Win a copy of Shady Characters for National Punctuation Day!

Terrible, horrible, no good, very bad photograph by the author.

It’s the 24th of September, which means it’s National Punctuation Day! To celebrate, I’m giving away a copy of Wydawnictwo Karakter’s gorgeous Polish edition of Shady Characters. (My terrible photo does not do it justice – this is a great-looking book.)

To enter, drop me a line via the Contact page and tell me what your favourite punctuation mark is and why. I’ll pick a winner on Sunday the 3rd of October. Be sure to let me know if you’re happy for me to share your entry!

I’ll ship to any continent except Antarctica (I’m looking at you, residents of Henryk Arctowski Polish Antarctic Station).

Good luck!

Miscellany № 91: interrobang archaeology

Funny how time gets away with you in a late-stage pandemic, isn’t it? Here are a few somewhat recent stories of a typographic or emojinal (?) bent that Shady Characters readers may enjoy.

If you recall, the interrobang came into being back in 1962 and was immortalised just a few years later in Richard Isbell’s Americana typeface of 1967. As the first interrobang to take its place in a fully-fledged typeface, Isbell’s “open” version has a reasonable claim to being the canonical form of the character. The holotype of the interrobang, so to speak.

Designed for Richard Isbell's Americana
Designs for the extra bold weight of Richard Isbell’s Americana. (Image courtesy of Fritz Klinke on Flickr.)

As far as I know, no digital version of Americana includes an interrobang patterned after Isbell’s original design. In fact, other than the hand-drawn example here, decent images of Isbell’s interrobang of any sort have general been hard to come by. Until now, that is: the indispensable Stephen Coles, editorial director and associate curator at the Letterform Archive in San Francisco,* recently tweeted that a clutch of Americana ephemera have been uploaded to the Letterform Archive’s website. You can find the complete collection here, but with the permission of the Letterform Archive I thought I’d highlight one particular example:

Americana Italic specimen, showing regular and italic interrobangs
1969 type specimen for Richard Isbell’s Americana, showing both regular and italic interrobangs. (Image courtesy of the Letterform Archive. View the original here. Thanks to Stephen Coles.)

As is traditional for type specimens, this one oscillates between the mundane and the absurd. What’s noticeable, though, is how prominently the interrobang features: either Isbell or ATF must have been quite taken by Martin Speckter’s new mark. Quite an achievement for a symbol of punctuation that had only been invented only a few years earlier, even if its Warholian five minutes never quite translated into a durable presence on the printed page.

Elsewhere, the National Law Review reports on “The Case of the [Allegedly] Stolen Ampersand”. Moshik Nadav Typography, a boutique type studio, filed a claim against Banana Republic stating that the clothing company had stolen an ampersand from Nadav’s Paris Pro typeface. That first suit was dismissed for lack of evidence, but Nadav has filed a second on more limited grounds. Head over to the article to judge (ho ho) for yourself, but Nadav looks to have a pretty solid moral case even if the court has found their legal arguments to be less convincing.

Over at the MIT Technology Review, Tom Mullaney writes about the first digital Chinese fonts. I wrote in The Book about China’s invention of movable type and the problems that hampered its wider adoption, and Tom’s article is a neat account of the parallel issues that affected digital Chinese typography. Aficionados of early desktop computing will appreciate the details:

At the advent of computing and word processing in the West, engineers and designers determined that a low-resolution digital font for English could be built upon a 5-by-7 bitmap grid — requiring only five bytes of memory per symbol. Storing all 128 low-resolution characters in the American Standard Code for Information Interchange (ASCII), which includes every letter in the English alphabet, the numerals 0 through 9, and common punctuation symbols, required just 640 bytes of memory — a tiny fraction of, for example, the Apple II’s 64 kilobytes of onboard memory.

But there are tens of thousands of Chinese characters, and a 5-by-7 grid was too small to make them legible. Chinese required a grid of 16 by 16 or larger — i.e., at least 32 bytes of memory (256 bits) per character. Were one to imagine a font containing 70,000 low-resolution Chinese characters, the total memory requirement would exceed two megabytes. Even a font containing only 8,000 of the most common Chinese characters would require approximately 256 kilobytes just to store the bitmaps. That was four times the total memory capacity of most off-the-shelf personal computers in the early 1980s.

And finally: the king is dead, long live the king! Keith Broni of Emojipedia writes that on Twitter at least, FACE WITH TEARS OF JOY (😂) is no longer the most popular emoji. Enter, instead, the even more hysterical LOUDLY CRYING FACE (😭). Keith wonders if the change is down to the Covid-19 pandemic (“Is there simply less to laugh about now?”), which seems like an eminently sensible conclusion in these challenging times. But as ever, the shifting sands of the emoji lexicon mean that the most straightforward explanations are not always the right ones. Nowadays, ‘😭’ is often used to mean “I’m laughing so hard I’m crying” — which is exactly the same ground once staked out by ‘😂’. Emoji users may not be distressed by the pandemic so much as following emoji fashion.

Stephen can be found at on the web at and, and on Twitter

A reminder: email subscriptions are changing

As I mentioned last week, email subscriptions are changing. If you subscribe to Shady Characters by email, or you’d like to do so, please read last week’s post to learn more. This will be the last post that goes out to existing email subscribers, so please make sure you update your subscription soon!

(If you’ve already updated your subscription to the new system, why not take a look back at my recent visit to Vindolanda, a fort near Hadrian’s Wall?)

Changes to email subscriptions

Feedburner, the Google service that Shady Characters uses to send email notifications of new posts, is closing down its email service. As such, I’ve updated the subscription widget at the bottom of the site to use’s email subscription feature instead.

To avoid cluttering anyone’s email inbox, I will not be moving existing subscribers to the new service. If you’re already subscribed to email notifications and you’d like to continue to receive them, please do one of the following:

In either case, don’t forget to check your email inbox to confirm your subscription.

If you’d prefer to follow Shady Characters on social media, you can find me on Twitter and Facebook. All of the posts made here are also publicised to those sites. You can also use good old fashioned RSS, if that’s your jam.

If you have any questions, please either leave a comment below or drop me a line directly.

Roman all over the place: a Shady Char­ac­ters field trip

Hadrian’s Wall is quite a thing. Its construction is linked to a visit to Britain, in 122 CE, of the Emperor Hadrian, although work may have been underway before then. Conventional wisdom says that Hadrian wanted to keep the restive Celts out of Roman Britain to the south; another interpretation is that the wall was a means to collect tolls and duties from whomever might have cause to pass through it, Celt or otherwise. Whatever the case, the finished wall was eighty miles long, running almost from coast to coast, and it became the abiding symbol of Roman rule in the island of Britain.1,2

Vindolanda is just a mile south of Hadrian’s Wall. It’s the site of a Roman auxiliary fort and an associated village, or vicus, both of which predated the wall but subsequently became part of its supporting infrastructure. The fort was rediscovered some time before 1702, which was the year that a doctor named Christopher Hunter3 described:

a square room, strongly vaulted above, and paved with large stones set in lime, and under this a lower room, whose roof was supported by rows of square pillars of about half a yard high: the upper room had […] two chimneys on each side of every corner or square [.]

All of which sounds exactly like the warm room, or tepidarium, of a Roman bathhouse, with a hypocaust below the floor and chimneys within the walls to convey hot air from the furnace.4

Despite the find, little happened in the way of excavations until the land on which the fort sat was acquired in 1814 by one Anthony Hedley, an Anglican priest and enthusiastic amateur antiquarian. Hedley’s purchase saved what remained of the fort from the stone-robbing and looting that blighted much of the rest of the wall, with Hedley himself rescuing a Roman gravestone from the attentions of an over-eager tenant farmer in 1818. Many more finds would follow and, since the 1930s, the site has been subject to near-continuous archaeological investigation.4

Today, Vindolanda is run by a charitable trust as combination of an open-air museum and an ongoing archaeological excavation.5 It’s a great place to visit if you’re in the vicinity — some of the nearby sites on Hadrian’s Wall itself have a more dramatic outlook, but the scale of the fort and vicus, along with a well-presented museum that houses nearly a century’s worth of archaeological finds, mean that Vindolanda more than holds its own.

On a recent visit, I couldn’t help but linger at a collection of Roman inscriptions — well, replica inscriptions, since the originals are in nearby Chesters Museum — originally discovered by Anthony Hedley himself. Together, they provide a fascinating snapshot of how stonemasons in Roman Britain approached writing and punctuation.

First up is the gravestone that Hedley found in 1818. It’s dated to between 43 to 410 CE, which is archaeology’s way of saying “we don’t know when it was made but we’re pretty sure it came from Roman-occupied Britain”,6 and it marks the death of one Cornelius Victor.

Reproduction of Roman gravestone at Vindolanda
Reproduction of Roman gravestone at Vindolanda. The text is recorded as inscription 1713 at Roman Inscriptions of Britain. (Picture by the author.)

The inscription reads:

D(is) M(anibus)
Corn(elius) Victor s(ingularis) c(onsularis)
mil(itavit) ann(os) XXVI civ(is)
Pann(onius) fil(ius) Saturni-
ni p(rimi) p(ilaris) vix(it) an(nos) LV d(ies) XI
coniux procuravi(t)

Or, translated and with its abbreviations expanded,

To the spirits of the departed; Cornelius Victor, singularis consularis, served for 26 years, a Pannonian tribesman, son of Saturninus, a senior centurion, and lived for 55 years, 11 days. His wife had this set up.”7

All the usual quirks are there: uppercase letters only; dots between words; abbreviations for familiar phrases; too-long words broken across lines. It’s a time capsule of Roman writing customs.

With the gravestone are three altarpieces that all display the same traits:

Reproduction of Roman altar at Vindolanda
Reproduction of Roman altar at Vindolanda. The text is recorded as inscription 1685 at Roman Inscriptions of Britain. (Picture by the author.)
Reproduction of Roman altar at Vindolanda
Reproduction of Roman altar at Vindolanda. The text is recorded as inscription 1686 at Roman Inscriptions of Britain. (Picture by the author.)
Reproduction of Roman altar at Vindolanda
Reproduction of Roman altar at Vindolanda. The text is recorded as inscription 1687 at Roman Inscriptions of Britain. (Picture by the author.)

Moreover, all three show a quirk of Roman numerals that I hadn’t previously thought much about. The inscription for the first, for example, reads as follows:

sacrum Pi-
tuanius Se-
cundus prae-
fectus coh(ortis) IIII

Or, translated,

Sacred to the Genius of the commandant’s house Pituanius Secundus, prefect of the Fourth Cohort of Gauls, (set this up).8

The thing I found interesting is that the number four is rendered not as “IV”, as I had expected, but rather in the “IIII” style found on some clocks.* “IIII” is said to be additive and “IV” subtractive, but very little (that I can find) has been written about when or why the Romans switched from one to the other. The only real convention seems to have been that the additive notation was preferred for inscriptions, especially official ones.9 The colosseum, for example, used “IIII” rather than “IV” for some of its gates.10 Beyond that, the choice seems to have come down to taste and/or context, even within the same document.

I’d love to know if any readers can shed some light on this!

David Breeze, “History of Hadrian’s Wall”
Jarrett Lobell A, “The Wall at the End of the Empire”, Archaeology Magazine, May–2017. 
Gordon Goodwin and F Horsman, “Hunter, Christopher (bap. 1675, D. 1757), Antiquary and Physician”, Oxford Dictionary of National Biography, 2004. 
Robin Birley, “Roman Researches from Camden to Anthony Hedley, John Clayton and Eric Birley”, in Vindolanda : A Roman Frontier Fort on Hadrian’s Wall, 2009. 
“History of the Trust”, The Vindolanda Trust
“An Introduction To Roman Britain (AD 43–c.410)”
“RIB 1713. Funerary Inscription for Cornelius Victor”, Roman Inscriptions of Britain
“RIB 1685. Altar Dedicated to the Genius Praetori”, Roman Inscriptions of Britain
Simon Hornblower and Antony Spawforth, “Numbers, Roman”, in The Oxford Classical Dictionary, 1996, 1053. 
Cycler48, “Colosseum”, Flickr, October–2012. 
It’s amazing the things you notice when writing a book about the pocket calculator