My First Bilingual Dictionary

So, I always vaguely believed that we didn't have dictionaries until, generally speaking and depending which language you're looking at, a few centuries ago. But the other day I was in the library, translating quietly away, when I suddenly got stuck in a complicated little phrase in which I couldn't just work out the grammatical import and leave the meaning of the word(s) for later, because it wasn't clear what was noun and what was verb and was was adjective and so forth. So off I trotted to the section holding Latin dictionaries. And what should I find there but - well, a Latin dictionary. Several, in fact, but one of them was in Middle English.

That is to say, while most volumes on the shelf were in the expected form of Latin -> Modern English, this battered little book had all its entries in the form Middle English -> Latin. A thoughtful man called Father Geoffrey compiled it in about 1440, presumably for the aid of those people who were now beginning to write some documents in Latin but had not been trained to it[1]. And I spent about two hours being utterly charmed.

There were the simple entries:

CUMPANY comitiva.

Then there are the ones that need a little clarification, because they could mean different things:

CRYSTEN manne or womanne [ie, adjectective/noun rather than verb]. Cristianus, Cristiana.
LENTE, holy tyme. Quadragesima.
CROCE of a byschope. Pedum, cambuca, crocea.

And, the croce being the cross-topped staff a bishop carries, I'm suspecting that croce and crocea are the ancestors of our 'crutch'. Then he has to distinguish between different uses of the same word in English for which the Latin translation will vary:

CLERE, as wedur ys, bryghte. Clarus, serenus.
CLERE, as water, or oþer licour. Limpidus, perspicuus.
CLERE of wytt, and undyrstondynge. Perspicax.

EEM, moderys brothere. Avunculus.
EEM, faderys brothere. Patruus.

The root word seems to usually come before its derivatives, even if they ought alphabetically to precede it. So CLEYME comes before CLEYMARE (and, for some reason, CLEYSTAFFE precedes them both). And sometimes the order is just a bit random:

FEE foedus.
FEDYN' wythe mete. Cibo, pasco, esco.
FEEDE chyldryn, wythe pappe mete. Papo.

And of course, this being Middle English, sometimes a word is not only listed out of order relative to the words immediately surrounding it, but under an entirely different letter to what you'd expect:

IVYL, or wykkyd. Malus, iniquus.

Verbs are all listed, not by their infinitive, but with a final -YN'. I can't help but wonder if that was the editor mistaking the infinitive for the gerund and putting in an apostrophe to indicate it.

KNAWYN', or gnawyn, or fowly bytyn. Corrodo.

DRYNKE [the noun]. Potus, poculum, pocio.
DRYNKYN'. Bibo, poto.
DRYNKYN' a-yeen. Rebibo, repoto. [Really, how often do you need to use that?]

But this rule broke down a little when it came to 'have', because of the multitude of little phrases that use that verb to form another meaning (like 'avere fame', 'to have hunger' (be hungry) in modern Italian):

HAVE ABhomynacyon, and have disdeyne [despise], supra in HAN.
HAVE in mende [remember]. Recordor, memoro, memini.
HAVE levyr [prefer]. Malo.

There are words that delightfully evoke the texture of daily life:

LYNT, schauynge of lynen clothe. Carpea, secundum sururgicos.
BEKEN with the iye, Annuto, conniveo. Connivet hic oculis, annuit ipse manu.
A-HA Euax.

And words you really hope people wouldn't have had cause to use nearly so regularly:

GELT MAN spado, eunuchus.

Then there are the words that you've never come across before but are just wonderful words that you immediately resolve to use in your own everyday life:

DROBLY, DRUBLY. Turbulentus, turbidus.

Is this a dictionary? Well, maybe not by modern standards, the same way we might choose not to call anyone living in the 14th century a real historian, no matter how accurate their chronicle. And it's only one way - for the English speaker writing in Latin, no help to anyone translating the letter they received in answer. The irregularity of spelling and alphabetisation mean that it's a bit tricky for use as a reference tool, and you have to be willing to search under different spellings or synonyms to find the word you want. But that would not be such an inconvenience to people who are accustomed to taking a long time over books and letters as it would be to a 21st century writer who expects information to spring to one's fingertips.

I don't think it is a dictionary quite yet, depending on your interpretation of the term, of course. I'd call it a word list, a translation tool, because I'd expect a dictionary to be more comprehensive both in the number of words it includes and its explanations of those words. Still, in compiling it, you can see Frater Galfridus considering and dealing with questions of scope, intent and ordering, questions that may be answered differently by the next person to undertake a similar task, but all the more easily for having been posed before.

It's a start!



[1] Frater Galfridus. Promptorium parvulorum sive clericorum. Ed. Way, Albert. Camden Society OS 25, 54, 89. 3 vols. London, 1843-1865. Camden Society listing here. There is also an EETS version published in 1908, but I couldn't find that.

Lingua latina in bucca... libri?

Latin isn't a real language. No, really. No language is a real language unless babies can pick it up. If it has to be consciously learned it's a code, no matter how sophisticated.

And this was surely even more true in the Middle Ages. We have a country that is, overall, tri-lingual: Middle English, Anglo-Norman and Latin. Out of these, the concept of spelling only seems to apply to Latin. And I just don't understand that. Having consistent spelling, with maybe a few errors, irregularities or regional variants, I understand. Considering spelling irrelevant and variable, I also understand. I do not understand how you manage to speak and write two or three different languages, and yet consider consistency in spelling important only for one of them. Granted, Latin isn't entirely consistent - one might contract 'occupaverunt' to 'occuparunt', or waver in how one wants to notate -oe- - and the vernaculars weren't entirely inconsistent, but overall it still bemuses me how the human brain can preserve such a distinction in their attitude to language.

This post was mostly prompted by a certain entry in the Anglo-Norman Online Dictionary, which is, if a little excessive, not atypical. I think this would be a good moment to quote it:

iluec, iluc, ilucs, ilueches, ilueckes, iluek, ilueke, iluekes, ilueks, ilueques, iluesques, iluk, iluke, ilukes, iluks, ilunqes, iluoc, iluoqes, iluque, iluques, ilux; elec, eleckes, elekes, eleuc, elluc, eloc, eloec, eloques, eloek, eluec, elukes; ilec, ileches, ilecqes, ilecques, ilek, ileke, ilekes, ileks, ileoc, ileoches, ileok, ileoke, ileokes, ileoks, ileoqe, ileoqes, ileoqs, ileoques, ileoskes, ileosques, ileqe, ileqes, ileqs, ileque, ileques, ilesqes, ilesques, ileuc, ileuk, ileuke, ileukes, ileuks, ileuqe, ileuqes, ileuques, ilex; illaoques; illec, illecqe, illecqes, illecques, illecqz, illecs, illecus, illeekes, illek, illekes, illeocqes, illeocques, illeoke, illeokes, illeoq, illeoqe, illeoqes, illeoqez, illeoqs, illeoque, illeoques, illeoquez, illeosqe, illeosqes, illeosqs, illeosque, illeosques, illeouqes, illeouqs, illeq, illeqe, illeqes, illeqoes, illeqs, illeque, illeques, illequez, illesqes, illesques, illeuc, illeucques, illeucus, illeuke, illeukes, illeuqes, illeuqs, illeuque, illeuques, illeuquez, illeusqe, illeusqes, illeusqs, illoc, illock, illocqes, illocques, illoec, illoek, illoeke, illoekes, illoeks, illoeq, illoeqe, illoeqes, illoeqez, illoeqies, illoeqs, illoeque, illoeques, illoequs, illoeqz, illoesks, illoesqes, illoesques, illok, illoke, illokes, illoks, illoq, illoqe, illoqes, illoqs, illoque, illoques, illoqus, illosqes, illosques, illouke, illoukes, illouqes, illouqs, illouque, illouques, illouquez, illousques, illuc, illucque, illucs, illuec, illuecqes, illuecques, illuekes, illueq, illeuqe, illueqes, illueques, illuesqes, illuesqs, illuesques, illuk, illuke, illukes, illukis, illuks, illuoqes, illuoqs, illuqe, illuqes, illuqs, illuques, illuqes, illusques, illux; iloc, iloces, iloche, iloches, ilocq, ilocqes, ilocs, iloec, iloeces, iloeches, iloecques, iloek, iloeke, ilokes, iloeqe, iloeqes, iloeqs, iloeque, iloeques, iloesqes, ilok, iloke, ilokes, ilokis, iloks, iloqe, iloqes, iloqs, iloque, iloques, ilouc, ilouke, ilouqes; ylec, ylecques, ylekes, yleok, yleokes, yleoqe, yleoqes, yleoque, yleqe, yleque, ylesques, yleuk, yllecques, yllekes, ylleoqes, ylloqs, yloc, yloec, yloeques, yloques, yloucus, yloukes, ylueqes, ylueques (hiluk Rot Parl1 i 389; ieluec S Audree 1512; ileu Anc Test (E) 7359; iley Durham i 174; ilioqe Rot Parl1 ii 87; illence (l. illeuce) Negotiations 102; illoik BOZ Cont 100; ilokis ROUGH 1; illoiques Edw Ring 66; illonqes (l. illouqes) Stats i 197 (var.); illonquez (l. illouquez) Readings 215; illovecques (l. illouecqes) Foedera iii 488; illus Rot Parl1 iii 91; iluekas Becket 4621 (var.); iluch Anecdota 13; ylooques Art 852)
adv. 1 (local) there, in that place; (local) (motion) there, to that place 2 (temporal) then; then, in that case;

Honestly. Did court scribes hold inter-county tournaments to see who could come up with the most obscure new ways to spell it? All that copying work must have got pretty boring, after all.

So, my conclusion is that the initial question rests on a false assumption. Latin and the vernacular(s) did not occupy the same space in most people's minds, and so they need not match in their rules. Perhaps Latin was learned from the page and largely stayed on the page, while written Anglo-Norman and English were considered a transcript of speech, rather than absolutes in themselves. Latin is a set of rules and signs, almost entirely self-referential: the vernaculars grow and change and steal and adapt. Latin was an internationally recognised code and had to remain consistent in order to BE Latin, but the vernaculars belong in the mouth of each speaker.

And, because I am worse than a dog with a bone[1]: see? Not a real language.


[1] I've yet to see a dog pay any attention to a bone for more than an hour tops. Solace buries hers within five minutes then runs around stealing the other dogs' and burying those.


Bonus points to anyone who recognises the phrase mangled in the title.

Thoughts on 'venire'

Working again, after a five-year hiatus, on memorising Latin noun and verb forms (and I have to say, no wonder all the Romance languages today have dropped the habit of declining nouns - how did anyone ever construct accurate sentences without years of formal grammatical training? Surely there must have been a simpler vernacular!), one word struck me curiously, after the continuous repetitious muttering chant of memorising verb forms that goes on for so long the words are reduced to meaningless shapes of sound.

The chant in question went "uenio, uenis, uenit, uenimus, uenitis, ueniunt". Occasionally, because my focus is on mediaeval Latin rather than classical and it doesn't really matter, it would go "venio, venis, venit", etc. The verb is "venire", "to come"[1], and I spell it with the "v" here because "venire" is also the modern Italian infinitive of "to come". What suddenly struck me as I was muttering, slurring syllables into musical chants, was the way "venio" can have three syllables ("ve-ni-o") or two ("ven-yo"). And this second form sounded oddly familiar. Not from modern Italian - today, the verb runs "vengo, vieni, viene, veniamo, venite, vengono". Italian "vengo" and "vengono" ("they come") are pronounced exactly as they look, each consonant alone ("ven-go", "ven-go-no").

And with one of those pleasing little moments when one more linguistic link falls into place, I realised where I'd heard the form "ven-yo" before:

I' son Beatrice che ti faccio andare
vegno del loco ove tornar disio;
amor mi mosse, che mi fa parlare. (Inferno, ii.70-72)
I am Beatrice who bids you go,
I come from whither I long to return;
Love moves me, and moves my speech.

And the third person plural derived from it:

Quali colombe dal disio chiamate
con l'ali alzate e ferme al dolce nido
vegnon per l'aere, dal voler portate. (Inferno, v.82-84)
Like doves summoned by desire
wings lifted, then still, to the sweet nest
they come through the air, borne by will.

It's not modern Italian, but Trecento Tuscan. In modern Italian, "gn" (as in "gnocchi", "agnello", "magno") is pronounced "ny", so Dante's "vegno" is pronounced "venyo" - ie, very similar to the way Florentines were probably pronouncing the Latin "venio" at the time, as "io" in Tuscan was almost invariably a diphthong[2].

A quick search on Dante Online (which I adore) confirmed that the form "vengo" never appears in any of Dante's works, and "vengono" only twice, both in the Convivio.

But of course, this is a modern (online) edition of the text. The extant manuscripts - of which there are many - vary wildly in their spelling, abbreviations, punctuation, capitalisation, and aeration (where one word ends and the next begins), even if one confines oneself to manuscripts produced in Tuscany in the fourteenth century.

For example, comparing one line across nineteen manuscripts at Dante Online that fit those criteria, we have:

Sol con un lengno e co(n) quella co(m)pagna
sol chon un lengno e cchon quella chonpangna
sol con un lengno (et) con quella compagna
sol chon u(n) legno (et) co(n) quella compagna
sol con un legn (et) con quella compagna
sol con un legno (et) co(n) quella co(m)pagna
sol con un legno (et) co(n) quella co(m)pagna
sol con un legno,' et conquella compagna
sol con un legno (et) conquella co(m)pagna
sol conu(n) legno eco(n) quella co(m)pagna
sol con un legno (et) co(n) quella co(m)pagna
sol con un legno (et) cu(n) quella compagna
solchonun lengno eco(n)quella compangna,
Sol chon vn lengno et chon q(ue)lla (com)pagna.'
sol con un legno (et) co(n) q(ue)lla co(m)pagna
sol con u(n) lengno e con quella conpagna
sol con un legno et con quella compagnia
sol co un legno (et) con quella compagna
sol conun legno (et) co(n) quella co[m]pagna

Letters or words in brackets are rendered by abbreviations imported from Latin. Spelling of the sound rendered nowadays by "gn" varies between "gn", "gni" and "ngn". But it's worth pointing out that the "gn" in both "legno" and "compagna" are descended from the Latin "gn", rather than "ni". And in the same manuscripts, the spelling of "vegno" is actually far more stable (the text is Beatrice's speech quoted above):

Bibl. Medicea Laurenziana: Pluteo 26 sin. 1 3v

Bibl. Nazionale Centrale: Fondo Nazionale II.I.36 5v

Bibl. Medicea Laurenziana: Strozzi 152 2r

Bibl. Medicea Laurenziana: Ashburnham 828 2r

You could have hours of fun comparing the spelling of just those four little extracts. "Beatrice" appears twice as "Biatrice", "I'" appears once as "I" (without any indication of the elided "o", twice as "Io" and once as "Io" with the "o" subpuncted - marked for deletion by a discreet dot underneath. The "or" in "amor" and "tornar" appears in three of them with the half-r, which, in the fourth manuscript, is also irregularly used after "a". Not one is even internally consistent in how they join "che" to the following pronoun ("ti" in the first line, "mi" in the third). The consonant is sometimes doubled, sometimes not, and the second manuscript spells it "che" in the first line and "ke" in the third. Three of them change "di" to "del" and attach it with a doubled consonant to "loco", but the fourth doesn't, effectively omitting the definite article, which could be an interesting Latin-derived mannerism in itself, or just a mistake. One has "dove" for "ove", which means much the same thing but is more common in modern Italian.

But one word that stays consistent - save for the immaterial distinction between "u" and "v" - is "vegno". Perhaps the "gn" in words derived from a Latin "gn" (such as "legno" and "compagna" above) was pronounced more nasally (ng+n), leading to irregular spelling, where the "io"-descended "gn" was harder to confuse as it was one consonant ("n") followed by a diphthong.

Nowadays you wouldn't distinguish "legno" from "vegno" - except, as I said before, that "vegno" has vanished. The consonants reversed, and hardened to "vengo". So at some point, farther down the track, the pronunciation was distinct enough to stratify in the spelling.

I wonder when that happened?



[1] Yes, I know it's conventional in Latin to use the first person singular indicative - "venio", "I come" - to "name" a verb. No, I don't care. To me, the infinitive of a verb is the name. Favouring one finite form over another is silly.
[2] Although, an exception occurs in both of the above quotes - both contain the word "disio", "desire", in which the primary stress falls on the second "i" and it is therefore a full syllable in its own right.

Except where specified as manuscript transcripts, quotes are from the Commedia seconda l'antica vulgata, Edizione Nazionale of the Società Dantesca Italiana, under the general editorship of Georgio Petrocchi (Florence 1994), accessed at Dante Online. Translations are mine.

Well, thif if a nice meff you've goc us inco!

It isn't a typo if you meant to spell it wrong, right? But what if you're a mediaeval scribe, sitting around with his pen and treated pig skin, copying out someone else's crabbed handwriting into your own crabbed handwriting, and you make a mistake? Is that a scribo? Scribble, maybe? Middle English spelling is often said to - well, to not exist at all, really, but most scribes drew the line (as it were) at substituting one letter for a completely different one - say, a 'w' for a 't'[1].

But sometimes you're just not sure. My edition of Gawain (Anderson's) has the following line in the description of the five five-fold virtues represented by the pentangle on Gawain's shield:

Now alle these fyve sethes, for sothe, were fetled on this knyght, (656)
[Now, Gawain possessed all these five sets of virtues.]

Anderson glosses "sethes" and "fetled" in this line: the first as "sets (of five)", the second as... "settled".

I find this choice of Anderson's odd. Why write 'fetled' and not 'setled'? The scribe of the single surviving manuscript of the poem uses the long 's', the one that looks like an 'f' without a cross-stroke. In his handwriting, it's often very hard to tell the difference between the two letters, because the cross-stroke is very small. In fact, sometimes he's just careless and puts the cross-stroke on 's' or leaves it off 'f'. His 'c' and 't' are also close to indistinguishable, which means that we're still not sure whether the name of the Green Knight's alter ego is "Bertilak" or "Bercilak". So (fo?) a modern editor has to make a decision based on a) context and his/her knowledge of the English language and b) alliteration. A) is complicated by the fact that this poet is writing in a rather obscure dialect, and several forms of words - even a few words in their entirety - exist nowhere else in our surviving literature. "Runischly", for example - any editor who glosses that is just guessing what it means by the context in which it appears.

In this case, Anderson seems to have gone with b), and settled on "fetled". This would mean he reads 'f' as the alliterating letter in this line, which in turn means he thinks the main emphases of the line lie like this:

Now alle these fyve sethes, for sothe, were fetled on this knyght.

I just can't agree with that. I like to think my status as a failed singer gives me a little more appreciation for rhythm and the music of words than that. To me, it's clearly the 's' that alliterates, which means the line must be:

Now alle these fyve sethes, for sothe, were setled on this knyght.

Technically, it could be either. So far as I can make out in my facsimile of the manuscript, the cross-stroke of the 'f' is visible in this line - though, as I noted before, that isn't always a sure sign with this scribe, particularly when he's just recently written two 'f's which did require the cross-stroke and had a habit of just getting into a roll with these things. Elsewhere, he repeats whole words from earlier lines that are clearly inappropriate.

The OED does cite "fettle" as a verb at this time ("To make ready, put in order, arrange. Now only dial. to put to rights, ‘tidy up’, scour; also, to groom (a horse), attend to (cattle).") and cites this line from SGGK as one of its examples. It also cites one other example from a more or less contemporary English text, which it mysteriously calls E.E. Allit. P. B. 585: "He at fetly in face fettled alle eres. Ibid. C. 38 In e tyxte ere yse two arn on teme layde, Hit arn fettled in on forme" Clearly, in both of these examples, the alliteration demands 'f'.

Still. Just because the word exists doesn't mean the poet used it here. Perhaps its existence makes the scribe more likely to use 'f' (though this one didn't really need an excuse for his scribos), but I don't think the OED's meaning of "fettled" suits the occasion, and I find it slightly ridiculous that even Anderson thinks "settled" makes so much more sense than "fettled" that he glosses one with the other. I think I fhall fly in the face of Anderfon, Fir Ifrael Gollancz, Tolkien and Gordon, and everyone elfe who'f ever edited thif text and declared the point fettled. I lay my saith and considence in my ear, and shall hencesorth servently believe that the line should read "settled".



[1] Well. Mostly. The Gawain poet does spell both "Gawain" and "Guenevere" with an initial 'w' if he wants to alliterate on that letter instead...