Tuesday November 29, 2005

Claudian Letters

Discovered while spelunking in Wikipedia: the three Claudian letters, proposed by the Roman Emperor Claudius but not, in the end, adopted into the Latin alphabet.  They look weird and have cool names:

  • antisigma is a reversed C—more precisely, judging from the name, a reversed lunate sigma—that he intended to replace the consonant clusters BS and PS.
  • digamma inversum (sounds like an uncomfortable medical condition) is a turned F intended to replace consonantal U.
  • half H apparently has no cool Latin/Greek name (may I suggest demieta?).  It was intended to represent the Greek vowel written with upsilon, which was neither the U nor the I of Latin.

Far be it from me to criticize the emperor of Rome, but these all strike me as fairly tone-deaf (if that's the right word) letter designs.  The relationship between a backwards C, whether pronounced [k] as in Classical Latin or [s] as in the Greek alphabet, and the clusters BS and PS is obscure.  (And why not just adopt the Greek letter psi?)  A turned F for [w] is a little more motivated, since the obsolete Greek letter digamma had that phonetic value, but surely the average Roman-on-the-street (Flavius Sixpack?) would be more likely to think it sounded something like [f]—which isn't all wrong, come to think of it, given the eventual evolution of U into V.  I suppose if he had to base the letter for upsilon on a consonant, H is about as close as he was going to get to a vowel, but then why not base it on U or I?  Claudius clearly had the kernel of a good idea, since W and Y were eventually developed to represent the sounds of the latter two Claudian letters, but I think his proposal was doomed by his taste in graphemes.

Of course, they may still catch on.  There's already a proposal to include upper- and lower-case versions of the Claudian letters in Unicode—in fact, judging from the Unicode Proposed Characters page, it looks like the new characters (U+214E, U+2184, U+2C75, and U+2C76) have already been accepted.  For the other two, upper-case antisigma and digamma inversum, the proposal mentions that there are already pretty good facsimiles: U+2183 ROMAN NUMERAL REVERSED ONE HUNDRED (Ↄ) and U+2132 TURNED CAPITAL F (Ⅎ).  Maybe I'll start peppering my posts with them to see if they catch on.

I am The Tensor, and I approve this post.
12:20 PM in Linguistics | Submit: | Links:


TrackBack URL for this entry:

Listed below are links to weblogs that reference Claudian Letters:


The demieta probably made some sense to Greek speakers, though actually introducing Y made a lot more.

Do we know for sure that consonantal U in that period was [w] and not something closer to [v]?

Posted by: David Moles at Nov 29, 2005 2:54:04 PM

Do we know for sure that consonantal U in that period was [w] and not something closer to [v]?

Well, that's the "classical pronunciation" I was taught in high school, but I don't know how we know. This book might have a convincing argument on the subject, but I haven't read it yet.

Posted by: The Tensor at Nov 29, 2005 4:28:29 PM

I can buy C-C-C-Claudius's antisigma. /bs/ likely surfaced as [ps] anyway, and the resemblance of lunate sigma and the p's curve suggests unifiability as one grapheme by removing the p's stem. Ecce!

Not sure about demi-, since Greek has semi-, and a half H (either half) comes close to looking like a barred I. Claudius probably had in mind the Italic scripts, where so-called "modified I" looks like a hemieta. Cf. Oscan "i-acute":


As for Latin /v/, I heard Donca Steriade thinks it surfaced as IPA beta. But I haven't heard her treatment first-hand, so I don't know what period she proposes it for.

Posted by: Angelo at Nov 29, 2005 4:36:06 PM

If I recall Vox Latina correctly, the evidence is that [w] had shifted to [v], or at least bilabial [B], by about 100, though was still [w] in the Golden Age. The name Valerius is transcribed in Greek with OY then later with B.

Posted by: aput at Dec 8, 2005 8:52:36 AM