Is there a replacement for the USFM marker \ul

Question

We have a project that we are trying to get uploaded into the DBL but our DBL currator says the project’s being rejected because of schema errors relating to \nd and \ul. She says that these two are deprecated.

I do not see \ul in the USFM 2.4 documentation but I do see \nd. Is there a marker in USFM 2.4 that does underlining like \ul?

Thank you

Answer 1

I agree with most details in the two above contributions by @CrazyRocky and @PADev . Still I am seeing a language, where a substantial list of homonyms which are not homophones are not easily understood by the native speakers, because the written representation is not giving them the full information they would get from oral.

This is about short verbs, where diacritics are already in use to show some TAM-information and grammar, so not easily fixed. And the team has got a nice collection of real-life or real-journalism examples where this old myth “the context will always tell” is just not true.

Not true even for the area of nouns, so the population has decided a few years ago from a real need, to mark minimal pairs for nouns and it has proven very helpful for the authors/journalists and for the readers.

Now underlining is indeed tricky and full of technical pitfalls. Luckily there are Unicode diacritics which just look like underlines and which should be treated by most systems like any other ugly or not (combining) diacritics. I am thinking of stuff like U+0332 COMBINING LOW LINE : underline, underscore.

This would be closer to PADev’s advice to “use something in Unicode” and CrazyRocky’s who has also mentioned diacritics. It would just look line underlining. And luckily most websites and apps do no longer use underlining (and bright blue) for links, so in a few years this feature might be liberated for local use. And yes, it can and should be taught locally like any other part of the orthography.

I am not recommending to use this idea without a real study of pros and cons. But I have also learnt that most languages have quirks and “weird details” and that coming in as an outside org to “help and fix” stuff is not always welcome and might also damage the identity or charm of a given languge. Just try to tell the Francophone to get rid of their cedilles for something more mainstream…

Personally I would not enjoy typing or processing entire words which would use the combinging-diacritic-technology on each character to distinguish words from each other. But I believe PT8 has got enough tools (like autocorrect.txt) to potentially handle exotic stuff like this and even publishing in print or modern media should be possible.

So the OP has asked about how to do this potentially. And a forum like this can give input that there might be enough reasons for the owners of that orthography to consider changing it. But that is always emotional and such input should be given with love and with much listening. I would love to hear more from @MSEAIT/LT if and how this detail was handled finally and how it is being received.

Sometimes I see contributions in other fora like “you should not do this, you should not even want this” and often I have a feeling that it means “… because I / we / our_beloved_tool cannot do it”. This is why I added a few words that I believe PT8 can do it, if those concerned want to go ahead this way. fwiw

Comments

Good morning,
As the underscore was used in a previous printing, there was no consideration of changing it. The local community understands it’s use. I made the original post in the context of getting it submitted into the DBL…you can imagine that there is zero interest in changing an accepted orthographic practice so that an already published (in print form) project can be submitted to the DBL.
MSEAIT/LT

Answer 2

I hate it when people tell me in a forum: “You do not want this, you want this other solution (which I happen to know about)…”.

I also hate it when people tell me in a forum: “You should not want this…” And often all they mean is “I do not know any answer.”

But then again, I find it rather helpful in a forum, where I need to ask something exotic or unusual or normally-frowned-upon, when I or other users explain why they want this. And what purpose it should serve.

I looked it up, and I did not find a marker for underline, sorry. I also found a red warning again, that it is strongly discouraged to use such “direct formatting”. And normally I might smile at such warnings.

But in the special case of underlining I need to caution at least: Underlining is really problematic, typographically speaking. You will find it much less in professional contemporary publications, and mildly-young people just think “hyperlink” and try to click on anything underlined.

So if you would share what you are trying to do, this forum might find a plan B with you.

I might get yelled at, but you could also look up the \z namespace. It allows you to create custom markers. You should only even go there, if you feel that the team using such has a full understanding on “all things USFM” and that there is good (and open-minded) support. So it depends on local competence and on the importance (or not) of what you are trying to do.

Imaging you have indeed a special need. You mark your special stuff \zu my special text-bit\zu* and later you convince or bribe your friendly typesetter or app-maker to style those as underlined. It is not hard to do, but can be tricky for your visual results, for example if your line-spacing is not really ready to hold underlines.

Answer 3

\nd is not deprecated. However, what you want to be careful of whether \nd ...\nd* is nested properly in some cases where it is being used. For example, within footnotes (inside \ft), it needs to be \+nd ...\+nd*

USFM does not officially include an \ul ...\ul* marker. Uploads to DBL need be be USFM (and therefore USX) compliant. You can use a feature provided for uploads to DBL which can allow you to apply a st of changes to a text prior to conversion to USX and upload. The changes do not apply to the project itself, but just to the converted data. Changes are listed in a file called dblChanges.txt (saved as UTF-8 in the current project folder).

"search for" > "change to"

Regular expressions are supported.

If your text has \ul ...\ul*, you could convert this to a valid USFM markup using this changes method, without changing your project text. There is not an ‘underline’ marker in USFM. You could (perhaps) convert these to \em ...\em* (emphasis), if that’s suitable. You would need to instruct a publisher that your intent for \em ...\em* is underline.

jmkla

Comments

I was trying to refer someone to DBLChanges.txt, but PT8/PT9 Help makes no mention of it. Seems to be a secret feature. It would be good if its existence as you’ve explained it here could be added into the Help system.

Answer 4

Dear MSEAIT/LT,

\nd is not deprecated. Regarding \ul and other formatting, you need to remember that the Paratext usfm’s are meant to identify the text, not actually define how it will look in the print job. How it looks is always secondary. So, with \nd, we are identifying the text as a “Name of Deity”. Once the text has been identified, we can define how it should look when publishing, and we have lots of choices. With \ul, the text is not being identified, and so it is not a good sfm marker for Scripture. Think instead, why am I underlining this text? What is it? Then use a marker that fits that identity, perhaps \k is what you want. At publishing time, you will then be able to define \k words as underlined. You should also make sure that you don’t misuse a marker just for the sake of how you want the text to look.

Custom.sty is the tool we use when we want to change how a particular usfm marker looks in Paratext or when we cannot find a marker that properly identifies the text. The Paratext Help menu has lots of good information on custom.sty.

Going back to \nd, as with all character formatting, you need to be careful to nest it if you are wanting to use two character styles at the same time. E.g., like using \nd in a footnote which is already a character style. Nesting uses the + sign for the inner set of character formatting so you would use \+nd …\+nd* when you want that character style inside a footnote.

Blessings,

Shegnada James

Language Technology and Publishing Coordinator, SIL Nigeria

Text Processing Specialist – Complex Script, GPS, SIL Intl

Skype: Shegnada.james.

[Email Removed]

+1 972 974 8146

Admin edit: Fixed formatting from e-mail.

Answer 5

Thank you all for replying.
This particular project was done quite some time ago and by a partnering organization. The team decided that the best way to resolve an orthography issue where two words are spelled the same, pronounced the same, but have different meanings can be differentiated by underlining one of the words. The team used \ul…\ul* to handle the underscore.

I’ve been given permission from the Partnering Org to change out \ul for \k. We’ll adjust the custom.sty to have \k underline the spanned text.
We can also edit a style in Scripture App Builder to have the app underline the spanned text.
We’ll have to figure out what’s up with the \nd issue. It maybe that some are nested but didn’t have the + applied.
Thank you again.

Comments

Thank you for giving the reason which prompted your underline. Local orthography rules always should have priority and using underlines to distinguish minimal pairs is rather clever, depending on frequency and other typographical considerations.

Hope it will work out all the way to the printed page or the app-screen. And glad I learnt about \k which had never made it unto my radar so far. Hope also this solution will not conflict with any “real” keyterms in your project.

Answer 6

The concern that homonyms will be confused if they are not typographically distinguished seems overblown. Most languages have these and they are not a problem in speaking or in writing because context allows native speakers to always know the correct meaning.

Using a typographical feature such as underlining or italics is not likely to work well, unless this convention is already being taught in schools and has been widely adopted by the community. I would definitely try and verify if this is something that the language community, not the publisher, wants. If not, I would recommend that the publisher drop it.

If a distinction does need to be made, most languages do this with diacritics or alternate spellings.

Comments

I agree with CrazyRocky: using
underlining in an orthography is unconventional enough to lead
to all kinds of problems. For example, standard spelling tools
don’t pay any attention to underlining. On the Internet
underlining is used to represent links. Texting systems usually
don’t support underlining, or if they do, it will be too much of
a nuisance for people to use. Keyboarding solutions support
pretty well everything in Unicode, but not usually underlining.
If people don’t use this feature in their everyday
communication, it will just become an artifact of Scripture.

If contrasting spelling is
genuinely important, use something in Unicode to do it properly.

PADev.

Answer 7

A question and a comment:
Is the underlining at the word level or character level?

• If underlining is at the character level, then any of the following could be used
  ◌̠ ‎0320 COMBINING MINUS SIGN BELOW
  ◌̱ ‎0331 COMBINING MACRON BELOW
  ◌̲ ‎0332 COMBINING LOW LINE
• … for two letters that need to be underlined
  ◌͟◌ ‎035F COMBINING DOUBLE MACRON BELOW
  I tend to favor the macron, since it is a proper diacritical mark

---

• If underlining is at the word level, then you can co-opt \pn …\pn* markup, which is intended for Chinese to mark Proper Nouns spelled phonetically using characters: eg: 以赛 Isaiah. This is a typographical feature that is also used for spelling disambiguation, not unlike the case you are describing.