TYPE BATTLE: Week 9 (20 March)

Christian Robertson's picture

TYPE BATTLE: Week 9 (20 March)

Here’s your chance to stretch your type muscles on a weekly basis. Each game kicks off on Monday with a new challenge, and closes at midnight PST on Friday. Anyone may submit a design response to the challenge. You may enter as often as you wish. Post anytime. Critiques and comments are welcome throughout the game, from participants and spectators alike. Smack talk is encouraged.

Winner take all, no holds barred. May be the best designer win.

Link directly to this thread: http://www.typophile.com/battle09

Requirements:

- Size: 600 width x 400 height. 72 dpi
- Color: Black and white only
- Format: Please save your graphics as PNG
- Only respond with the characters posted in the challenge.

// THIS WEEK’S CHALLENGE:
Design three new characters for the latin alphabet representing the sounds 'ch' 'th' and 'sh'. We're looking for new characters, not ligatures. Show the characters in context.

2m1's picture

thanks

engelhardt's picture

Quoted from the Typophile Forum Posting Guidelines ...

Image technical requirements:
The best way to post an image is to use the “Insert Image” function below the “Comments” box. To utilize this function, you must have the Flash 8 player installed. If you have an image to accompany your post, please ensure that it conforms to the following guidelines: do not exceed 600 px wide, save images as an RGB jpg/gif/png (tips: gifs are usually crisper for black and white type renderings and not all browsers support png).

Go HERE to download the Flash 8 player.

pica pusher's picture

Here's a hack-and-slash job on Trajan...

duncan's picture

You'll have to tell me if these are too close to being ligatures. I felt like these new characters should relate to, or derive from the original characters if they are meant to be understood as the phoenetic combination. Does that make sense?

Jongseong's picture

I am just going to say at this point that it would be a good idea to distinguish between the voiced 'th' of 'this' and the unvoiced 'th' of 'thin'.

There are of course variants of the Latin alphabet (such as those used by Icelandic and Turkish) that include letters to represent those sounds, but I'm guessing we should stay away from those letters?

Jongseong's picture

Here's a sketch of my attempt at the characters (UC and lc) for 'sh'. I can't think of natural shapes for the other characters yet.

dtw's picture

Well, I was going to say that we don't need to design new Latin-alphabet characters for 'th', because we already have them; the thorn (þ) and edh (ð). Why avoid them? (By the way, I think we should reintroduce these; they'd be great for texting, wouldn't they? I'd much rather see 'the' rendered as 'ðe' than 'da'...)

Still, for 'sh' I thought I'd cross the integralesque curled long s ʃ (used in the IPA to represent 'sh'; hmmm, to see that you'll need a Unicode-happy browser) with the Pitman shorthand hook thing for sh, to give a sort of swung s.
And for 'ch', maybe adapt the Cyrillic ч, but give it a longer, right-pointing tail so as to avoid any potential confusion with u, y or µ...


______________________________________
Ever since I chose to block pop-ups, my toaster's stopped working.

Jongseong's picture

I thought the point of the exercise was to design new characters. There simply are too many options already available if we just want to adopt existing symbols.

For the voiceless 'th', we have the Greek theta (θ) as well as the thorn (þ). We could also drag out the appropriate symbols from ancient Hebrew or Arabic.

For the voiced 'th', in addition to the edh (ð), we have variants like 'đ', or we could use the Greek delta (δ) in reflection of the Modern Greek pronunciation. Again, ancient Hebrew or Arabic symbols could be used.

For 'sh', the symbol 'š' is used in some Central European languages, 'ş' in Turkish and other languages, and 'ș' in Romanian. The IPA and some African languages use esh (ʃ). Cyrillic has sha (ш), and Hebrew and Arabic have corresponding letters.

Similarly, for 'ch' we have 'č' and 'ç', 'ʧ' in the IPA, che (ч) in Cyrillic, and so on and so on.

So I've been trying to devise alternative extensions to the Latin alphabet that avoid those existing solutions. For example, my 'sh' comes from imagining: what if mediaeval scribes added a fancy swish to the long s to represent the 'sh' sound and that eventually evolved into its own character? I find it really difficult to come up with reasonable, logical, natural-to-write characters that already do not look like some existing symbol in the IPA, for example.

jselig's picture

Not to be rude about this, but using pre existing characters defeats the purpose of the exercise. Not to mention the close resemblance to other common characters where the sentence reads "I pink dese upsticks are too sort". Aside from the ð character the rest could simply be mistaken for one of the other 26 alphabet characters, depending on the stroke of the person if this were written out as opposed to typeset. IMO, that's were the proposed solution fails.

dtw's picture

Fair enuff... :-(

jselig's picture

:)

Mark Simonson's picture

Not to mention the close resemblance to other common characters where the sentence reads “I pink dese upsticks are too sort”. Aside from the ð character the rest could simply be mistaken for one of the other 26 alphabet characters, depending on the stroke of the person if this were written out as opposed to typeset.

I think this will be unavoidable. If you put any unfamiliar character shape in with familiar ones, it's natural to try to interpret it as a letter you already know. This works to your advantage when you are designing a logo.

Something else to consider is how the new letters would be written, either printed or longhand, or even in a calligraphic hand. I think you need to figure that out before playing with typographic forms.

So far, I think Brian's sh (Shellfish) works the best. I didn't have any trouble reading it. I think it's because it retains the ascenders we are used to seeing in the word. I'm not sure how it would work in handwriting, though, and I don't think the capital form looks like a capital. I though it was a lowercase variation until I read the explanation.

pica pusher's picture

As I see it, the challenge here is this: letterforms just don't follow phonetic logic; they are pure symbols in that they do not visually depict anything from the real world.

(I'm told that devanagari scripts actually do relate to the shapes the mouth makes when pronouncing their related phonemes, but I have gone over and over them without seeing any such relationship...)

My idea was to start with some majuscule ideas that obey the "rules" of the first carved Roman characters (mostly rectilinear, with a few curves, serifs to facilitate chiseling) and that don't easily get misread as other characters. From there, hand-writing the majuscules over and over (starting at the upper-left) should give a good idea of how such characters might have developed over time into minuscules... at least, that's my hope.

Jongseong's picture

That's a good and logical solution, except that as someone who deals with the IPA a lot, I can't force myself not to see the 'hooked h (ɦ)' symbol used for the voiced version of the 'h' sound. Curse you, IPA, for using nearly every simple variant of Latin glyphs as symbols.

claes's picture

Interesting to see Christian's result, because I was doodling earlier and I came up with the same hooked h design for "sh". I couldn't come up with any good ideas for any of the other sounds though so I gave up.

Jongseong's picture

I'm happier with the 'sh' and the voiced 'th' (as in 'The feather') than with the 'ch' and the unvoiced 'th'. I'm aware that the unvoiced 'th's look like '5's.

Anyway, here's the doodling that inspired the letters:

track and kern's picture

hrmm... i know i don't post much, but i was going to jump in on the type battle this week. i feel that my knowledge is not of the caliber that will allow me to do so. if anyone might point me in a direction that would perhaps give me so inspiration or more knowledge on these specific characters, it would be great. I checked the old wiki already. all in all though, good stuff so far for those who posted. maybe i will play next time.

pica pusher's picture

So the caps don't match Minion yet, and all the characters have some weight issues... Still, I'm satisfied that these new characters aren't easily misread as others; if I had learned them as a kid I could probably read this (?)

franzheidl's picture

i know it always appears a bit easy to jump in without contributing something myself, but that task is really an interesting one…

The alphabet, or, to be more precise, language and letters, as it's materializatipon/physis, as a collection of symbols, is so deeply implemented in our brains that it appears so hard hard to just invent something new. As Mark Simonson rightly states, when reading a text you're expecting to find letters that are familiar, you WILL try to interpret anything you don't know as something that you know. And if you go down the 'phonetic' alley, you'll (inevitably?) end up with something that at least seems to be a ligature*: If you want to make it "readable" at first sight, it has to resemble something that you know, if you invent something completely new, readers will have to learn it first. Remember, we can't read when we are born, we have to learn that system first, thus anything remotely new will have to be learned too – prior to the ability to read and interpret it as intended. And if you can live with that, almost anything is possible that fits in with the latin script system (or any other, for that matter)
but anyway, that is not meant as being demotivating, i'm just wondering if it can be done at all and what the task actually means. But the proposals above are surely highly interesting. keep em coming!

* for example, i, being german, did interpret the proposal of an hooked lc h as a ligature of the outdated long-s (see german dbls) and an ordinary h. And another one above made me think of the indefinite symbol instantly.

Christian Robertson's picture

Exercises like this are always going to be an uphill fight. In the competitive environment of symbols, established symbols have an obvious advantage. Not only does the individual reader have to rewire their brain to interpret the new symbol, an entire community has to do the same.

So for this exercise the designer has to consider three things: 1) How well will this thing work once the reader has been trained, 2) How hard is it for the reader adopt, 3) How strong are the readers motivations to adopt the new notation.

In this case, it's obvious that there is no motivation for english speakers to change the way they notate ch th and sh. However, the exercise forces the designer to step back from their training as readers and start to look at the letterforms more abstractly. The designer has to retrain his/her brain to read the new characters and evaluate their performance. There are other ways to get this more abstract perspective. One way is to study other writing systems. Another is to simply turn your native writing system upside down.

The designer also has to consider how users will learn the new characters. Will they be intuitive (based on the community's shared past experience) or will they force the user to learn something new?

An interesting recent example of a community adopting a new system of writing are the l337 h4xx0rs who invented a new notation for online chatter. While many of the character/word substitutions resemble traditional notation, much of it is impossible to decipher without some training. In this case, there was sufficient motivation for users to train themselves in the new notation (no one wants to be a n00b), and it was widely adopted.

track and kern's picture

@ christian: Could you post a link to some more info on the I337 h4xx0rs. I am familiar with some internet lingo from other message boards, and the whole number substitution deal is not that new to me. I always just thought that is was something obscure though, and not widely used. I guess I am still a internet meme n00b.

track and kern's picture

@ pica pusher: you seem to really be making the most of this type batter. I am have been really interested in the characters that you have developed. On the whole, my only comment would be that your character that represents the "th" sound, resembles

Not sure if that occurred to you, and its not a criticism either, just conjecture.

P.S.- I could not get that silly ASCII character to render any larger with the typical font attribute tags. Can anyone site how exactly I might do this, and yes, I did already check the formatting page. I not the HTML/XML guru, but I can get around for the most part.

typotheticals's picture

Just a thought that occured.

fontplayer's picture

Fwiw,
I played around with a rasterized t & h in this font and came up with this. It is close to a ligature, but it is the farthest I want to get from what is familiar at this time.
: )

pica pusher's picture

h4ck3r5 f1r57 5t4r73d u51n6 num3r41s 1n5734d 0f 13773r5... aw screw it

hackers first started using numerals instead of letters to avoid email watchdog programs which took their cues from search strings. Probably unnecessary example: if you wrote in your email "Here's how to hack into xyz government office" that email might be caught by a program looking for the string "hack into" but if you wrote "H3r3'5 h0w 2 h4x 1n70 xyz g0v3rnm3n7 0ff1c3" your email would slide right through.

track and kern's picture

this is my first ever try. i think that it resembles an "s" a little too much still. originally it was not cantered, but then i changed it back, as the font i based its letterform off of was an italic face. oh well. so much for a first try.

dishdesigner's picture

Gosh, hats off to all who've submitted so far.....lots of smart stuff and fun to read! So, here's my $0.04.

I agree with Jongseong that in addressing this problem you must consult existing diacritical marking systems and alternative alphabets like IPA. Jongseong also rightly points out that singular new glyphs for these consonant clusters would fail to encompass voiced and non-voiced pronunciation variants, and after all, the problem as stated asks us to represent the "sounds"....

But we may be losing sight of the fact that extended IPA and other sound-based alphabetic systems have never been adopted as working alternatives to the Latin alphabet for daily use in any country. Why? Because they're too complex. Too many characters. They are only useful and necessary for linguistic specialists, typographers, and singers.

The beauty of what the Greeks did in simplifying their alphabet was to make learning accessible to all people. Alphabets before had so many characters and intricacies that only priests, scribes, and the wealthy and well-educated could read. Reducing the number of characters was critical to the democratization of written communication.

Along with that reduction comes a fair amount of abstraction. You must now rely on context (and some learning) to infer the intended pronunciation of consonant clusters and other language elements.

But I disagree with 'pica pusher' that letterforms are so abstract and unconnected to phonetics that they don't represent anything in the real world. Take the letter 'v' for example, and its origins in the diagram for an ox plow, indicating that its oneiric function includes pushing, reworking, molding, incising, carving, etc.....

Many people have written about the mystical representative aspects of our Latin letterforms' evolutions. We should not forget that most of our letters did indeed evolve from the need to represent real objects, forces, phenomena and ideas (regardless of whether those connections are still relevant to the modern world).

Given all this, my answer is that it is simply not necessary to propose distinct new Latin letterforms for these three sounds beyond the existing IPA. First, they would never gain everyday usage anyway; and second, this would set a dangerous precedent in frivolously adding characters to the alphabet to represent sounds directly which begs the question: why not then have a symbol for EVERY possible sound and consonant combination? Then we're right back to an alphabet too complex to easily learn and down comes written language!

Instead, I propose what I think is more appropriate: merely a new diacritical marking that could indicate a consonant's coupling with "h".

I quickly analyzed which characters used diacritical variations to the basic Latin letterforms for sound-based representation in non-English languages so as to avoid duplication. I focused on diacritical "pairs": the caron / circumflex, the acute / grave, and the macron / breve.

My aha! moment came in noticing that of those couples, only the macron / breve are actually different marks instead of simple formal inversions or reflections of the same mark. The macron and breve are used to indicate long and short vowels respectively. So, I thought, why not have a mark that indicates "long consonants"?

I also noticed that inverting the breve captures very compactly the most distictive formal part of the additive "h", the upper part of the bowl. I am not sure of any significant language usage of the inverted breve outside of logic symbolism, poetry scansion marking, or music (the dotted 'fermata').....as for naming, I thought it appropriate to invert the Italian 'breve' to 'esteso': succinct or short to extended or long.

A couple notes:

1. I thought it odd that they asked for 'ch', 'sh', and 'th', but not the significant 'ph', so I included it here, even though it's not really needed to clarify a distinct sound: the simpler 'f' does that. But still, adding the 'h' modifies the 'p' consonant sound....

2. I left it at those four characters because after consulting the Oxford English dictionary (I have the searchable CD-ROM version, which totally rocks...), I felt there were no other significant added-h variants that produced new sounds. Other consonant examples with added h's that don't really affect pronunciation: lhasa-apso = "lasa", khaki = "kaki", jhula = "jula", rhodium = "rodium".......perhaps a nod to 'z' and Eastern European languages for "druzhina" and the like?.....it would be easily drawn.

Jongseong's picture

Hmm. I think we're looking for characters, not a new system of diacritical marks (although the distinction is not always clear-cut). Also, when we talk about the 'sounds' represented by writing systems, we usually mean phonemes, theoretical units distinguishing meaning in a language that actually encompass a range of different sounds. The 'p' sounds in 'post', 'spot', and 'stop' are all different sounds (and careful IPA transcriptions will distinguish those), but are all manifestations of a single underlying phoneme represented by the letter 'p'.

Theoretically, there should be a one-to-one correspondence between the phonemes and the symbols in an alphabetic system. In practice, there can be differing analyses of what the phonemes are in a language. Also, there are solutions meant to reduce the number of symbols necessary to represent the phoneme inventory of a language. The h-digraphs in English are an example, as is the practice in some languages of doubling letters to represent long vowels (if you recognize short and long vowels as separate phonemes).

I was making the distinction between the voiced and unvoiced versions of 'th' because they represent different phonemes in English, as evidenced by the minimal pair 'loath' and 'loathe'. The only difference in the pronunciation of those words is whether the sound represented by 'th' is voiced or unvoiced, so we know this is a meaningful distinction in the English language. Of course, it is possible to analyze the situation differently and say that there's a 'silent e' phoneme at the end of 'loathe' whose effect is to make the 'th' voiced, so it is a matter of opinion whether the two 'th's are separate phonemes or not.

DiSH, you might be surprised to learn that several African languages have actually adopted IPA symbols for their native phonemes. You say IPA has too many characters, but that's because it's meant to transcribe the countless sounds used in any natural language. Different languages have different ways of organizing sounds as phonemes, so the IPA has to be provide for all those possibilities. But the writing system of a given language only has to take the native phoneme inventory into account, so it will be much, much simpler. A given language will typically have between twenty to fifty phonemes, a few dozen tops. That's why most European languages have done well adopting the twenty-odd symbols of the Latin or Cyrillic alphabets. With the judicious introduction of digraphs, trigraphs, diacritical marks or ligatures, all the phonemes of a language can be written with the existing symbols.

If a new alphabet for English were to be created from scratch from IPA symbols on the one-symbol-per-phoneme basis, it would be entirely manageable. Making new symbols for 'th', 'sh', and 'ch' makes sense, because the sounds they represent count as separate phonemes in most analyses. (They are not 'long consonants'; English doesn't make long consonant distinctions, and the 'th' sound is certainly not a long version of 't', etc.) A new symbol for 'ph' wouldn't make sense, because it's just a historical variation of 'f' for words of Greek origin based on Latin spelling (notice how Spanish for example replaces that digraph with 'f'). No-one would seriously advocate new symbols for every consonant combination, because most consonant clusters in English would be analysed as combination of multiple phonemes. So applying the one-symbol-per-phoneme scheme would not lead to 'an alphabet too complex to easily learn'. It would certainly not be more complex and difficult to learn as the current system of English spelling.

I hope I explained clearly enough why this is not an arbitrary exercise but one that makes sense (as opposed to, say, designing characters for 'st', 'rs', and 'mn'). There's no reason for English not to have a separate symbol or two for 'th' other than that Latin (for which the alphabet was devised) didn't have the sound and lacked a dedicated symbol for it. If Latin didn't have the 'f' sound, the letter wouldn't have existed and we probably would have to do with something like the 'ph' digraph for the sound in English as well.

pica pusher's picture

Apologies; I didn't mean to suggest that written language has no iconographic basis - just that it naturally develops away from visual representation. Imagine, for example, V was actually a detailed silouhette of an ox-plow. Apart from the readability issues, the symbol would actually become less meaningful because, seeing it, one couldn't help but think "ox plow" and not "V." Note, too, that there is absolutely no relationship between V and F, although both are nearly identical phonetically (labiodental fricatives). This strong visual difference between phonetically similar characters is key to alphabets' power, because it makes reading similar-sounding words easier. As I see it the same logic can be applied to these new characters: if the new character foo looks like an SH, a reader will think "SH" and not "foo."

dishdesigner's picture

Jongseong and pica pusher:

Delightful! Your insights are inspiring. What complexity lies beneath such a simply stated question...

Jongseong: Thanks for the IPA lesson. The Wiki articles for "H", "digraphs", and trigraphs were all illuminating after reading your response.

But does your statement, "with the judicious introduction of digraphs, trigraphs, diacritical marks or ligatures, all the phonemes of a language can be written with the existing [Latin] symbols," support my point that a diacritical solution may be enough here? Do new characters for 'sh', 'ch', and 'th' save us enough work to justify the imposed learning curve? Or, is it your position that the learning curve would actually be lower in the long run since people would no longer have to assimilate digraph/trigraph/diacritical "band-aid" solutions and thus spelling complexities?

Pica pusher: a nice point about the abstraction of a letterform reducing its specificity (and therefore increasing its general symbolic power) and its need for visual difference, even from phonetically similar characters......

OK - 12 hours left...I'll give new characters a go tonight!

dishdesigner's picture

By the way, Jongeong: can you post the careful IPA transcriptions you mentioned of the different ‘p’ sounds in ‘post’, ‘spot’, and ‘stop’? I'd love to see them.....

fontplayer's picture

Pasted together from Aichel

3d5's picture

Hi, I'm still a noob here, but hey I can give it a try :)

Jongseong's picture

DiSH, a diacritical solution certainly would not only be enough, but quite logical and natural. In fact, I think several shorthand notations already replace h-digraphs with diacritical marks. But such a solution just wouldn't be as interesting and fun as designing whole new characters.

IPA transcriptions can be as detailed as the amount of phonetic information you want to include. I've seen some examples of 'narrow' transcriptions and the details included there are phenomenal. I'm not trained for such detail, but since you ask I'll just give transcriptions that are just enough to distinguish between the phonations of 'p' in 'post', 'spot', and 'stop' (I hope these display on your browsers). The 'p' in 'post' is the aspirated [pʰ]. When it occurs in the combination 'sp' as in 'spot', it is unaspirated, so it is most often written simply as [p], although the Extended IPA probably provides for detailed transcriptions of the state of the glottis if one gets to that. The pronunciation of 'p' in 'stop' varies tremendously according to speaker and whether one is pronouncing carefully. An unreleased [p̚] is most frequently heard, but I guess it could be [pʰ] in careful pronunciation (i.e., you would hear the 'puff' of 'p' at the end). It could be reinforced with a glottal stop, especially in British English: [ʔp]. And in general, you can probably add a mark below ([p̺]) to indicate that the English 'p' is apical in terms of articulation.

dishdesigner's picture

OK.........very late.....last try!

dishdesigner's picture

BTW, thanks again Jongseong for your eye-opening IPA knowledge.

I downloaded the IPA alphabet and was surprised to see that many diacritical marks have direct relationships to the character of voicing implied (like a "central" articulation or a "breathy" phonation)! So, I see how I was playing fast and loose with the idea of a "long" consonant and only looking at the formal qualities of the diacritical mark.....

And thanks for the 'p' notation above.....that stuff is interesting to me because I'm also a singer. It's amazing to really peek under the hood and see what underlies spoken language, no?...much less try to visually capture that in letterforms.

Cheers!

jazzsammich's picture

DiSH, I can't help but see your sh as "ls", as in ISelly sells sealsells by the sealsore.

I'm feeling pretty sealsore myself.

--Jim K.

dishdesigner's picture

HA! Well, Jim, honey, you better get some salve for that and quick....

I completely agree with your "ls" reading of the lowercase, but I see "LS" in the uppercase. But if you'll note, they also both bear a lot of formal associations to "b" and "B": note the upward curvature of the lowercase character's bowl as it connects to the stem versus the uppercase flat intersection, like "B"...

I think it goes back to the discussions above showing that once an alphabet is learned, any new variants introduced will always first be compared to your learned mental pictures of the existing characters' shapes.

For example, if you didn't already know the alphabet, you might have a similar reaction to the letter "d", mistaking it for a "cl"...

Others noted above how hard it is to find any "natural" shapes for new characters like these that are easily drawn in handwriting (which is why I chose to render them in News Gothic - to not rely on serifs or much stroke variation for legibility), but that are also not too close to existing characters in either Greek, IPA, or other Latin-alphabet languages. It's very difficult.

Did you see pica pusher's example above? They are lovely ideas, but again, it was pointed out that the proposed "th" is essentially the infinity symbol. And when using less 'natural' shapes, it looks to me like the "Th" and "Sh" are rather Cyrillic in tone. But, look at the proposed "Ch" -- it seems very natural to Latin alphabet languages to me in ease of rendering, similarity to existing character shapes, identification of the distinct 'C' and 'h' forms, and overall character width. I think that one character is the most successful of all proposed.....props pica!

The real problem I have with the lowercase "ch" and "sh" shapes I proposed is that the counterforms create smaller and more detailed negative spaces than any other normal alphabetic characters. Therefore, when you squint at the example setting, these characters stick out because they are busier or fussier shapes than other characters.

Fun exercise though....a real brain stretcher.

jazzsammich's picture

Here's my entry into the fray:

The letters Sham, Thay, and Chey.

In preparing these, I thought back to the last time that our alphabetic heritage had separate characters for these sounds, and attemped to retrace their evolution as if they had remained in use (or split into two separate letters, as with the case of Chi becoming Chey/X.)

As I was designing these characters, I paid special attention to their chirographic qualities. That is to say, in order to be successful, the character had to be easily scrawled and still fit fluidly in the normal handwriting scheme, without being misinterpreted as any existing letters or letter combinations.

A couple summers ago, I was preparing myself to do a series of art projects designed to investigate the question of what helps the brain recognize written non-ideographic language as being language vs. abstract symbols or squiggles. The project was based primarily on the creation of pseudo-alphabetic or pseudo-abjadic writing systems: nonsense characters or scripts. The project was put on hiatus due to financial constraints, and I haven't looked back at it since then. I'm curious to look back now and see how these new characters compare with the older explorations. I suspect I probably came up with similar solutions back then.

jazzsammich's picture

Re: the proposed esteso...

Darrin,

What I'm having trouble understanding is what the esteso would accomplish that the caron doesn't do already. What is the difference between the caron letters and the esteso letters?

dishdesigner's picture

Jazzsammich:

Yes, you're right: there would be no phonetic difference in say, Czech, between the caron and esteso I drew. I believe they even call the caron the hacek (pronounced soft 'ch'), no?

But, formally, I think the visual difference is significant in that the upper bowl of the added 'h' is diagrammatically captured, as if the 'h' were hiding behind the altered character and baseline-raised so as to be slightly visible.....but now that you mention it, there might be an interesting option to use the esteso 'c' to represent soft 'ch' in church since it relates to 'h', but then use the caron 'c' to represent hard 'ch' in choir since the caron relates to 'k'?!!

I like your letters by the way.....They are unique enough to be distinguished from other alphabetic characters, and they don't capture awkward counterforms (see my complaint about my characters).

But, have you gone perhaps too deep in designing new letters as opposed to just new characters to represent specific sounds?

The more I read the careful wording of the question and the comments by Jongseong, the more I think the problem is simply to design characters that have a direct and unique correlation to these three sounds as stated (and clarifying which sounds exactly you're designing for: voiced versus unvoiced, etc.). They would probably only ever be used in analytical settings and not in daily writing as letters. Letters are more general than that; most letters can represent multiple sounds in different contexts and when combined in multiple digraphs and trigraphs.

For example, we still haven't solved Jongseong's complaint that your new 'chey' would need two characters to represent the 'ch' sounds in "choir" and "children" and 'thay' would need two characters to represent the 'th' sounds in "they" and "thistle".

My thoughts on your letterforms:

1. Thay: Would there be potential confusion between the lowercase and an oldstyle 7 with a descender? Why not include the backward loop on the lowercase thay descender as shown in the uppercase and make it a formal cousin to q? (it's there in the medieval english too...)

2. Chey: The turn in the uppercase descender not appearing in the lowercase descender seems odd. It's apparent in both cases of medieval english. In fact, I wonder if the uppercase chey needs a descender at all (other than to distinguish from a J with top bar)?....perhaps just cut it flat at the baseline and leave the lowercase descender flat too, further distinguishing it from the backward curled descender of lowercase thay?

3. Sham: speaking of the chirographic qualities of the strokes - it seems that both the Greek and Medieval characters are drawn by first establishing the verticals and then adding a horizontal stroke with a downward curve. Your new letterform makes a change in this order by requiring one stroke to establish the second vertical and downward curve, a stroke that may be awkward to produce? Most bowled characters are drawn counter-clockwise (to wit: O is awkward for right-handers to stroke clockwise). The ones that are stroked clockwise (b,B,D,p,P,R) only require the right half of the circle to be stroked before connecting to a vertical stem. It's the drawing of a vertical line upward instead of stroking it downward that causes you to break a pencil lead.....

Fun stuff to look at, and thanks for the generous letterform etymology diagram to reveal your design thinking!

jazzsammich's picture

Darrin --

That's an interesting point about capturing the h or k in the shape of the diacritical mark... I hadn't looked at it that way before.

When I first learned the caron, I learned it as the hachek in the context of learning Bosnian and Croatian folk songs. For their soft 'ch', they actually have two diacriticals, the hachek for the ch as in church, and the acute for an even softer ch, made with the middle of the tongue instead of the tip of the tongue. It's not a difference most English speakers would identify automatically.

As to the nature of the assignment, I guess I was responding more to the "latin alphabet" part than the "sounds" part. ;) Perhaps in an alternate universe, Typophile Battle #9 is to design new characters for hard chey and unvoiced thay. :D

About the specifics of the letterforms:
Thay: There could totally be confusion between the lowercase and an oldstyle 7, but I imagine it wouldn't be any more than we already find with 0 and O. The lowercase thay has most of its curve before the baseline, with a straight descender, where I think an oldstyle 7 would have a gentler curve to it. I'm glad you noticed the analogy with q... I envision that the backware loop on thay would work in a similar way -- a viable option for script or display work, but absent from most text settings. If I had a bit more time on my hands, I'd post an example.

Chey: Chey also has analogies to Q. Q has one of the only natural uppercase descenders in a roman face (with the exception of some faces that have a long J), so I gave Chey a similar sort of tail so Q wouldn't be so lonely. Also, I think it works well to distinguish it from J with a top bar -- I think it changes the bouma enough to prevent easy misreading. I think lowercase chey should probably remain loopless, though.

Sham: Somehow as a kid I learned to do my O clockwise despite my right-handedness, so it seemed like a natural stroke to me. How is it that you do your vertical in D? Bottom-up or top-down? In my handwriting, I do a downstroke followed by an upstroke, THEN followed by the bowl. Perhaps these are the sorts of individual peculiarities that help give designers their individual voices?

Thanks for the insightful commentary, Darrin! This has definitely been my favorite battle so far. It seemed to bring a lot of thought and debate to the surface in ways that the others haven't. Good work, everyone! ^_^

verdiinpink's picture

My tweak of Helvetica Neue Ultralight 25. Hope you like it.

(Why can't I post a picture, even though I saved it in .png format already and the width doesn't exceed 600 px?)

So, please check have a look at this http://asia.pg.photos.yahoo.com/ph/verdi_55/detail?.dir=/dfc0&.dnm=171c....

and my scribble here at http://asia.pg.photos.yahoo.com/ph/verdi_55/detail?.dir=/dfc0&.dnm=1193....

It would be nice if one of you guys help me post these two pics instead so that to avoid browsing away from this wonderful thread, please please ... (*-*)

--- pinky kinky winky ---

jazzsammich's picture

Pinky --

Take a look at the second item in the thread.

In the meanwhile, here are your images:

Paul Cutler's picture

I don't know about the typographic approach but the message is certainly inspiring…

peace

verdiinpink's picture

Many thanks to jazzsammich
... and I'm sorry with the taboo, don't mean to be rude but this phrase just came up to my mind.

--- pinky kinky winky ---

Tomi from Suomi's picture

Here is my contribution:

Syndicate content Syndicate content