"Sagispeak"

[volleo6144]

I'm not aware of any evidence that /kw/ is a difficult consonant combination for the average human. Though I agree that if it collapsed to /k/ we'd have a problem. And we'd also have a problem if some speakers collapsed it to /w/ since that's the language-permitting alternative to /dʒɐt/ ("jat").

Spanish and French use <qu> to denote just /k/ for some reason (<quinientos> /kinientos/ 'five hundred' in Spanish, <cinquante> /sãkãt/ 'fifty' in French), which might cause people to collapse it to /k/ if they don't realize.

I needed to verify that it was pretty common cross-linguistically, and I found this https://phoible.org/parameters which appears to be a ranking of all phonemes' popularity across all languages, which is pretty darn awesome if that's the case!

Yup! I've also seen it used in this series reviewing conlangs to verify that a preaspirated bilabial nasal /ʰm/ occurs in "approximately no languages".

...the initial "ng" problem...

Note that /ŋ/ is also only allowed to appear in the coda of a syllable in Mandarin Chinese (and yes, pinyin romanizes it as <ng>). Also Japanese phonotactics are pretty well-represented by the holes in the kana syllabary: there's a grand total of one consonant that's allowed to appear in the coda of a syllable, and the only consonant clusters that appear are /tʃ/ and /ts/, which are analyzed as allophones of /t/ except in loanwords.

[Dave Keenan]

Thanks for that review. I've added the following note to the Sagispeak table for Magrathean on the previous page.

qu = /kw/

Even if qu is collapsed to /k/, this will only be a problem for the two rare symbols (out of thousands) that consist of a 1-tina or 2-tina accent against a bare shaft (and of course their inverses). That's because /k/ is the sound of a right flag and can only occur as the onset of the final syllable, kai or kao, of a Sagispeak name, but any other use of tinas would follow a mispronounced ki or ko with another syllable.

[yahya.abdal-aziz]

Dave, is that magnificent table "Sagittal-SMuFL-Map.pdf" also available in other formats? I'm thinking particularly of Microsoft Excel spreadsheet or Word document formats, or their open source software equivalents.

I'd like to be able to print it out as a reference (ideally as an A3 wall chart), and those formats make it easy to have repeating heading rows and paginate, with the text at any desired - and readable! - size, using standard page sizes such as A4, A3 or US Letter. Not to mention tarting it up with a bit of colour!

[Dave Keenan]

Dave, is that magnificent table "Sagittal-SMuFL-Map.pdf" also available in other formats? I'm thinking particularly of Microsoft Excel spreadsheet or Word document formats, or their open source software equivalents.

I'd like to be able to print it out as a reference (ideally as an A3 wall chart), and those formats make it easy to have repeating heading rows and paginate, with the text at any desired - and readable! - size, using standard page sizes such as A4, A3 or US Letter. Not to mention tarting it up with a bit of colour!

Hi Yahya.

Douglas (@cmloegcmluin ) created it. He will be pleased at your characterisation of it, and your planned posterisation. It is magnificent isn't it.

Yes indeed. If you visit https://sagittal.org/ and scroll down to the fourth "shell point", you will find it in Excel and Libre Office spreadsheet formats.

[yahya.abdal-aziz]

Thanks Dave, and Douglas.

1. I've downloaded and installed the Bravura fonts and downloaded the spreadsheets, in both formats.
2. As I only have an old version of MS Office (2003) on this old Win7 desktop, I also have the compatibility pack, which automatically converts Excel 2007 format (.xlsx) to Excel 2003 format (.xls), and I note that using this old [fudge] the Olympian and Magrathean Sagittals display as (rather beautifully calligraphic) Chinese characters …! Not at all what displays in the PDF document. At a glance, everything else looks fine.
3. As soon as I can make time, I'll check out the Open Office format on a Win8.1 laptop and a Win10 desktop.

[yahya.abdal-aziz]

Just installed the latest LibreOffice 7.0.4.2 (x64) (with locale:en-AU; UI: en-GB) on my Windows 8.1 laptop, then loaded the .ods spreadsheet.

Unfortunately the Olympian (U+E3F4 - U+E3F7) and Magrathean (U+E3F8 - U+E40B) symbols still appear as (mostly) Chinese characters, with a few communications-related icons thrown in (pencil, scissors, envelope, postcard, analogue telephone).

I've checked that the font chosen for the first column is indeed Bravura Text. Indeed, I had to resize these problemaic symbols from 30 pt to 24 or even 22 for them to display fully inside the current column width. Why would just these be oversized, when all the other accidentals aren't? Strange!

Perhaps there's a problem with the font at certain sizes? But no, I tried several characters at various sizes, and they scaled as expected.

Any better ideas why it fails just here?

[Dave Keenan]

Hi @yahya.abdal-aziz ,

I'm sorry you had to go through that. But thanks for debugging our spreadsheets for us.

Please try the new version of the spreadsheets, and the new font to go with them, in the 4th shell point at https://sagittal.org.

The reason the previous versions didn't show the Olympians or Magratheans is that they were only accepted into the SMuFL symbol set (version 1.4) a few months back, but the W3C haven't yet released it, and so Steinberg hasn't yet updated Bravura and Bravura Text to include them. They are due any day.

In the meantime, I have uploaded a version of Bravura Text that we have modified to include the Olympians and Magratheans, and named Bravura Text SC, because any modified version must legally have a different name. And so I had to also update the spreadsheets to use Bravura Text SC.

I note that it took Windows 10 several minutes to install Bravura Text SC on my computer. But once installed it seems to be working fine. But I note that, in programs whose menus show font names set in their own font, Bravura Text SC will look like a black blob. But it will be in alphabetical order. This can't be avoided, due to other changes we've made in the font, for other purposes. Namely that the characters have zero "advance width" and so to even get them to show up in the spreadsheet, every symbol is now followed by an "em space" (U+2003).

I also made another update to the spreadsheets, that is not yet in the PDF. We recently decided to name the comma part of each symbol's pitch description using "directed names" with slashes, instead of the old "undirected names" that used colons (which can still be seen in the "SMuFL description" column on the far right). For example is now called "1/5-comma" while is called "5-comma". So one can immediately see which would be used for the harmonic and which for the subharmonic.

[yahya.abdal-aziz]

Thanks again, Dave! - for all the hard work, and your speedy response. It's always my pleasure to help, whenever I can, to test or debug any application or system so that it delivers a more satisfying user experience.

So I'll certainly download "the new, improved version" as soon as practicable.

Pity you have to fudge by adding the space after every character; surely a less tiresome work-around is possible?

I appreciate the user-friendliness of the "directed names", which will immediately inform users of the direction of the named comma.

[yahya.abdal-aziz]

Just downloaded the new Excel spreadsheet and the Bravura Text SC font, and find that the troublesome accidentals now display correctly on Windows 7. Interestingly, in the Excel formula bar, every one of the Bravura Text SC characters still displays as Chinese! Why they do, I don't know - yet.

[Dave Keenan]

Nicholas Denton Protsack and I have introduced a new flavour of Sagittal based on a set of symbols we call "Stoic" which has the same semantics as Athenian but only uses two types of Sagittal flag, the left barb (Sagispeak consonant "p") and the right scroll (Sagispeak consonant "n"). It uses up to 3 of each flag type.
See viewtopic.php?p=4449#p4449



This makes Sagispeak pronunciations challenging for the Stoics.

We have an existing rule where a doubled consonant is replaced by that consonant followed by an "h". This is currently only used for:
pp→ph or f
kk→kh or ch
ss→sh

But since "nh" is pronounced the same as "n" in English, and since we haven't used the "ng" phoneme, I suggest a new rule:
nn→ng (pronounced /ŋ/)

We have already used "g", but only as an alternative to "pat" , and there will never be a symbol spelled "npat-" since "n" is for a right flag.

But what do we do with triple-p's and triple-n's? I suggest we just translate the first two occurrences according to the above rules, and nothing more. So:
ppp→php
nnn→ngn

|(	nai	[existing]
|((	ngai
|(((	ngnai
/|	pai	[existing]
/|(	panai
/|((	pangai
/|(((	pangnai
//|	phai	[existing]
//|(	phanai
//|((	phangai
//|(((	phangnai
///|	phpai

Can anyone see any problem with those? Does anyone have alternative suggestions?

I'm hoping @cmloegmcluin can come up with IPA spellings for whatever we settle on — maybe with some alternatives for speakers for whom "ng" is a foreign phoneme, if we go with "ng". I expect "php" would be "fəp".