Question for Polish speakers: help with ligatures

[agisaak]

This topic was imported from the Typophile platform

Hello,

A font I'm working on contains an L_L ligature, and I've also opted to add Lslash_Lslash. I'm unclear, though, on whether I also require L_Lslash or Lslash_L. I was unable to find these pairs in a cursory examination of a few Polish newspapers, but since I don't speak Polish I don't want to take this as evidence of their nonoccurrence. Do any polish words contain the sequence LŁ?

thanks,
André

[Michel Boyer]

The open office Polish spelling dictionary does not contain the sequence lł (I searched with grep -i) in its 280963 lines.

[agisaak]

Thanks, Michel

André

[froo]

Indeed, there is no lł and łl sequence. The ll and łł are quiet rare, but worth of a ligature.

Look at this document by Adam Twardoch:
At the page 38 (37) you can find a good łł ligature.
On the page 34 (33) you can see a difference between printed and handwritten letters. (Note, that the someone forgot to add the slash to the uppercase Ł).
Page 35 (34): ł can look like T, but let the bar fly over the stem a tad.
Let's come back to the page 38 (37): if you want, you can lead the outgoing stroke of the second l through both letters, so you draw the whole ligature with one movement. Very romantic.

[Michel Boyer]

I get a few words with the łł sequence including mełli, półleżąc, dwuipółletni.

[Michel Boyer]

i.e. click listen Dowgiałło i Jagiełło półleżąc mełli dwuipółletni pieprz (the little loudspeaker on the Polish side)

[JanekZ]

[edit] see below
thanks Michel!

[Grzegorz Rolek]

Those with a pół sequence are compound words, with the pół itself being quite frequent and distinguishable stem in compounds, so I would be careful not to blend the ligature too much.

[froo]

Right!

Anyway, if this can help in something:
mełli, mełliby, mełlibyście, mełlibyśmy, mełliście, mełliśmy, mełł, mełła, mełłaby, mełłabym, mełłabyś, mełłam, mełłaś, mełłby, mełłbym, mełłbyś, mełłem, mełłeś, mełło, mełłoby, mełłobym, mełłobyś, mełłom, mełłoś, mełły, mełłyby, mełłybyście, mełłybyśmy, mełłyście, mełłyśmy, mełta, mełtą, mełte, mełtego, mełtej, mełtemu, mełto, mełty, mełtych, mełtym, mełtymi

[agisaak]

Thanks everyone for the feedback.

Just to clarify, the ligatures I'm dealing with are uppercase, not lowercase, and this is a sans display face rather than a script face. I'm basically nesting one L above the other in an attempt to reduce whitespace.

At least in English, the result isn't terribly satisfying when it occurs across morpheme breaks (e.g. in compounds). Do all of the Polish examples of ŁL given above involve morpheme boundaries? The face I'm working on has *lots* of interacting glyph variants, so the addition of any ligature actually involves creating numerous new glyphs. If these examples are restricted to morpheme breaks, I might opt not to include it. If they occur within a single root, then I'd want to include them.

André

[Grzegorz Rolek]

No, in the examples above only the łl as in the pół- morpheme crosses the boundary. Nevertheless, this particular morpheme is part of a few common compounds and could possibly be seen far more often than all the other words in question, so if you face the all-or-none dilemma, I would rather go for no ligature at all.

[froo]

Definetly. If you plan nesting letters, there should be no ligature here. The double ł appears in few foreign words like mułła (mullah) or the above mentioned Lithuanian names. The only exception are forms of the past tense and conditional mood of one verb, mleć (to grind), which seem to be a linguistic ballast, and aren't practically used in daily language. There, in plural, the łl appears. It appears to mark a difference between masculine and feminine forms: -łli -> m., -łły -> f. So nesting the second letter there, would be a bit misguiding. But in the case of the suffix pół (half, półlegalny = semi-legal) - where one probably can create much more compound words than those few "łl's" we are familiar with - crossing the boundary would be purely distructive.

[agisaak]

Hi Marcin,

I'm unsure if I'm correctly interpreting your post. As a non-Polish speaker, the łł sequences I've run across have all been in proper names. Are you saying that these are all borrowings and that the sequence is not normally found in Polish? If so I appreciate the info. It would not have occurred to me since AFAIK ł is only found in Polish and Navajo, so I'd assumed these were native Polish words, though on reflection I should have realised that borrowings often adopt nativized spellings.

From the comments here, it would appear that the ŁL/LŁ ligatures aren't needed, but now I'm wondering about whether I actually required the ŁŁ in the first place.

Thanks,
André

[hrant]

Is the "Ł" supposed to indicate a "deep" el sound? Because in the word for mullah for example it's not supposed to be deep; in Arabic only "allah" is accorded the deep el sound. I wonder if the bar is sometimes used to emphasize a word's foreignness...

hhp

[agisaak]

No, Polish ł is pronounced as [w] (and Navajo ł is pronounced as [ɬ], similar to Welsh ll). I'd assume the Arabic sound you're referring to is [ƚ], i.e. a velarized (or possibly pharyngealized) variant of l.

André

[hrant]

So the word for mullah is pronounced "muwwa"?

hhp

[agisaak]

Based on the spelling, I'd assume so. Bear in mind that l > w is fairly common historical shift.

André

[Synthview]

Hello,
if you ask for a ligature between two uppercase /Ł that I’ve never seen before and is very rare in Polish, can’t you just make the diagonal stroke longer?
It will cut the empty space without any ligature need, and it can exist in Polish. You can even make the stroke to start from the vertical stem.

I’ve read on typophile that a ligature should be avoided with compound words, and in Polish the sequence łl is compound only (I guess).
If you want to design a font adapted to Polish, focus on designing ć ń ó ś ź with the special Polish acute sign kreska ;)

@hrant : ł is a sound between u/w and l, but closer o u for not native speakers.

[hrant]

I’ve read on typophile that a ligature should be avoided with compound words

I think that makes great sense for display setting, but for text the opposite can actually be argued: you want compound words to be see as one thing - you don't want the reader tempted to wonder about etymology during immersive reading.

hhp

[jcrippen]

Polish ł used to be pronounced [ɫ] or [lᶭ], i.e. a velarized or ‘dark’ [l]. This sound is found in North American English dialects in words like ‘ball’ [bɑɫ] and ‘built’ [bɪɫt]. In Polish the lateral [l] part was gradually lost, leaving only the velarization as a voiced velar approximant [ɰ]. Labialization came along at some point, producing [w]. (The [w] sound is not produced with a single gesture, but is instead actually both a bilabial and a velar approximant, i.e. [ɰʷ].) Apparently the letter ł developed back when it was still a lateral sound and hence was paired with the letter l representing the ordinary voiced lateral approximant [l].

Because some Native American languages distinguish a voiced and voiceless lateral sound, the Polish ł letter was borrowed to indicate the voiceless lateral. Thus Athabaskan languages like Navajo use ł to represent their voiceless lateral fricative sound [ɬ], tł to represent the aspirated affricate [tɬʰ], and tłʼ to represent the ejective affricate [tɬʼ]. In handwriting the Native American ł is distinctly different from the Polish ł, even though both are similar in print. Handwritten ł in Native American orthographies usually appears as a looped l with a crossbar, similar to a cursive t. In such hands people tend to write cursive t without a loop. In non-cursive hands the ł is often still written with a loop, but unconnected to neighbouring letters. Some people write it more like the IPA ‘belted l’ symbol ɬ instead, with the crossbar connected by a loop to the upright; this style of handwritten ł is where the IPA symbol came from. Finally, it can also be written in unconnected hands with a simple vertical stroke like l but with a slanted crossbar, very similar to the usual sans-serif appearance of ł. I would attach some examples but I don’t have a scanner handy. Maybe I’ll post some tomorrow.

In Native American languages it’s not impossible to encounter sequences of two łs in a row, but it isn’t exactly a common thing. Representing them as a ligature would be confusing to most people, however. The only thing that might be advisable to do would be to ensure that the two crossbars don’t overlap.

[froo]

Thank you!
I would add, that the 'dark' l [ɫ] was a part of the formal spoken language (politics, movie, theater) until 60's.

@agisaak:
Are you saying that these are all borrowings and that the sequence is not normally found in Polish?
Three of them: mułła - Arabic, Dowgiałło and Jagiełło - Lithuanian. The other two I've found, ancient verbs mleć and pleć (to grind, to weed) that have the łł/łl in conjugation, are Polish, but extremally rare in literature (and in the daily life - swapped as too hard to pronounce).