I frequently check in with (isfogailsi) …. never explained the voice from your mouth, a feisty and frighteningly learned blog written by Kristina, a grad student exploring the remoter reaches of Japanese history who always has interesting things to say about East Asia, academic life, and whatever else strikes her fancy. Ever since I first started visiting I have vaguely wondered about the name of the blog, but never got around to asking her what “isfogailsi” meant and (at least as important) how it’s pronounced. Now she’s answered my unasked question, and I feel like an idiot because I should have figured it out (it’s Old Irish, and I’ve studied Old Irish), so as public penance I’ll explain it here.

Old Irish has the most complicated verbal morphology I’ve ever had the pleasure of studying; a drastic reduction of unstressed syllables that took place before the Old Irish period turned a language that looks rather like Latin in the few grave inscriptions we have (and that probably would have been as easy to learn) into a nightmare of vanishing morphemes. This particularly affects verbs, which very often have prefixes that in prototonic forms (stress on the first syllable, meaning the prefix) get mashed into the verb stem with appalling results. Thus the deuterotonic (normal) form of the verb ‘they say’ is as-berat, with the stress on -ber-, but the prototonic form is -epret. (Did I mention, by the way, that t is pronounced d and p b?) LIkewise, do-lugai ‘(he/she) pardons’ gets mashed into -dilgai.
So the verb fo-gleinn ‘learns’ has deuterotonic forms like fo-glésed (past subjunctive) and fo-giguil (future), which are bad enough, but the “verbal of necessity” (comparable to the Latin gerundive) is always prototonic, in this case giving the form fogailsi, pronounced FOgalshi (palatalized s is pronounced sh). And since the verbal of necessity always occurs after the copula (which in Old Irish, conveniently, looks just like English: is), we arrive at the phrase that gives the blog its name: is fogailsi ‘it must be learned.’ (Compare Latin delenda est ‘it must be destroyed.’) The fact that it’s written as one word in the blog title gives me a feeble excuse for not recognizing it.
The above explanation doubtless makes no sense to anyone who hasn’t studied old Irish, so if your eyes glazed over, here’s the gist: it’s pronounced “iss FOglshi” and it means ‘it must be learned.’ And if you have any suggestions for a new version of “you never explained…,” Kristina wants to hear them.
Update. Isfogailsi disappeared sometime late last year—it didn’t just stop being updated, it vanished from cyberspace, leaving not a rack behind. Furthermore, its creator hasn’t responded to e-mails. Kristina, if you see this, could you please drop me a line to let me know how things are going? I miss your online presence in general and your blog specifically; nobody else discusses the things you did. Where am I going to go for my Ainu toponyms now?


  1. Discendumst.

  2. It’s not the nicest way of writing it, to be sure, although of course Latin inscriptions and Old Irish manuscripts often blend together with no word spacing. In our standardized texts, we had a dot between the is and the fogailsi, which of course isn’t really easily produced in ASCII. So I mashed it together instead.
    Spacing is very important in recognizing words, I’ve found. Running English all together makesitveryhardtoread; conversely, now that I’m used to Japanese as it is printed (without spacing), the emails that students sometimes post with spaces between the words takes me much longer to decipher. (Not that I’m fluent still, or anything, of course….)
    Old Irish did end up to be a very unstable language, at least at the time of its romanization. My favorite verb classes, (verbs ended up being what our class mostly focused on) were the deponents, which were far more regular. Our professor told us, actually, that aside from the glosses we focused on in the first half of the class, very few texts were purely Old Irish: most show the signs of Middle Irish already in that era. Even the grand ol’ reference couldn’t list verbals of necessity for most verbs, only speculate them; particularly since, if I’m remembering the notes right, the verbals of necessity were mostly used in translating Latin (like the glosses). Many of more mundane paradigms were also incomplete….
    The old line about Sanskrit, isn’t it? That it was “more perfect than the Latin, more complete than the Greek”: that is, that it had more tenses than Latin and more cases than Greek, we were told early on that Old Irish was even more so. It makes me sad that I didn’t keep up with it better. (As I eye a $150 grammar for Manchu…. Not enough time for all the languages.)

  3. · or · will give a centred dot in almost all of the browsers out there.

  4. Alas, but doesn’t help you with urls, which is where I first used the name. I suppose I could change it on the page itself, but… Well, I’ll think about it. (I’ve now gotten used to it, rather.)

  5. No point changing it; it’s distinctive, and we’ve all gotten used to it. And I doubt the Old Irish scribes were very scrupulous about word breaks.

  6. Robert Schwartz says:

    “Spacing is very important in recognizing words, I’ve found.”
    Wow a connection between dispate topics.
    there was a thread about word processing:
    to which I contributed the following:
    Historical Note:
    “Ancient Greek was written in upper case characters with virtually no word or sentence division and without punctuation.”
    This was probably less true of Semitic languages (Hebrew, Aramaic, Arabic) see:
    In this respect formatting text is a modern innovation, related to the adoption of moveable type printing.
    So, Is spacing an innovation that came from technology, or is it a feature of languages?

Speak Your Mind