Pre-Rome, the Greeks themselves did not carve large monumental inscriptions… their stone-carved text was very small and to the point. While these letters were often beautiful, they do not stylistically align with the Trajan Column's letters.

In their Empire, the Romans did in fact carve large monumental inscriptions in Greek, at least on occasion. If a Greek would find these letters Greek enough is not for me to say.

And now I'm gong to speculate, with no feet standing in the realm of my own factual knowledge: Perhaps the Byzantines, who were not Romans at all even though they regularly called themselves as such, made some monumental lettering. They did think that they were good Romans after all, and it was the Romans who did that sort of thing. Except that the Byzantines (at least after a certain point) spoke Greek instead of Latin. So…

As for Cyrillic, after the second World War, the Russians did carve a lot of imposing monumental Trajan-like lettering. I mean, why wouldn't they? I bet you would, too, if you had been them. Moreover, their client-state artists did as well… so East Berlin, for example, has some odd stuff. I bet that Poland and Hungary do, too.

the Russians did carve a lot of imposing monumental Trajan-like lettering
Yeah, I recall one of your Berlin photos, Dan; ‘Trajan Cyrillic?’

There isn't an absolute distinction between Romans and Greeks - by the second century CE, and even more so in the third, it's probably fairer to say that some Romans wrote natively in Latin and some Romans wrote natively in Greek. Or that in the third century BCE very few Greeks were Romans, whereas in the third century CE most Greeks were Romans. The biggest and most monumental of inscriptions tended to be in Latin, because that was the official state language, but even imperial inscriptions were often in Greek in the Greek-speaking provinces.

So Greek inscriptions of the Roman period are Roman, but they're also Greek at the same time, and entirely a part of the Greek cultural tradition. For this reason, in fact, the exact style of monumental capitals used in Latin, like on Trajan's Column, isn't found much in Greek lettering of the same period; in general, the Greek lettering style tended to be somewhat more square, and the serifs are treated differently (often more angular). But the elements of the Latin monumental style were derived from Hellenistic Greek inscriptions, so it's all related. Byzantine inscriptions have a quite different style, looking "medieval" rather than "Roman" (think Russian icons, for example).

To get back to the original question, yes there were grand and imposing monumental Greek capitals, especially in the Hellenistic and Roman periods. No, they do not look exactly like Trajanic lettering, but they are culturally related and thus similar in various ways. You can see a ton of examples, big and small, of various periods, at the Aphrodisias inscriptions site.

How all of this is perceived by modern Greeks is another question entirely, and one I wouldn't know the answer to.

Dan, here is a photo of some Greek inscriptional lettering from within a Roman cultural environment. Unfortunately, I took this picture a long time ago, and I don't remember where. It could by in Lyons, France, or in Rome.

http://www.flickr.com/photos/36766351@N00/1345229290/

Thanks Dan and John! Those are both great examples! There are some other good examples in that Flikr pool that it's a part of. And Dan, you're holding out on some Greek from the British Museum!

I believe that I do have some Greek photos from the British Museum buried in my flickr space. But John's image shows much better what I was hinting at!

I'm looking at that Inscriptions of Aphrodisias page, and they have some cool stuff. Not always the style of lettering that I'm looking for, but check out this:
Building dedication for Aphrodite, Emperors and Demos

That's heavy metal!

I have uploaded two more examples to Flickr, two halves of a single inscription in the Imperial Forum, Rome.

http://www.flickr.com/photos/36766351@N00/2269083743/
http://www.flickr.com/photos/36766351@N00/2269083653/

While you're at it, you might also like to take a peek at this item:

http://www.flickr.com/photos/64189496@N00/125299697

There are similar-styled bilingual inscriptions to be found at Ephesus as well. Someone posted photos of them once upon a time (perhaps here, perhaps on Flickr). Think it was Hrant. (BTW, does anyone know if Hrant's okay? He's been mighty quiet lately.)

Regards,
Ernie

Thanks again guys.

So here's my scenario as it plays out in my head. You have Ioe Schmoe, the Roman letter cutter, and he's handed a pile of Greek letters to add onto his inscription that has a deadline coming, literally. Looks at the letters and says, "Hey, most of these letters letters are the same as Latin. And the ones that aren't? Vell, I'll just drav them the same vay I drav Latin."

Which looks to be the same way the guy did it on solfeggio's example. Cool beans!

So in my font, the Greek letters are drawn the way I think the same person drew the Latin. Make sense?

Here's another bilingual one, this one from a bilingual region (eastern Sicily). It's about a century before Trajan, so the classic Trajanic style isn't fully developed in it, but it shows a deliberate attempt to make the two scripts visually compatible while retaining their own style still, like in solfeggio's example.

The photo David [archaica] posted is interesting indeed, particularly for those curvaceous epsilons (rather than straight-back forms) and the lunate sigmas (don't confuse those with an uppercase "C", Dan). But it is precisely those forms, and in particular the lunate sigma, that gives me pause about both the dating and the provenance of the inscription: perhaps memory is playing tricks on me, but my recollections are that these forms were predominant in Eastern Greek scripts (even turning up on coins found in India). Let's hope someone with appropriate cites to hand will pipe up soon and unmuddle this.

Still and all, Dan, giving the Latin a proper "Greek wash" rather than Romanizing the Greek (as is so often done) might make for an interesting point of departure. In any case, I wish you luck.

Regards,
Ernie

I was wrong about the "eastern Sicily" part - it's from Palermo, in western Sicily. It's an odd one because the lunate epsilon and sigma weren't all that common until later, but the spelling of the Latin looks early - for example the "heic" instead of "hic" is mainly pre-Augustan. The grammar's a bit dodgy in either language. People tend to date it to the early 1st century CE, generally with a big question mark. It's ILS 7680, CIL 10.7296.

Thanks for the clarifications, David. That's a tasty little oddity you've come across.

Regards,
Ernie

Dan, Robert Slimbach got conflicting messages from different authorities, and ended up doing two sets of small caps in Arno, one with accents, one without. Go figure.

I figure do your homework but at some point you have to make up your own mind. So what if your designs have a foreign accent? The locals might find it appealingly exotic and sexy, like Chevalier or Garbo. (Before my time too, but you know what I mean.)

Thanks Nick. I have a feeling, completely without foundation I might add, that when set in stone the Greek reader would have had to figure things out by context, i.e., no accents.

We do the same in English when txt messaging, etc.

Nick: Robert Slimbach got conflicting messages from different authorities, and ended up doing two sets of small caps in Arno, one with accents, one without.

That's not the result of conflicting messages. Historically, one finds smallcaps with and without accents, so there is a matter of preference, which recommends a mechanism by which the user can choose to display accents on smallcaps. So Adobe is, like other developers, using the feature to contextually supress the display of smallcap accents, allowing the user to turn off this feature if they decide they want to see the accents. Now Adobe have a policy of providing one-to-one glyph name to character mapping, so that if Acrobat needs to reconstruct text strings from raw glyph IDs it can do so accurately. This means, among other things, that they include e.g. /A.sc/ and /a.sc/ as smallcap variants: duplicate glyphs with different names for use with different layout features. The Greek smallcap glyphs with and without accents are another instance of the same policy, i.e. a smallcap Alpha with no visible accent or breathing mark may still need to be mapped back to a diacritic character, not to a plain Alpha.

Is it worth noticing what the little bilingual inscription is telling us? The Latin text, using lowercase, is Tituli heic ordinantur et sculpuntur aidibus sacreis cum operum publicorum. Books on inscriptions do not insult their readers by giving translations from Latin. At any rate I have not seen one for this inscription, although we are told that it is bad Latin (and Greek). So here is my very cautious attempt: ‘Inscriptions are prepared and cut here for sacred and public buildings’.

Some of the words are easy. Titulus is a common word for an inscription. Sculpere means ‘cut’. But what about ordinare? Does it mean ‘we write the words’, or ‘we put the words in order and choose the letterforms’? Is it the earliest advertisement for a specialist in the layout of texts? And does the Greek help us with ordinare?

The frustrating thing is that books on Latin inscriptions, like Giancarlo Susini’s The Roman stonecutter (English edition 1973) talk about the text without showing the inscription. The only picture of it that I know is a pale halftone in a good little book in Italian, Ida Calabi Limentani, Epigrafia Latina. (I have the edition of 1983.) She mentions another advertisement in the Vatican and suggests that it is probably for memorial inscriptions: Titulos scribendos vel si quid marmorari opus fuerit, his habes. Something like ‘You can have inscriptions written here and cut by our workmen’?

The image of the Palermo stone in the post is the best I have seen. The moiré suggests that it is from a printed work. Can we know what it is, please?

The Greek renders roughly thus: "stela here chiseled and inscribed [for] sacred temples and public works." The "for" is implicit, hence bracketed. The second "and" could likewise be understood as "as well as..." The "public works" isn't specific, so it might include anything from slabs to denote that, say, "Hadrian's cousin built this place" or perhaps even the more mundane: mile-markers and "This way to Rome" signs. Just can't say. But yes, it clearly is an advertisement for an early chisel wielder.

"Does the Greek help us with ordinare?" Not much at all. But if my Latin's not too rusty, "ordinare" can mean to "set in order" or if you prefer, especially for the case present, "compose."

Hope this helps.

... when set in stone the Greek reader would have had to figure things out by context, i.e., no accents.

And no spaces between words to boot, e.g., in the image above the words for "sacred temples" (ναιος ιεροις) are run together as are τυπουνται και and συν ενεργειαις. Lots of fun to read, ain't it? Standard practice back then. (Please, let's not turn next to complexities of boustrophedon. That's way too near a mindbender.)

John, this from the Arno specimen book:

"In developing these two approaches, we received valuable input from a few of our outside Greek langauge experts, who engaged us in a spirited exchange of ideas, before we settled on the two-option model."

That led me to believe there was a difference of opinion amongst the experts as to whether polytonic small caps should have accents or not.

I solved the problem in my Greek fonts by only having the small caps feature work for monotonic.

This means, among other things, that they include e.g. /A.sc/ and /a.sc/ as smallcap variants: duplicate glyphs with different names for use with different layout features.

A little off-topic, but should "correct" fonts now include two sets of small cap glyphs? -- one for substitutions from caps, one from lower case?

 The Jewish Temple in Jerusalem had warning signs posted in Latin & Greek saying something like, "No Gentile is permitted past here; whoever is caught will have himself to blame for his ensuing death". I don't know if there's an extant Latin version, but the there's a picture of the Greek at http://www.abu.nb.ca/Courses/NTIntro/InTest/JerTem.htm, about ¾ths down the page. This would predate Trajan's inscription by probably less than 100 years.
—Joel

I like the Alphas with the bent crossbar.

Nick: should "correct" fonts now include two sets of small cap glyphs?

A safe answer might be a noncommital, "It may well depend on your audience."

To the extent one can legitimately generalize, contemporary practice in Greece has been this: Greek text set in all caps and/or small caps (and yes petite caps, too, if anyone's curious) uses practically no diacritics. The sole ones to retain are for the vowels with dialytika (diaeresis). Would polytonic or monotonic alter that? Not one whit. Of course an absent accent means one can't rightly tell, without context, whether "ΜΑΛΑΚΑ!" was to mean "μαλακά!" (softly!) or "μαλάκα!" (wanker!), but that's another matter.

Frankly, I've never seen Greek set in smallcaps with accents apart from some ugsome adverts produced for the Greek market by a foreign firm. But I've seen (and set) plenty of smallcap Greek without accents, quite independent of whether the text was Katharevousa (Purist), Archaic, Koine, or contemporary Demotiki.

If this sounds dismissive of the "two-option model," well, yes, I guess it is. Just can't see the point of it, since one can either read the text or not. There may be historic reasons for doing otherwise but I'd like to learn (and see) what they are, as I'm fairly certain most "old timers" who set lead would simply snort and term it a mistake.

Hope this helps.

Nick: That led me to believe there was a difference of opinion amongst the experts as to whether polytonic small caps should have accents or not.

Understandable. I was one of the people from whom Adobe sought advice, and while the discussion was lively it wasn't really contradictory. Most of it resolved around the best way to implement the historically observable options*; there wasn't a lot of disagreement about the need to support those options. I think at several points in the discussion Adobe were on the point of giving up on the whole idea of polytonic smallcaps, as they realised the amount of work involved in getting the casing rules implemented correctly. Its a testament to their commitment to quality typography that they stayed the course and that Arno is as good as it is.

* Or at least, the major options. Once one starts looking at the full range of Greek typography, one encounters all sorts of strange practices that are well beyond what became the (relatively) standard norms of polytonic setting, e.g. accents after letters instead of before them.

A little off-topic, but should “correct” fonts now include two sets of small cap glyphs? — one for substitutions from caps, one from lower case?

You mean in the general sense, i.e. not Greek-specific? Adobe obviously want to encourage it, but I don't think they would use the term 'correct'. They recognise that it is a choice that depends on how much effort one wants to make in supporting Acrobat text reconstruction. Given than more and more PDFs are being produced using software that embeds the original character string in the document, the need for such text reconstruction is in decline, but it is still common enough to be a serious issue. Generally, I do not bother, but I always ask my clients how important it is to them that PDFs created from print streams, with raw GIDs instead of character codes, be accurately searchable. So far, other than Adobe, only one client has wanted to pay the extra money for this functionality in their fonts.

RCC: Frankly, I’ve never seen Greek set in smallcaps with accents apart from some ugsome adverts produced for the Greek market by a foreign firm.

One sees smallcaps with accents in renaissance Greek printing (i.e. outside of Greece but typically with the involvement of emigré Greeks); indeed, it seems to have been the norm, with smallcaps apprently being treated as a variant of lowercase text rather than a form of all-caps.

In this, as in other aspects of Greek typography (e.g. the 'prosgegrammeni' vs ypogegrammeni contention for uppercase letters with mute iota) the issues are complicated by the independent traditions of typography for classical and Biblical scholarship outside of Greece. These traditions are not only independent of the Greek tradition, they are also independent of each other: German conventions differ from French conventions which differ from British conventions.

John, thanks for those clarifications. (I wasn't being set-in-stone adamant, after all, for mine was a plainly hedged bet: "To the extent one can legitimately generalize, contemporary practice in Greece has been..") If you could post some representative examples (or cites), I, for one, would very much appreciate it. (Utterly off topic: Sorry to have missed a chance to meet with you, and others, when you were last over here, but a heart attack and hospitalization sent all my plans off the rails for quite a spell.)

Regards,
Richard

The museum in Thessaloniki had many serifed Greek inscriptions which surprised me
http://www.p22.com/photoalbum/ictvc3/ictvc3-Pages/Image148.html

The moiré suggests that it is from a printed work. Can we know what it is, please?

It's a quick-and-dirty scan from Martin Henig's A Handbook of Roman Art. The ordinantur means roughly "are arranged"; it probably refers to the layout of the text but could maybe also have something to do with helping the client pick the right words to use. So the translations given above are correct. The Latin means something like "Inscriptions are arranged and cut here for sacred temples and public works". The Greek means something like "Stelae are formed and cut here for sacred temples with public actions".

The general meaning is clear enough, but there's room for doubt on the nuances because the wording is odd and awkward in either language. James, if you want to know more about it or maybe get a better image, the person to ask - who would know far more than I do - is Jonathan Prag at Oxford (Merton).

(Apologies, Dan, for taking your thread so far off-topic.)

Richard, I have posted a one example of renaissance Greek smallcaps on Flickr. Click on the 'All sizes' option to view a larger version of the image.

http://www.flickr.com/photos/36766351@N00/2272500310/

This is probably the most famous example of Greek smallcaps, and one of the earliest, from Laurentius de Alopa's edition of Apollonius of Rhodes, Florence 1496. Note the petite adscript mute iota. The scan is of the beautiful collotype reproduction in the original edition of Scholderer's Greek Printing Types. Where smallcaps are used for a body of text, as here, accentuation is probably more important than it would be in a title or other short use of smallcaps.

I have a theory, which I would need to test against more samples, that the distinction between accenting smallcaps or not depends on whether the smallcaps are understood to represent lowercase/mixedcase text or all-cap text. In Latin script typography, we use smallcaps sometimes to represent lowercase letters, e.g. as a stylistic device in the first line of a chapter, and sometimes to represent uppercase letters, as in abbreviations such as UNESCO, but there is no visual distinction between the two. Renaissance typographers found that they could maintain a distinction between Greek smallcaps as variants of lowercase and Greek smallcaps as variants of all-caps by including or excluding accents. I don't know how consistent this use was, how widespread, or how long it persisted.

Here is another example of Greek cap forms with accents from the Estienne edition of Isokrates of 1593.

This is interesting because the title has the accents dropped, but the subhead has the accents. Although the capital letters in the subhead are technically not smallcaps, i.e. they are full size caps from an intermediate size of type, they are being treated as if there were smallcaps, which to this publisher at least means putting accents on them. But the typesetter has a problem, because accents on smallcaps go above the letter, not before the letter as in regular capitals, but he is using full size capitals from a different font so does not have sorts with accents above caps. So he has to insert the accents after the letters.

Perhaps Arno should have three sets of Greek small caps, also including a version with accents after the letters? ;-)

So he has to insert the accents after the letters.

There may be more to it than that. Perhaps he made the distinction that small caps deserved their own style of accenting, so it didn't look like wannabe lower case, and so that it didn't take up too much vertical space (given that small caps are taller than "kappa" height.) And he came up with the "somewhat after and somewhat above" style. Note the different treatment of Omicron Varia and Omicron Oxia.

I don't think this is an attempt to create a new style for smallcap accentuation, Nick. I think the typesetter was making do as best as he could with what was in his type case. He's able to raise and slightly offset the marks over some letters, probably with the help of a file, using mark sorts that were designed to go over lowercase letters. Have you read the analysis of Aldine Greek printing in Peter Burnhill's Type spaces? The lengths that renaissance Greek printers went to is extraordinary.

Thank you, John. The images are interesting indeed.

In the "Isocrates" example above, capitalization is all over the map: Δημόνικον set partly in caps and the remainder in lowercase because of a syllable break (ΔΗΜΟΝΙκον — egad!) and Παραινεσεις capitalized in one spot, but the declined παραινεσεως not so a few lines lower. This is truly wild stuff. Doubt an authentic "pattern" emerges apart from your comment "the typesetter was making do as best as he could with what was in his type case." Nevertheless, it's fascinating.

Yet an unstated question hangs: Though examples from a relatively distant past document such small-cap usage, is there today a good rationale for maintaining what some might view as little more than a curiosity? (More baldly stated: Do you see a genuine market for this? The extra work entailed isn't that hard, but would it likely ever be used?)

Thanks again. Love those ligatures.

Yes, I've read Type Spaces.
You're right -- the typesetter appears to have used the lower case accents (they seem to be the same height), and has kerned them backwards where possible. By putting them after, and not before in the usual titlecase manner of cap accenting, he is pointedly saying "these are accented more like lower case than upper case".

Do you think this conforms to any scribal practice, or is a typographic invention?

Do you see a genuine market for this?

What's possible, and what's available, creates usage -- that's what's happened with the casing of certain Greek characters that only existed in "upper case" or "lower case" forms. So now we have serif fonts with that "lightning bolt" character (can't remember its name) in both cases, which AFAIK, was only ever a capital character, and sans serif.

It would seem that if a typographer today were going to use the "accented small caps" style, they would also want the Gamma that looks like Pi with a short leg (did I get that right?) and the extravagant ligatures, otherwise "authentic" isn't really authentic.

Nick: It would seem that if a typographer today were going to use the “accented small caps” style...

Well, I guess more plainly stated, that if to me, at least at my present geographical remove, seems to loom rather large. I'm genuinely curious whether anyone is using these forms today for anything short of (potentially) pseudo-authentic reproductions. Would it be, in other words, merely a question of "perceived value" — cranking up the overall kilobyte weight by including scads of glyphs no one will likely ever access — or is there a bona fide contemporary need and usage?

I'm decidedly in favor of offering extensive choices to users, mind you, but where, in the case present, is the tipping point at which one's merely adding something that's tantamount to an alien-faced alternate smiley, if you get my drift?

Nick: ... the Gamma that looks like Pi with a short leg (did I get that right?)

Have I overlooked something? Where did you see these? In the example John posted above there are a couple of capital Π characters with shortened right legs (ΠΡΟC is one instance, ΥΠΟΘΕΣΙΣ, towards the bottom of the image is the other). Are those what you meant perhaps?

Yes. But why are there two kinds of Pi?
Is there a grammatical or typographical distinction?

Nick: But why are there two kinds of Pi? Is there a grammatical or typographical distinction?

Neither, actually. It's merely an older form of Pi and, in the example shown, is more "decorative" than anything else. One might suppose, at first blush, that it was selected as an initial cap (for instance in ΠΡΟC) but it's also been used in ΥΠΟΘΕΣΙΣ. Rather disorderly — if not to say devoid of apparent logic. Which is why I so readily agreed with John's observation that "the typesetter was making do as best as he could with what was in his type case."

is there today a good rationale for maintaining what some might view as little more than a curiosity? (More baldly stated: Do you see a genuine market for this? The extra work entailed isn’t that hard, but would it likely ever be used?)

I don't know. One of the reasons I recommend the 'calt' feature approach is that it means the accents are suppressed by default: a user has to take an action in order to display them (and in some applications this isn't even an option, since they don't provide a UI mechanism to turn off the 'calt' feature). My take on this is that the range of conventions of typesetting Greek is so wide and varied that providing for options is a good idea. I know that I tend to look further back than some people when considering what those options might be -- Yannis Haralambous was surprised, I think, by some of the older material I showed him --; I like being able to say 'Look, this thing was possible once, and now it's possible again'. I'd rather people have the option and decide not to use it than not to give them the option.

Do you think [the placement of accents after the letters in (sm)all-caps] conforms to any scribal practice, or is a typographic invention?

I suspect it's a typographic invention. Maybe it was invented during the typesetting of this particular page. I seem to recall Gerry Leonidas saying that he had seen similar things. I'm not aware of any scribal precedence, but then many Greek manuscripts were effectively monocase, using only lowercase letters.

One of the reasons I recommend the 'calt' feature approach is that it means the accents are suppressed by default: a user has to take an action in order to display them (and in some applications this isn't even an option, since they don't provide a UI mechanism to turn off the 'calt' feature).

Not wishing to become entangled in a coding snarl here, John, but wouldn't stylistic sets work equally as well? That is, why not have default smallcap and caps-to-smallcaps behavior for Greek, both mono- and polytonic, without accents (dialytika excepted) and then optionally apply accents to all these smallcaps via a stylistic set? If additional and/or optional typographic "niceties" are deemed requisite for said accented smallcap ranges, such could then be applied via a second stylistic set, right? So doing would appear to fulfill your criteria: accents suppressed by default, requiring users to actively choose to display them.

In fact, this approach seems so straightforward that I wonder why "contextual alternates" was chosen as the feature route to follow rather than "stylistic sets." Issues of application UI support, perhaps? Less point-and-click maybe? What am I failing to notice on this count?

The matter appears to turn more on "stylistic" (at least by historic period) rather than "contextual" distinctions, but there may well be a Procrustean coding rationale at work in the shadows too. Please explain. Thank you.

If you use the stylistic sets approach, then you are obliged to include a full set of Greek diacritic character smallcap variants with accents and a full set without accents. If you are following Adobe's approach on supporting Acrobat text reconstruction, then you are already including such sets, but if you are taking a more efficient approach to glyph set size and avoiding duplicate glyphs, then using 'calt' makes more sense. Also, the handling of the dialytika needs to be contextual, e.g. άι needs to be replaced by ΑΪ, and should be linked to whether or not the other accents are displayed, so it makes sense to include these contextual lookups in a contextual layout feature, rather than in a stylistic variants feature.

A further thought regarding use of accented smallcaps: in the Laurentian edition of Apollonius Rhodius, smallcaps are used for the main body of text, and obviously one wants the reader to be able to read such text with ease, and not be concerned about possible ambiguities introduced by the absence of accents. Even if modern typographers choose to suppress accents in typical smallcap use, there may still be circumstances in which they would find it linguistically helpful to display them, perhaps even for individual words within otherwise unaccented text, in order to avoid ambiguity. Your 'wanker' example above might be just such a circumstance.

Getting back to part one of the original question, might I mention Monumenta? I've no clue if the Greek is "authentic" to Latinized inscriptions, but the credits do say, "These letters are based on Roman and Greek characters carved on stone."