includes French names with nasal vowels; e.g., Morin gets transcribed as moh RAI. However, “AI” shows up in some transcriptions that have no nasals, like Laperriere  luh PAIR ee air, as well as the key’s example word above. In fact, of 69 instances of AI, only 13 are not followed by R.

I also found it interesting that the mid-front is the only vowel borrowed with nasalization; Jean and Simon are both transcribed with nasal consonants.

AH for short A (as in father)
AW for broad A (as in talk)

OK, so officially the NHL has no [a/ɔ] merger. But how can “father” and “cat” both have “Short A”?

EE for long E (as in meat)
EH for short E (as in get)
UH for hollow E (as in “the” or French prefix “le”)

Maybe the French citation is a bad(ly labeled) example. Sounds like UH means schwa, so far.

AY for French long E with acute accent (as in Pathe)

Note the missing accent in the example. That’s how it looks in the pdf.

IH for middle E (as in pretty)

The phrase “Middle E” implies between long E ([i]) and short E ([ɛ]) – not inaccurate here, but still a bit imprecise.

EW for EW dipthong (as in few)

Seems to support the claim that this is a rising diphthong; rather than the glide being part of the onset.

IGH for long I (as in time)

Personally, for a non-IPA system, I thought this was pretty smart.

EE for French long I (as in machine)
IH for short I (as in pity)

Short I = Middle E.

OH for long O (as in mote, or “ough” as in though)
AH for short O (as in hot)
AW for broad O (as in fought)

So the [a/ɔ] merger is maintained at this end of the orthographic space.

OO for long double OO (as in fool or “ough” like through)
UH for short double OO (as in foot or “ouch” as in touch)

Ouch indeed! So far UH has been [ә], [ʊ], and [ʌ].

OW for OW dipthong (as in how, or “ough” as in plough)
EW for long U (as in mule)
OO for long U (as in rule)
U for middle U (as in put)

This would have been fine, except for the part about “short double OO” above.

UH for short U (as in shut or hurt)
K for hard C (as in cat)
S for soft C (as in cease)
SH for soft CH (as in machine)
CH for hard CH or TCH (as in catch)
Z for hard S (as in decrease)

See above for frustrated comments.

G for hard G (as in gang)
J for soft G (as in general)
ZH for soft J (as in French version of Joliet)

If you can’t contrast S and Z, why bother with ZH?

Seems to me that the moral of the story is, phoneticians and phonologists are valuable and under-used resources.