What is your favorite letter?

zeno333's picture

In the video here...

http://www.youtube.com/watch?v=F691weEVpwc&feature=related

....Erik Spiekermann says his favorite letter is the "small a"...I totally agree....It is the most complex and has the most potential for variation as he says....That is one of the reasons I have always been fascinated with the "Galliard" typeface, it is revolutionary since it dares to "throw away" much of the potential variation of the small letter a and replace that with just a uniform width straight line!!!....INGENIOUS I say.....

Té Rowan's picture

It sounds something like a 'kh', then?

John Hudson's picture

It sounds something like a 'kh', then?

What does that mean? In whose orthography? That's my point: unless you have a description of how a sound is made by the vocal system, you're not really describing anything unambiguously meaningful. For instance, if I see 'kh' I presume an aspirated velar stop, because that is what 'kh' most often indicates in English orthography. But the same letter sequence might indicate something quite different to someone else (including, possibly, a pharyngeal fricative). This is why the IPA alphabet exists, to provide symbols that correspond to the specific descriptions of how sounds are made, rather than the ambiguous mappings of conventional alphabets.

Té Rowan's picture

The 'kh' I was thinking of is probably a voiceless velar fricative, if the 'pedia is to go by.

hrant's picture

You can't clear phlegm with a "kh".

hhp

John Hudson's picture

The 'kh' I was thinking of is probably a voiceless velar fricative, if the 'pedia is to go by.

Ah. So similar to the Arabic ح in that it involves constricting airflow between the tongue and another surface, but further forward in the mouth. I suspect a lot of English speakers trying to learn Arabic begin by trying to fake the ح sound with a velar fricative, because they're not used to making sounds so far back, in the throat.

hrant's picture

And just wait till you try the ق. :-)

hhp

dezcom's picture

Sounds like visiting my uncle Stathis at the assisted living center.

Typogruffer's picture

My Latin favorite is lowercase g
My non-latin favorite is sheen ش

Té Rowan's picture

Pharyngeal... reminds me of a time when an Icelandic TV reporter got an impromptu lesson in pronouncing Qaqortoq. The 'q' sounded like it came from deep in the throat.

quadibloc's picture

I thought that the Armenian alphabet had a simple and uncontroversial origin, like the Glagolitic, from some obscure form of cursive Greek writing.

In seeking the materials from which Mesrop Mashtots worked, so that I might suggest how a voiceless pharyngaeal fricative would have been written in Armenian, though, I have found that it was not so simple.

Gardthausen favored Greek, but this was already deprecated by Taylor, who favored the advanced form of Zend found in the inscriptions at Kapur di Giri.

And there is the recent theory, mentioned in this forum, that there is more than a mere superficial resemblance between Ge'ez and Armenian.

So I thought I would check on Glagolitic; surely the origin of that from cursive Greek is uncontroversial. And I found a paper suggesting that Cyril and Methodius would have been rather too late to have invented that script... because it derived from Linear A. Somewhere around Greece still, but...

However, Isaac Taylor at least also exhibits the presumed cursive Greek antecedents of the Glagolitic letters.

And while Mhedruli doesn't look much like Armenian, Hutsuri does... but am I surprised to see Georgian nationalists claiming Mesrop had nothing to do with the origin of their script? (Taylor feels he used a better version of Zend for the Georgians because with them he was working from a clean slate... the Armenians already using an inferior version of Zend to write with, Mesrop was constrained.)

...the Armazi script is proposed as one of the more likely alternative sources of Georgian; but I've now stumbled on one source that says that it, rather than Zend, was what Mesrop used, in addition to Greek uncials, to create the Armenian alphabet!

At least with so many candidates, I should have no problem finding enough shapes with which to extend the Armenian alphabet so that it can be used to write Sanskrit.

And, of course, let us not forget the Albanians of the Caucasus (no relation to the Albanians of the Balkans) and the alphabet Mesrop gave them.

hrant's picture

It's not clear exactly what Mesrop used as inspiration; some scholars even claim his design was totally original* and it was only the scribes later on that made our script look more like others (because they were used to copying things in Greek, etc.). In fact it's entirely possible that Mesrop might have have yelled "What the Hell are you doing?!" at those braindead copyists.

* Which is highly plausible because the whole reason to invent the Armenian alphabet was in fact to fight assimilation, and copying other scripts goes directly against that.

am I surprised to see Georgian nationalists claiming Mesrop had nothing to do with the origin of their script?

Can you spot the key word in that phrase? :-) Most Georgians had no problem accepting that an Armenian had designed their writing system... until we started demanding more respect for the Javakhk region!

I should have no problem finding enough shapes

But frankly why even worry so much about historic precedents? To me they can only hold us back; the best shape for that sound would come from an analysis of what readers need today.

hhp

altsan's picture

Going back to the original question, I think I like 'R' and 'e'... partly because I find there's so much scope for having fun in designing them; also, you can tell so much about a font just by looking at those two characters...

Moving away from Latin, I'm also fond of ね. (Perhaps ゑ even more so, but that one isn't used in modern Japanese so I'll relegate it to an honourable mention.)

quadibloc's picture

@hrant:
To me they can only hold us back; the best shape for that sound would come from an analysis of what readers need today.

Hmm. My thinking was simply that starting from a historical precedent, even if it came from a theory which would not withstand critical scrutiny, was an improvement on just making a shape up.

Your point is definitely valid; my first reaction was to think that this would simply be an after-the-fact constraint to ensure any new shape used is legible and distinctive. On further thought, however, there's also the issue of the learning curve.

Instead of digging into historical precedent to find out what Mesrop might conceivably have done, which did make sense for the original intended application as I understood it, if one was talking about adding new letters to Armenian for widespread adoption (which, incidentally, I've learned has already happened twice since Mesrop) the proper example to follow would be that of the Initial Teaching Alphabet, so as to avoid a steep learning curve.

That is, any new letters ought to resemble, although being differentiated from, existing letters related in sound, so that it would be immediately obvious what they were for.

timd's picture

I always enjoyed an e at the end of a line when calligraphing, crossbar flourish. Or an uppercase G at the beginning of a line.

Tim

hrant's picture

has already happened twice since Mesrop

Yes, the Օ and Ֆ... both of which are out of character with the rest of the alphabet (with the former also being a sad case of script-borrowing). Mesrop would never have made those that way; they are products of multiple levels of ignorance.

any new letters ought to resemble, although being differentiated from, existing letters related in sound

This is a classic conundrum for anybody contemplating the invention of letter symbols. Learning and doing are not the same, and are often at what I call "90 degrees", meaning they pull in somewhat different directions. A symbol that's easy to "figure out" because it resembles the shape of a similar sound becomes harder on readability when efficiency gets factored in (think speed and parafoveal depth) once you no longer need to make the connection. So,which to favor? Or rather: where to strike the balance? Which BTW is the beauty and power of Hangul: you can easily learn the basic alphabetic components that join to make up the 2-4 letter syllable shapes, but once you know a compiled syllable shape you can read much faster without bothering with the letters. And this is exactly a power that comes from conscious design, unlike the headless-chicken evolution typical of most scripts.

For example, is it good that the words "opposite" and "apposite" have opposite meanings? :-)

hhp

John Hudson's picture

For example, is it good that the words "opposite" and "apposite" have opposite meanings?

This seems to be a distinct sort of question from that regarding letter shape. Almost all languages have root systems that link related words, even if the relationship is one of opposition as in this case. Sometimes, these root systems are reflected in pronunciation, often in spelling (even when pronunciation changes) and sometimes both.

quadibloc's picture

@hrant:
the Օ and Ֆ...

No, I counted that as once, although since it happened in the Middle Ages, the two letters could have been added on separate occasions.

The second time, at least according to Wikipedia, was և (a ligature for եւ, I suppose) and ՈՒ (or ու in lower case) which, although looking like a digraph, is counted as a new letter.

hrant's picture

Yes, my "apposite"/"opposite" parallel was a bit... metaphorical.

The և is supposed to only be used like an ampersand - it's not a letter. The ՈՒ/ու was a Soviet joke at our expense; it's not a single letter for about half of the world's Armenian population. Note BTW that the people who regard ՈՒ/ու as a letter no longer regard Ւ/ւ as such (which is part of the same joke).

hhp

quadibloc's picture

@hrant:
Note BTW that the people who regard ՈՒ/ու as a letter no longer regard Ւ/ւ as such (which is part of the same joke).

Hmm. Would this be like making QU/qu/Qu a single letter in English, so that we would no longer regard Q/q as a letter?

On the other hand, our Dutch friends do regard ij as a single letter.

hrant's picture

Basically no word starts with Ւ/ւ, and since the Soviet-era spelling reform (carried out in the 1930s) only has it after Ո/ո they decided to change the alphabet by replacing Ւ/ւ with ՈՒ/ու. BUT in the Mesropian orthography* Ւ/ւ is used much more. I'm a reform-minded person myself, but even more important is Intent, and it's no secret that the Soviet intent with the reform wasn't to help kids learn spelling faster, it was to further alienate the homeland and the diaspora. And sadly it worked, because even two decades after independence we're still arguing -sometimes bitterly**- about which spelling is Correct.

* Which in Armenia is referred to as "classical" to avoid offending Mashtots. :-/

** You should see the two comments -one from each side- submitted to that recent Armenotype piece*** that I couldn't approve for posting. :-/

*** http://armenotype.com/2012/08/the-armenian-language-dilemma/

If/when the IJ/ij is a single letter, it should look and act like a single letter, which the Dutch haven't succeeded in pulling off yet. The recent resolution of the UC eszet is a nice contrast to that. BTW: http://typophile.com/node/34111

hhp

quadibloc's picture

It's hard enough for Armenians in the diaspora to convince their children that continuing to use Armenian is worth the effort!

To feel that it's worth the effort to learn Eastern Armenian - so that there is no longer "two of everything" - and, incidentally, to adopt the unfortunate tamperings with the language you decry - when one speaks Western Armenian just makes it that much harder.

BrettR's picture

For me it has to be either the "lowercase k" or the "lowercase b"

Martin Silvertant's picture

I know this is an old post but I noticed some interesting statements I want to ask about. Perhaps some people will have the interest to respond.

That is one of the reasons I have always been fascinated with the "Galliard" typeface, it is revolutionary since it dares to "throw away" much of the potential variation of the small letter a and replace that with just a uniform width straight line

What is he referring to?

The Armenian lowercase հ. It has so many stories to tell, not least about the future.

Could you elaborate a bit? I'm curious why this is your favorite.

And upside-down, or in certain styles slightly rotated.

I don't think I've seen an upside-down asterisk. Why do you prefer this, and what does the rotation do?

I would say it sounds like the beginning of gradually trying to clear phlegm from your throat. :-/ [about the pronunciation of the Arabic ح]

Does it sound like Dutch by any chance? ;)

quadibloc's picture

I used to call it "bicameral", but Kent convinced me (via Bringhurst IIRC) that that's better reserved for describing writing systems with two cases (such as Latin and Armenian).

Using bicameral in contexts other than referring to legislative systems with a House of Commons or Representatives below, and a Senate above, risks confusion.

For example, are we talking about a "g" that hears voices?

Yes, that was a reference to The Origin of Consciousness in the Breakdown of the Bicameral Mind by Julian Jaynes.

Actually, though, Kent's advice makes sense - for discussions here, such as about the problems of Greek typography, it would be useful to have a quick, unambiguous term for writing systems with an upper and a lower case.

If we were to call the Latin script polymorphic, to take into account upper and lower case, small capitals, and italics, though, then we would run into another problem. Just about any script can be bolded for emphasis - i.e. there's no problem with having a bold version of Chinese characters. (And they certainly can be optically slanted as well, although that somewhat collides with Chinese typographical conventions. Also, Song typefaces could be considered analogous to Roman, with Kai typefaces as the equivalent of italic.)

Lower case is clearly more fundamental to the Latin script than the weight variations used for emphasis - and small capitals occupy a strange intermediate position.

Martin Silvertant's picture

i.e. there's no problem with having a bold version of Chinese characters. (And they certainly can be optically slanted as well, although that somewhat collides with Chinese typographical conventions.

Why does it collide with the Chinese typographical conventions? I never bothered to ask why some scripts seem less sophisticated when it comes to features and typographical capabilities. I can certainly see why the Chinese scripts don't have small-capitals, but I can barely imagine not having different ways to emphasize and distinguish and having wide ranges of weights.

By the way, I know the movable type system was first introduced in China but somehow I never realized how tedious printing must have been for such an elaborate logographic systems. Does anyone know how optical sizes were employed? Is this perhaps a reason why the typographic capabilities in such systems have never been expanded (as simply adding different pt sizes for the whole logographic system would have been an absolutely tedious job), or is there inherently no need for it?

quadibloc's picture

Why does it collide with the Chinese typographical conventions?

I was thinking about the practice, now generally in disuse, of writing Chinese first vertically and then from right to left, instead of first from left to right and then vertically.

In the traditional script orientation, italicization would not be sloping the characters in the direction of reading, which may be significant.

Martin Silvertant's picture

Amazing thread. Thanks for the link!

Fournier's picture

¶ My favorite glyph is the ampersand in its italic form. Versatile & lovely!

Martin Silvertant's picture

The ampersand is great, though I don't enjoy designing it.

I love a, g, e and B. Hmm, I only now realize they're all binocular-like, with 3 or more horizontals. I don't think that's a coincidence. I guess it's the amount of detail involved in these letters which I like.

Fournier's picture

¶ My favorite glyph is the ampersand in its italic form. Versatile & lovely!

Especially·from·Frederic·Goudy!

Martin Silvertant's picture

How awesome are these though?


Martin Silvertant's picture

And here's Goudy goodness:






Fournier's picture

¶ All your examples of ampersands show the ancient calligraphic flavor of this glyph.

¶ Thank you for your wide array of ampersands: a·nice·bunch·of·flowers.

Martin Silvertant's picture

I just found out the "Goudy" ampersands I showed are actually part of P22 Goudy Aries.

jimmydurant's picture

Many years ago an academic requested that I add a letter to one of my Armenian fonts; he was writing a paper about pre-Mashtots Armenian and claimed that we used to have a sound like the Arabic ح (which I'm not sure how to properly describe) that got lost when our alphabet was implemented. According to his research the word for "lion" for example (առիւծ, which sounds like "arrouydz") originally started with that sound instead of an "a" sound (which actually makes sense if you know what ح sounds like :-).

examples of danish letters check out here
http://www.laan-info.dk/
http://www.laanogsparpenge.dk/

hrant's picture

Martin, the Armenian lowercase հ is my "favorite" because in most designs (especially these days) it is a glyph that's fully Latinized, and the furthest from what it needs to be; this is partly because its authentic form is so intractable (which to me actually makes it valuable) and also because it's the same sound as the Latin "h". But as a result of the debasement it suffers, the հ is also the key to our typographic salvation. When enough of us start making it correctly, it will mean we're finally on the right track again.

The fact that it's the first letter of my name is a bonus. :-)

hhp

Thylacine's picture

I've always had a fondness for the eth and the thorn. Tossed out of the English alphabet like a couple of lost boys from a polygamous sect, they've hung on for years in Icelandic living a lonely and isolated existence. They might not be the prettiest letters or the most fun to draw, but I feel sorry for them and think they deserve a second chance.

Maybe I should also feel bad for all those old Cyrillic letters that were cast aside, but I don't. Most were ugly, deformed and deserving of a dignified burial.

quadibloc's picture

One problem is that the distinction between voiced and unvoiced th in English is no longer phonemic. So if English did feel a need to have a letter for "th", they would probably take the Greek theta instead, it being more well-known, since only one letter and not two are wanted.

The lower-case version from Greek, though, wouldn't fit with the Latin alphabet well, so they would likely take the lower case of the Cyrillic letter removed from Russian by the new orthography instead! (So that gives us Ѳѳ.) And they would probably take the sh letter Шш and the ch letter Чч from current Russian, and maybe also zh, Жж, since these are other sounds used by English which don't have a single letter.

Té Rowan's picture

As long as the Icelandic language exists, so will '[Ee]th' and '[Tt]horn'. Sure, the '[Dd]croat' can sub for the former – and has, on recent occasion (read: Facebook) – but they are not scraping landfills for a living up here in the Little White North.

hrant's picture

If we had a good king/queen, he/she could reintroduce the thorn and eth back into English where they belong. Cheapskate English printers using canned Dutch fonts ruined everything.

hhp

dezcom's picture

I knew we could count on you. Té ;-)

Thylacine's picture

Yes, thank you Té for keeping the eth and thorn safe and warm during their exile. We owe everyone in Iceland a debt of gratitude for providing gainful employment to these two castaways.

quadibloc's picture

I was wondering what on Earth a [Dd]croat was, but I see that Croatian has the letter Đ or Đđ (U+0110, U+0111) which (in upper case) does look a lot like an eth, Ðð (U+00D0, U+00F0), as opposed to a thorn, Þþ (U+00DE, U+00FE). In looking those up, though, I see there's another forgotten letter - wynn - Ƿƿ (U+01F7, U+01BF) for which we might long.

And not far from Wynn, there's an African D, Ɖ (U+0189) as well!

Ah, but Wynn simply represents the same sound as the letter W; like Thorn, it was borrowed from the runic alphabet. So, since we already have W, there's no demand for Wynn.

Still, that suggests that the rune Ing could be used as the basis for adding a letter to the Latin alphabet for the consonant cluster -ng at the ends of some words.

riccard0's picture

Still, that suggests that the rune Ing could be used as the basis for adding a letter to the Latin alphabet for the consonant cluster -ng at the ends of some words.

That spot would be reserved to Benjamin Franklin’s eng:
http://steve-lovelace.com/modern-spelling-reforms/

quadibloc's picture

In that case, one might also go to Pitman's Phonotypic Alphabet for Esh and Ezh instead of to Cyrillic...

riccard0's picture

since we already have W, there's no demand for Wynn.

Personally, I would welcome the substitution of |w| with a “proper” letter: it always seemed to me a bit of “frankenglyph”, or a very bad case of keming*. Ugly, uninspired, underdeveloped, and too wide. One exception is perhaps the cursive “asymmetrical w”: https://www.flickr.com/photos/52362783@N00/6129692536/

* http://www.ironicsans.com/2008/02/idea_a_new_typography_term.html

Let’s try:)
Personally, I ƿould ƿelcome the substitution of |w| ƿith a “proper” letter: it alƿays seemed to me a bit of “frankenglyph”, or a very bad case of keming. Ugly, uninspired, underdeveloped, and too ƿide.

Uhm… a bit too “p” to ƿork ƿell.

white055's picture

I'd go for all letters, every letter is important & also full of potential.

hrant's picture

Indeed. But you still choose only one to marry. Or at most four. :->

hhp

Syndicate content Syndicate content