[tex-hyphen] Names of files in OFFO

[Claudio Beccari]

Arthur the two cases you are illustrating are not simple hypotheses, 
they are real cases.

When I was in junior and in high school in my country, where I had eight 
years of Latin, from 6th grade to 13th and I was from 10 to 18 years 
old,  I got a strong imprint about Latin; we started reading the 
classics, the easy ones, such as Cesar's De bello gallico, and our 
books, in spite of dealing with classical Latin used a totally modern 
orthography: distinct u and v, no ligatures æ and œ, phonetic 
hyphenation. The first patterns I conceived were actually a small 
variation of Italian, and actually I published at the beginning of the 
nineties an article on TUGboat dealing with one pattern set valid for 
two languages, Italian and Latin. After a couple of years I admitted 
that, even if it was feasible only for modern Latin and Italian, the 
idea was not a good one, and I detached Latin patterns from the Italian 
ones, and modified Latin patterns so as to include also the medieval 
ones; may be I was directed to this goal by my reading of the book De 
vulgari eloquentia, written in the XIII century by Dante Alighieri, in a 
period of time when he was actually setting the bases for modern 
Italian, but the scholarly works were still written in (medieval) Latin. 
Just to tell something that is little known outside Italy, in high 
school we read without difficulty the whole of Dante's Divine Comedy 
written in XIII century Italian, while Chauser's Canterbury Tales 
written in XIV century (middle) English are read today only translated 
in modern English because apparently anybody else from a scholar does 
not understand XIV century English.

It was the university latinists that complained about the inadequacy of 
my (modern/medieval) Latin patterns with classical Latin. I explained 
them that I did not create such patterns because I did not succeed in my 
search of scholarly books dealing with such a problem; I also remarked 
that hyphenation, as we consider it nowadays, started with "mechanical" 
typography, i.e. with Gutenberg. Before that time handwritten codices 
very often did not contain any line breaks obtained with word breaks, 
and  justification was obtained by enlarging or shrinking the scribe's 
handwriting while using special signs such as the tilde over certain 
vowels (librũ instead of librum) or barred stem q or p for other 
shorthands. When none of these tricks was usable the line breaks were 
obtained by breaking words without any syllabic concept. They finally 
gave me the photocopies of some pages of an out-of-print book where I 
was able to extract the rules to get the "correct" etymological 
hyphenation of classical Latin. Of course they were "correct" according 
to the author of that book.

Your second case is less likely, but not impossible; why would any one 
typeset a Neolatin text, spelt (and hyphenatable) with post Council 
Vatican II rules, with what is considered good classic spelling and 
hyphenating as in pre Council Vatican I times? It would be something 
such as wearing a Roman toga together with British modern leather shoes.

I understand that in Catholic monasteries they are very much concerned 
with correct hyphenation, which for them means dividing words with 
hyphens to match the musical score with lyrics when they have to sing in 
Latin.  But again this is not what we might call classical hyphenation 
(spelling is generally in the modern variety) because singing is not 
writing, nor speaking, nor reading. When you speak your phonetics are of 
certain kind and when you sing phonetics may be completely different; 
and when you read you don't speak nor sing; writing and hyphenating, 
therefore, must adapt themselves to the purpose of the text: prose or 
lyrics?

So let us establish this point:  phoneticians require special analyses 
to establish the meaning of "syllable"; nowadays  they perform time 
dependent Fourier analyses and trace the time dependent voice spectrum 
to decide what is a syllable. Grammarians establish rules that may 
divide syllables according to specific criteria that are completely 
conventional. Typography adds some more constraints depending on the 
purpose of the text. Neglecting the cases of u/v, ae/æ and oe/œ (that do 
not forbid to hyphenate liturgical Latin with the conventional 
phonetic/modern hyphenation) we have two conventional sets of rules that 
I call phonetic and etymological. With the two sets preloaded in the 
format files, and with suitable .ldf files for babel and polyglossia, 
you can deal with both cases the way you like.

If in OFFO it is not possible to distinguish these two conventional 
hyphenation systems I do not know what to suggest, because I had never 
been involved with hyper text mark up and I do not know anything about it.

All the best
Claudio







On 11/03/2016 17:34, Arthur Reutenauer wrote:
> On Fri, Mar 11, 2016 at 04:53:22PM +0100, Claudio Beccari wrote:
>> Yes, it is correct.
>    Good.  Now consider the following use case: someone wants to typeset
> speeches by Cato the Elder, using the classical spelling convention of
> course, but that user doesn't like the patterns you devised for
> Classical Latin, they want to use the other ones that produce phonetic
> breaks (they're great fans of Pope John XXIII and want to use the
> post-Vatican II conventions).  They also know how to switch between
> hyphenation patterns and are aware that there might be some minor issues
> due to the u/v distinction that the modern patterns make.
>
>    Conversely, let's assume someone else wants to typeset a collection of
> Papal bulls in modern Latin, making the u/v distinction, but they've
> made the stylistic choice to actually use the patterns that break
> according to etymology (they adhere strictly to Vatican I).  They're
> also aware of the underlying technical issues and are ready to solve
> them as they encounter them.
>
>    Obviously, these use cases are not perfectly supported by the current
> setup and users making the choices described in the above two paragraphs
> should keep in mind that they may run into some problems.  With that in
> mind, wouldn't you agree that these are nevertheless reasonable use
> cases?
>
> 	Best,
>
> 		Arthur