Naxia, a text face featuring Greek

Jongseong's picture

I caught this thread on the Greek ductus a couple of weeks ago, and was inspired to sit down and draw letters to test out my idea of what a Greek text face could look like. I'm building it into a typeface I'm calling Naxia, and I would like your criticism, comments, and suggestions.

I wanted to recreate the rhythm of the classic Greek typefaces like Didot's with a modern edge. The lowercase letters are slightly inclined, but close enough to being upright to harmonize with the uppercase letters and to differentiate them from the not-yet-presentable italic. The letterforms are based on my handwriting, and alternate forms are provided for certain characters.

There is no real underlying logic in the patterns of stress other than doing what seems natural for each individual letter to serve the rhythm as a whole. The angles of stress are all over the place. I sliced off the ends of thicker strokes to create sharp angles rather than the rounded terminals seen in similar designs. This becomes a design feature that adds a subtle sparkle to the page.

Lacking nativity in Greek, I have no idea how much of this works if at all for native Hellenophones. So I would really appreciate help from Greeks especially. How acceptable would this attempt be? Could this work as a serious text face?

As for the Latin alphabet which was added afterwards, it's quite close to some of my previous unpublished designs, but it also takes some subtle design cues from the Greek. I haven't really spent enough time on the uppercase letters for both alphabets, so there's some stroke width issues going on among other things. Not to mention that all the spacing is preliminary and there is no kerning yet.

In short: long, long ways to go. Comments, please!

Naxia01.pdf81.84 KB
paul d hunt's picture

well, i don't want this to get lost in the shuffle, so i'll comment:
it looks pretty decent to my eyes, but i'm no expert in Greek. One question though, why did you choose a form of eta that does not descend as the default? just curious as to your rationale behind this choice.

hrant's picture

That and the alpha make me think that maybe Brian has
been looking at my own favorite Greek font, Apollonia. :-)
But I do think the descending eta is better.

Brian, my only worry (caveat: as a non-native) is that the stroke contrast texture is too jumpy; and that it's less jumpy than the Latin's (even factoring in Greek's larger tolerance for that). Maintaining the "atmosphere" in a multi-script system is very important, since the fonts should ideally be useful in the same way.


Thomas Phinney's picture

If that's the first time you've drawn Greek letters, it's very impressive!


Jongseong's picture

Thanks for all your encouraging comments. Hrant: I too feel a mismatch of tone ("atmosphere") between the Greek and the Latin owing to the different contrast schemes--is that what you're talking about? So if the Greek's too jumpy, but less jumpy than the Latin, which direction should I push which script? In the long term, I might tone down the contrast to make it work at text sizes, so that might help...

hrant's picture

My gut feeling (really not more than that) is that the Greek should usually be
somewhat more jumpy than the Latin. Just don't ask me to define "somewhat". :-)


Jongseong's picture

The Greek lowercase already looks jumpy enough to my eyes, but for a different reason than the innate character of Greek script--incline angle inconsistency. I think I'll concentrate on tempering it somewhat by maintaining the same minute optical incline throughout.

As for the non-descending eta, that's the form I prefer in my own handwriting. That goes for all the lowercase Greek forms I chose as defaults (it would be silly to extend that to the capitals, e.g. the Omega which I usually write as an O with an underline).

The enlargement of the alpha of Apollonia Hrant posted on the Greek ductus thread was indeed fresh on my mind when I drew my alpha. But I had not noticed the non-descending eta in the small sample of Apollonia on the Cannibal website (eta occurs just once there). I see now that the etas are more prominent in the Apollonia Polytonic sample, which I had missed before.

The non-descending eta in Apollonia doesn't surprise me, as it is common in contemporary Greek designs, as one can easily check browsing through the Cannibal website's samples.

As for why I personally use the non-descending eta... I don't know. My Greek hand was strongly influenced by the handwritten Greek I was exposed to when my family lived in Greece, and most of those had non-descending etas, as did a lot of typefaces I saw in use around me.

Perhaps, subconsciously, the eta with a descender bothers me because none of the other vowels have descenders. Then again, a descender might help to differentiate it from pi in some handwriting... Well, if I ever finish and release this a couple of years down the line, then I'll decide which form of eta to put as the default.

If that’s the first time you’ve drawn Greek letters, it’s very impressive!

Thanks, Thomas. It means a lot coming from someone who has designed the beautiful Greek of Hypatia Sans. But I've drawn Greek letters before, though this is the first I've put up for critique. One set of Greek letters I've drawn would surely draw Hrant's wrath for its latinized forms. :)

paul d hunt's picture

thnx for the thorough explanation, Brian! very insightful. i guess i need to look at a lot more native Greek handwriting.

crossgrove's picture

Sorry: another non-native comment: I think it's beautiful and lively, and you've done something I've been interested in for a while: introduced a little of the varying stroke angle of Greek into the Latin lowercase. It livens it up and gives them a little more obvious relationship. Again, I can't read the Greek, but it looks very strong and comfortable to me. Nice!

hrant's picture

If you determine that the Greek is fine, I guess my preference would be a
careful increase in the stroke variance that Carl points out in the Latin.


John Hudson's picture

Good choice of specimen text, Brian. 'Ιθάκη' is one of my favourite poems. I don't know the Dalven translation, though, only the classic Keeley & Sherrard. But setting it as monotonic is cheating. :)

There are some lovely shapes in Naxia, but I think you have too much going on in the Greek: too many flairs and bent strokes, going in all sorts of directions, and this -- combined with the variable ductus -- is what is making the text jumpy. The individual letters have beautiful features, but they are not setting well together. I think you are going to need to rein the design in considerably. The good news is that it is easier to rein in an overly fanciful design than it is to inject life into a dull one.

As a general rule of thumb, I'd say that the more inconsistent the ductus, the more tightly constrained other elements of the design need to be, because that basic inconsistency is already pulling at the texture of the text, threatening to unravel it. If you regularise the ductus, then you can afford more elaboration in other areas, which is how something as amazingly elaborate as a page of Garamond or Granjon's heavily ligatured text holds together so well: the traditional Byzantine ductus provides a consistent foundation for the fanciful forms.

Some specific comments:

The upper left portion in the rho, sigma and final sigma is weak, and causes the letters to fall to the right. This is particularly bad in the sigma. This is a very common problem when one first starts making Greek letters. This area needs to be pushed out more.

The top of the final sigma is too long and too straight: it is confusable with the stigma ligature.

The bowl of the uppercase Phi should be wider but shorter, i.e. there should be more stem above and below the bowl, and compensation in the width to keep the counters nicely open and balanced.

If you make the part of the uppercase psi that crosses the vertical stem squarer, you can open the counters a bit more. At the moment, this letter looks a bit cramped.

Your uppercase tonos accents are too close to the letters. They need a bit more room to breathe.

Jongseong's picture

Thanks for your detailed comments, John. I have made some adjustments since I posted this and started work on an italic, with a more regular ductus, but I haven't done any work on this for the last few weeks as I have entered a busy stretch at work.

I think I'm going to make Naxia regular more upright to set it apart from the italic more, making the adjustments John suggested in the lowercase rho, sigma, and final sigma (good point about the final sigma being confusable with the stigma). I will also open up the counters more. I see John's point about reining in the design. I had some fun exploring the idea of 'What if Greek mimicked the look of being written with the sharpness of broad nibbed pens, only crazier with more flair?', but yes, the letters need to work together more.

I am quite dissatisfied with the uppercase I have so far. I'm not sure they fit really well with either the lowercase Greek or the Latin. They feel perhaps too narrow overall. And maybe I should make at least the Greek capitals lighter.

I hope to get back to work on this soon. In the meantime, for the rest of the typophiles, are there any comments and suggestions about the Latin portion of Naxia?

hrant's picture

> I think I’m going to make Naxia regular more
> upright to set it apart from the italic more

What does the Italic look like? Because the Roman's slant is gentle, so a typical italic won't have any problem standing out. Plus keeping the slant in the Regular is a nice way of differentiating from the Latin.

> I’m not sure they fit really well with
> either the lowercase Greek or the Latin.

Well there's a severe limit to how much formal harmony the UC and lc can have. If you force it too much one or the other case stops working.

> comments and suggestions about the Latin portion of Naxia?

The difficulty is that intelligent comments would have to take the Greek into account. My own lack of expertise in Greek means I can only comment reliably on the Latin out of that context... but here goes:

The descenders are a bit too long for a text face.

The "e", "g" and "x" could be a hair wider.

The "v", "w" and "x" could used more activity.

Your spacing is too loose for the ideal target size
indicated by the vertical proportions and color.


John Hudson's picture

Brian, I don't think you need to make the Greek much more regularly upright, at least not so far as vertical strokes are concerned. As Hrant notes, the slant is very gentle, and will be more reduced once you modifiy the rho, sigma and final sigma, and probably straighten the leg of the mu somewhat (it certainly shouldn't slant outward more than the descender of the gamma, and often curls inward following more the descender of the rho). Most of the apparent slant comes from misapplied weight, rather than from the actual slope of verticals. The raised left side of the pi and tau, for instance, contribute more to the apparent slant than the angle of the legs.

If you are interested in experiments writing Greek with a broadnib pen, including a variety of approaches to ductus, try to find a copy of Epictetus' stoic aphorisms published as Virtue and Happiness by Shambala. The Greek texts are expressively written by Claude Mediavilla. I don't always agree with his choice for individual letters, and I'm not sure, overall, that the calligraphy reflects the meaning of the texts, which are pretty sober. But it is worth taking a look at, and there's a lot of beauty in it.

Syndicate content Syndicate content