Another great passage from Judith Flanders’ A Place For Everything via Joel at Far Outliers:
Arabic dictionaries also used nonalphabetical methods of organizing. The Mukhaṣṣaṣ, or The Categorized, by Ibn Sīda (d. 1066), was divided, as its title states, by subject or topic, beginning with human nature and continuing on to physiology, psychology, women, clothes, food, and weapons. Al-Khalīl Ibn Aḥmad (d. 791), in his Kitāb al-‘ain, The Book of [the Letter] ‘Ain, used sounds to organize his work: he listed entries in an order of his own, where each sound group was followed by subcategories based on how many consonants a word contained. …
These mainly nonalphabetical developments contrasted with the works of Hebrew scholars, who tended toward alphabetical order simultaneously with (and occasionally a little ahead of) their Christian contemporaries. At the end of the eleventh century, Nathan ben Jehiel (c. 1035–c. 1110) produced his Sefer ha’Arukh, The Set Book. Ben Jehiel, who had been born in Rome, spoke Arabic, Aramaic, Greek, Hebrew, Italian, Latin, Persian, and Syriac, and he drew on his knowledge of these languages to produce an alphabetically ordered book of root words occurring in rabbinic literature. It became one of the best-known dictionaries of its type—more than fifty copies survive—as well as being one of the first Hebrew books to be printed, in Rome sometime before 1472.
Many of these works, both in Arabic and Hebrew, and the scholarship that had produced them, became accessible to scholars in Western Europe for the first time as these languages began to be more widely translated into Latin. … That so many of these works returned to the West via Arabic was significant, for earlier Arab scholars had frequently added substantially to the originals, including details of their own work, which was far in advance of much of Western thought at the time.
The Western rediscovery of the classics had two results, one somewhat abstract, one concrete. More generally, the awareness of how many great works had been entirely unknown before the lifetimes of these new readers, and of how many more had been permanently lost, produced a sense that the current generation needed to ensure that this recaptured knowledge, as well as all the works produced under its influence, were preserved for future generations. Further, it created a drive to ensure that the details contained in all these new works could be found easily—in other words, readers wanted not merely to read the books, but to refer to them: they wanted search tools.
These recently translated manuscripts also brought to the West other elements that are crucial for our story. Educated European readers now became increasingly familiar with foreign alphabets. In Italy and France in particular, Hebrew had routinely been transliterated into the roman alphabet when manuscripts were copied; in the rest of Europe, the Greek alphabet had sometimes been used, but less and less as time went on. In Europe, apart from Spain, where Arabic was in common use, Arabic too had been almost always transliterated into the roman alphabet. By contrast, some in the British Isles were familiar with Old English runes, known as futhorc, or with the Irish writing system known as Ogham. Many more would have recognized, and used in conjunction with the roman alphabet, the Old English runic letters such as thorn (Þ, þ) and wynn (Ρ, ρ). For these reasons, “foreign”-looking letters were more familiar and less unnerving in the British Isles, and so Latin and Hebrew letters were both used, as they were from the ninth century in Germany, a regular destination for highly educated monks from Ireland and Britain.
(Ƿ ƿ) is wyn, (Ρ, ρ) is rho
Very true, but I just checked Google Books and that’s what appears in the book itself. Sad, but I can no longer honestly call it shocking. Shame on the publisher.
There’s more than one way to alphabetize. Lines of Classical Persian poetry are alphabetized by the last letter of the line, secondarily by the next-to-last letter, etc., highlighting the importance of the end-rhyme. Poems are generally “titled” by the first line; there are no separate titles. Therefore, in order to look up a poem, it will do you no good to remember the opening words, such as “Agar ān Turk-e Shirāzi…”, by Hāfez. You will have to remember that the line concludes with an aleph, preceded by a re. The line is found at the beginning of the index of lines or index of poems because the alphabet begins with aleph. The initial aleph of the first word of the line, agar, is irrelevant in this case. It’s the final aleph that determines the alphabetization.
The whole line:
agar ān turk-e shirāzi be dast ārad del-e mārā
If only that Turk of Shiraz would hold my heart in his/her hand
(There is no gender in Persian.)
High medieval Arabic dictionaries also use alphabetic order by last letter then next to last letter, as most notably in Lisān al-‘Arab. But Kitāb al-‘Ayn uses a much more linguistically interesting order: by place of articulation from lowest to highest, much as in Devanagari only without the latter’s rigid separation of plosives from fricatives. Which rather makes me wonder how early Indian dictionaries worked…
Edit: Incidentally, Arabic has two alphabetical orders even excluding such experiments: the original one, closely matching Aramaic or Hebrew and reserved today for numeration, and the normal one, which is basically the former reordered by letter shape.
Wynn is correct in the UK [Picador] edition. Lost in translation.
I weep for my country.
“English monks were less intimidated by the entire Greek and Hebrew alphabets as compared to their Continental colleagues because they were already comfortable with supplementing ASCII with edh/thorn/yogh/etc. when writing in their own vernacular” does not strike me as a particularly compelling argument. I suppose there may have been considerable regional variation in when reading fluency in Greek was lost amongst the (already none-too-numerous) literati – the Ven. Bede is generally said to have been literate in Greek but we probably know this precisely because that was already an unusual accomplishment in the relevant time and place and thus got mentioned by contemporaries.
One of the things that’s also odd about this narrative is treating “Western Europe” as meaning, apparently, “Paris and its environs.” Throughout much of Italy, ongoing contact with Greek-speakers (and writers-in-Greek) never stopped (local in the more southerly parts, as a result of ongoing long-distance commerce for those from Genoa/Venice/etc.), but apparently the fruits of that had more initial difficulty getting over the Alps than the fruits of contact with Arabic-speakers-and-writers in Iberia had getting over the Pyrenees.
I think it’s fair to separate Italy from “Western Europe” when talking about those times, when the Alps were a genuine barrier.
I think that the unbroken tradition of literacy in the vernacular in the languages of Britain (in English, up to the 13C; in the others, until today) has a good deal to do with it.
JWB, LH: The Alps were a barrier, but the sea was a road: travelling between Marseille to Genoa cannot have been that difficult (athough probably more expensive than walking over some of the many Alpine passes). But Italy being mostly a long peninsula provided privileged access by sea to and from many culturally important places.
But for France in the medieval period, the Northern half, centered on Paris, was not easily linked to Italy, from which it was separated not only by the Alps but also by the Southern half, which spoke and wrote varieties of Occitan (then called Provençal), a language that was also learned and used by poets in Northern France, Italy, Catalunya, etc. The major cultural centres in the Southern half were Toulouse and especially Montpellier, not Paris. These cities welcomed not only poets but medical doctors, scientists and scholars from the Mediterranean, including religious ones, many of whom were not Christian. The Cathar “heresy” was actually a different religion strongly influenced by Eastern manicheism. The Albigensian crusade ordered from Paris in order to destroy it was inspired much more by a desire to destroy the flourishing Southern economy and culture than to save souls. In this goal it succeeded, and the region never fully recovered from it.
Good points all; I guess there were really a Northwestern Europe and a Southwestern Europe at that time.
i think in this (as in many things) jewish diasporic geography is more helpful for imagining the cultural territory than the christian/feudal version: before the western european expulsions (spain & portugal being the last), the cluster of {sfarad (iberia + provence/occitania) & italy} stood against the cluster of {zarfat (lang d’oïl) + ashkenaz (rhineland + uppermost danube)}.
the Southern half, which spoke and wrote varieties of Occitan (then called Provençal),
¡¿Then?! It’s still called Provençal in Provence. You have to go west of Montpellier before anyone talks about Occitan. I’m referring to ordinary people, of course. I don’t doubt that academic linguists know about Occitan.
There’s more than one way to alphabetize. Lines of Classical Persian poetry are alphabetized by the last letter of the line,
No doubt David Eddyshaw can tell us, but how do they arrange the entries in a Welsh dictionary? If you have a good grasp of mutations you know that fam will be listed under M. But normally you don’t want to look up a very familiar word like fam, but one you’ve never seen before. How do you manage then if the context doesn’t give much clue of its gender, etc.?
Traditional Welsh dictionaries are alphabetic by basic form of the word, so you have to work backwards from any mutated forms encountered in text or speech. So fam is under mam, falch is under balch, while ficer actually really begins with f- in its basic form (borrowed from English “vicar” too late to be modified into something mutable). Combined with possible vowel changes and endings it can be challenging (e.g. nghymoedd which, obviously, is under cwm “valley”). No hope for it but to look up all the possibilities.
I once attended a talk by a representative of Gomer Press who publish the dictionary Y Geiriadur Mawr. He told us that they got an order for a number of copies from China every year – but no orders for other books from their catalogue or any explanation of what they were using them for. So it seems that somewhere in China a secret army of Welsh learners is ready to foil the plots of the Cymry… I hope they are managing to untangle the mutations.
But the question is not merely why there were barriers to diffusion of translations from Italy to Paris in the 12th century, it’s why those barriers seem to have been greater than barriers to diffusion of translations from early-Reconquista Castille to Paris at the same time, where you have not only the Pyrenees but those Occitan-speakers (and Cathars etc) similarly in the way.
nghymoedd which, obviously, is under cwm
I assume that that was tongue-in-the-cheek: obvious to a Welsh speaker, perhaps, but not to a poor Saxon.
Yes, I wasn’t being entirely serious about nghymoedd. Though initial ngh- can only come from c- and -y- in non-final syllables will often be from -w- in final syllables (including monosyllables), so not quite as unreasonable as it looks.
It was clearly easier to get from coastal Spain to coastal Italy than to go from Paris to Rome. OK, but there are still things I find difficult to understand. Modern educated Spanish is reasonably close to being mutually intelligible with modern educated Italian. The first time I went to Italy with my wife she had no difficulty communicating with Italians in the street or in restaurants if she spoke Spanish and they replied in Italian. So far so good, but Catalan is geographically between them, but more difficult to understand than either. How do we account for that?
Note FWIW that this is just after the era when the Alps had proved no barrier to Norman conquest of Sicily and much of southern Italy, putting quite a lot people competent to translate directly from Greek to Latin under the authority of rulers with connections to northern France. But I guess getting Greek works rendered into Latin to be sent back to northern France wasn’t one of the priorities of those Norman rulers?
Thanks. I’ll try to remember nghymoedd/cwm as it’s a nice example. I suppose that -oedd is a plural marker, so now you can explain if there is a simple way to guess the plural of a noun. When I was studying Teach Yourself Welsh half a century ago (there are not a lot of Welsh speakers in Marseilles, but I was living in Birmingham then) I found much of the language reasonably straightforward to grasp, except for the plurals, which seemed entirely arbitrary.
Geographically central innovations that haven’t reached Castilian or Italian.
Also, Standard Italian is essentially 13th-century Florentine frozen in time, which explains in part why it’s so conservative.
David Marjanović, Athel Cornish-Bowden: It’s more complicated than that. A MAJOR stumbling-block with oral comprehension of Standard (Barcelona) spoken Catalan involves some recent phonological innovations, notably the reduction of unstressed vowels: older /a/ and /e/ are realized as schwa, older /o/ and /u/ merge as /u/. This feature is absent in spelling, and I would be willing to bet that Catalan would be judged just as easy or even easier to understand than Castilian by Italian L1 speakers if reading comprehension alone was tested.
More broadly, it is important to remember than recent, superficial sound changes can dramatically lower intelligibility: Portuguese and Spanish are part of a common subgroup within Western Romance, sharing a number of common innovations, and yet by some accounts L1 Spanish speakers find Italian (which is definitely NOT Ibero-Romance!) easier to understand than (especially) European Portuguese, whose phonology (unstressed vowels especially) has, over the past few centuries, changed drastically and made spoken European Portuguese much more impenetrable to Spanish (and indeed other Romance) L1 speakers.
I lost what I was writing in reply to Etienne, but I won’t try to restore it now because it’s 1h40 and I should be in bed. (Normally I would be, but we’ve just finished participating in a homage of the Chilean Academy of Sciences to Humberto Maturana (who died earlier in the month), and the time that was convenient for them meant it started at midnight for us.)
Welsh noun plurals are easy, just like Hausa noun plurals; all you need to do is learn them individually as you go along. What could be simpler?
Actually there are some identifiable patterns to Welsh plurals, such as animal names tending to take -od, and most of the words that do the singulative -yn/-en vs plural null thing falling into recognisable semantic classes. Quite a lot of noun-forming suffixes go with predictable plurals too. But I suspect it would be like trying to remember such rules as there are for gender in French: more trouble in the end than just learning the items individually anyway.
(At least you don’t have to learn two verbs* for every English verb, like you do in Russian.)
* or more …
such as animal names tending to take
Reminds me of the Russian word опёнок (a variety of mushroom, usually Armillaria spp.), which is etymologically о-пён-ок (something like “by-stump-er”), but had fallen in with an etymologically unrelated but large (and productive) family of “baby animal” words with singular -ёнок, plural -ята.
[I don’t recall offhand how exactly did the “baby animal” words end up looking like that; I do vaguely recall that it’s a complicated story involving sound changes and/or analogical remodelling in Old East Slavic, and that the original singular was -я.]
So the prescriptive (and historical) plural is (the more generically regular) опёнки, but approximately everyone says опята.
The question is, why do languages go in for regular inflections at all? Why aren’t all plurals (or accusatives) suppletive? If we can memorize N different noun roots, it can’t be any harder to memorize 2N singulatives and their suppletive plurals.
While at it, make all inflections suppletive. It’ll be easier than trying to remember which German nouns take -e, -en, etc. in which case.
East Nilotic languages not only have unpredictable noun plural formations but in several languages mark the nominative case by unpredictable tone changes (which may differ from singular to plural.) So they are showing the way to Y’s linguistic utopia already.
And there is always Navajo, which I have seen memorably described as “a verb-centred language in which all the verbs are irregular.”
So what is needed is a mixed Copper-Island-Aleut-style or Michif-style language based on Navajo verbs and Maasai nouns.
Catalan was once spoken in parts of Italy, if I’m remembering correctly, although I don’t know if it was spreading out, or if it’s the language of a coastal area which has now retreated. (I just have a vague idea it has something to do with Christopher Columbus.)
Written Catalan at least looks to me like Castillian crossed with French. I don’t know if it would be seen as a kind of French(/Occitan/Provençal) if Castillian didn’t exist…
My brother teaches German as a foreign language and has recently provided the key to recognizing n-stems in German: all the masculines that end in -e, -ent, -ant*, -ist or Latinate -at, plus Herr, Bär, Mensch. Now all you need to do is remember which nouns in -e are masculine, because there are plenty of feminine and neuter ones, too!
(While I’m at it, there’s a single neuter n-stem, Herz; it has a wholly unique genitive, Herzens, and I have no idea why.)
* Including Elefant.
The city of Alghero in Sardinia (where it’s now dying out).
Catalan is indeed close to Occitan (to the point of being sometimes regarded as a dialect of it, though there is understandable political resistance to the notion.)
https://en.wikipedia.org/wiki/Catalan_language#Relationship_with_other_Romance_languages
@dm
re Herz/Schmerz, here are some old corpus results:
Medizinisch-naturkundliche Sammelhandschrift (‘Astromedizinisches Hausbuch’). 1485.
Vnd ist yegliche gut zelassen / fur allen smerczen / vnd ge- pressten / der prust / vnd des hertzen / Auch des magen / der huff / vnd der seytten / vnd fur allen smertzen / der lungen / vnd dieselben Adern / sol man lassen / an dem funfften tag / des vollmonets / oder dabey.
From this you see that:
1. the final s in genitive was at that time not used for Herz–Herzen was the full genitive form
2. der Schmerz had (and still has!) the en plural.
›Secreta mulierum‹ mit Glosse in der deutschen Bearbeitung von Johann Hartlieb (1465). Hg. von Kristian Bosselmann-Cyran. Pattensen (Hannover), 1985.
O mensch, gedenck des schmertzen deiner mueter, den sy von dir jn der geburt gehabt hat!
From this you see that Schmerzen was also possible in the genitive singular (unless you argue that “des schmertzen” is to be interpreted as genitive of das Schmerzen, not der Schmerz–I would consider that unlikely but OK).
[N. N.]: Artzney Buchlein wider allerlei kranckeyten vnd gebrechen der tzeen. Leipzig, 1530.
Zum dritten wirt der schmertze gestillet mancherley tzuthuung der ertzteyen/ vnd der yenigen die den schmertzen stillen
This is an unambiguous accusative den Schmerzen for der Schmerz.
@Paddy: one of the things every student of German language history learns is that the noun inflection classes all were thoroughly mixed, shaken, and stirred in Early Modern High German (Frühneuhochdeutsch), so that knowledge of how a noun is inflected nowadays doesn’t help you much for texts before the 16th/17th. The texts you quote are at the tail end of that period, and of course there are deviating inflections also later (coming from dialects, or words moving to another inflection class only later).
@hans
Thanks. I suppose the only real observation is that Schmerz also kept en in the plural. Is des Herzens then some kind of double ending (archaic genitive in en retained but s added when this became confusing?)?
one of the things every student of German language history learns is that the noun inflection classes all were thoroughly mixed, shaken, and stirred in Early Modern High German
Clearly the only sensible thing to do is to eliminate all noun inflection in German. It would provide almost as much relief for learners as getting rid of the characters in Chinese.
Is des Herzens then some kind of double ending (archaic genitive in en retained but s added when this became confusing?
I assume so, but to be sure, one would have to check historical grammars.
@LH: That language exists. It’s called Dutch. 🙂
Ah, thanks for reminding me – des Herzens isn’t unique, because des Magens has the same synchronically double ending. And Magen is masculine.
Oh, one more: Friede, also masculine, makes des Friedens. The fun part of that one is that it didn’t use to be an n-stem at all, but a *u-stem.
Conversely, Hahn has been retconned as an *i-stem (nom. pl. Hähne, gen. sg. Hahnes), but the prefix form remains Hahnen- because this word was a *n-stem at some point.
Schmerz is masculine and forms its plural with -en, but the gen. sg. is des Schmerzes. I think what we’re really looking at is the genitive being a feature of the written standard that most dialects lost hundreds of years ago…
Even some of Goethe’s writings are plain ungrammatical today: some nouns have changed gender, some details of adjective declension have changed…
There are more masculine words with a Genitive in -ens. In a few cases (Friede, Name, Drache) the -en- is still an oblique stem, in others (Magen, Morgen, Regen, Besen, Boden, Balken, Garten etc), the -en- is part of the word stem, if not etymologically, then at least synchronically. And even Friede, Name, Drache have by-forms (colloquial or even literary) with a nominative in -en, showing a rebuilding of the noun from the oblique stem. It’s interesting that the male n-stems with a gen. sg. in -en (Riese, Hüne, Sachse, Bayer etc.) don’t show such a development, at least in Standard German.
I agree, and I bet that it would be even easier for Northern Italians with some (even passive) knowledge of their local dialects.
Oh yes, aren’t Venetian and Catalan supposedly mutually intelligible?
Let me just confirm this – sometimes I actually have to think about Friede.
David Marjanović: Your brother sounds like the sort of German teacher I wish I had had: My first German teacher was Swiss and VERY thorough (some national stereotypes do have a grain of truth), but treated noun plurals as though each and every one of them was unique and no general rules were possible.
In answer to your question (“Aren’t Venetian and Catalan supposedly mutually intelligible?”), this is VERY difficult to establish today because of the absence of a sizeable community of either Catalan or Venetian monolinguals: unfortunately there is no way to filter out knowledge of Castilian among Catalan speakers and Italian among Venetian speakers.
Jen in Edinburgh: in answer to your question about Catalan (“I don’t know if it would be seen as a kind of French(/Occitan/Provençal) if Castillian didn’t exist”), the answer is definitely YES. In fact, it has been pointed out that grouping Catalan and Occitan together, in a subgroup from which which both French and Castilian would be excluded, is far more justifiable (linguistically!) than grouping French and Occitan together to the exclusion of Catalan, or Catalan and Castilian to the exclusion of Occitan.
David Eddyshaw: I really must take issue with your statement on rules for Welsh plurals (“But I suspect it would be like trying to remember such rules as there are for gender in French: more trouble in the end than just learning the items individually anyway”): having taught a course to French L2 learners which allowed them to accurately predict the grammatical gender of a few thousand nouns, learning such rules DEFINITELY eases the language learner’s task. Whether this is also true of Welsh (or German) nominal plurals I leave for you and others to decide.
I am reminded, in this context, of a remark made to me by a fellow linguist, a native speaker of Romanian, who, in answer to my bemoaning the fact that its nominal plural-marking morphology seems to be the least regular of any Romance variety, told me that if you know all the sound changes separating Latin from Romanian and have a solid foundation in theoretical issues surrounding morphology, you actually do have a better than fifty/fifty chance of correctly predicting the plural form of a word with a Latin etymon. No, I did not find that very encouraging either.
“But Kitāb al-‘Ayn uses a much more linguistically interesting order: by place of articulation from lowest to highest,”
A famous speculation was that Arabic dots work the same way (Revell, 1975).
Jein. (Ni oui, ni non, bien au contraire !) If you know the gender of a noun and have learned to recognize n-stems, you can narrow it down to 2 or 3 options, or probably even 1 sometimes, but I don’t think it’s possible to do better than that. After all, -er has been spreading irregularly through the lexicon for hundreds of years, and there is a bit of more or less regional variation within Standard German (Wagen ~ Wägen as pl. of Wagen, Generale ~ Generäle as pl. of General, Knie ~ Kniee as pl. of Knie*, Schi/Ski ~ Skier as plural of Schi/Ski – Skier looks so very Scandinavian, but it’s not…).
…though there’s not a lot of variation between the dialects, actually.
* I googled “kniee” to make sure I was really remembering that form. Here’s a prescriptivist condemning not only Kniee, but also “all the other regularly encountered plural forms” like Knieer, Knier and Knies, plus the interesting spelling Kniehe for Kniee. I had no idea. – While I’m at it, note that that prescriptivist blithely uses neuste. That form has recently** become very common** in Germany, but the word is neueste, dammit, with three syllables.
** or recency illusion, frequency illusion, I have no idea.
Now that I think about it, that may actually be where the story comes from – Catalan- and Venetian-speakers have above-average experience with Romance diversity, so they’re more likely to understand any other Romance language than a monoglot is.
David M.: Schi/Ski ~ Skier as plural of Schi/Ski – Skier looks so very Scandinavian, but it’s not…
I won’t claim to understand what you mean here. I’ll just rush to add that it probably doesn’t help that the original Scandinavian is irregular too, The current Norw. word is feminine in the singular and neuter in the plural:
ei ski – den skia – to ski – begge skia
There are dialects that preserve a neuter singular.
et ski – det skiet – to ski – begge skia
I think the feminine singular is a result of interpreting the definite plural as singular. It’s a word that often is used in the plural.
A similar case is sko. For me it’s masculine in the singular and neuter in the plural:
en sko – skoen – to sko – begge skoa
For my kids it’s parallel to ski
ei sko – skoa – to sko – begge skoa
About degree of mutual inteligibility as a measure of distance between two dialects.
Will the triangle inequality hold?
I mean, this*
The worst one here is #1. I can’t pretend that I udnerstand myself well:) #2 is bad too. But it never occured to me to think about the triangle inequality.
—
* the link is to the axioms of a metric space in Wikipedia:
1. d(a, b) = 0 ⟺ a = b
2. d(a, b) = d(b, a)
3. d(a, b) + d(b, c) ⩾ d(a, c)
…for exmaple, a, b, c are languages and d(a, b) is how well speakers of a understand speakers of b. Can you have pairs of very similar dialects (a, b) and (b, c) where dialects a and c are very dissimilar?
Do you mean
https://en.wikipedia.org/wiki/Dialect_continuum
or are you thinking about something more complicated?
d(a, b) = d(b, a)
I think there are examples where speakers of A can understand speakers of B but not vice versa. I can’t think of any, though (except of course when speakers of A are bidialectal in B but not vice versa)
I have the feeling this is to some extent the case with Portuguese (A) and Spanish (B), and didn’t somebody recently say something similar here about Czech and Polish?
I think there are examples where speakers of A can understand speakers of B but not vice versa.
Speakers of Bron Akan understand Twi Akan but not vice versa …
Actually the two dialects of Kusaal stand in this relationship to one another to some extent; Agolle speakers seem to understand Toende Kusaal without much trouble, but Toende speakers in Burkina Faso who haven’t been previously exposed to the Agolle dialect report low levels of comprehension of it. The situation may be complicated in part by sociolinguistic attitudes: Toende speakers believe that their (minority) dialect is “purer”, apparently. But it probably also has to do with the fact that Agolle has diphthongised some of the proto-common-Kusaal vowels, which remain unchanged in Toende, and undergone more in the way of intervocalic dropping of velar consonants: it’s (as it were) the Portuguese to Toende’s Castilian.
https://www.sil.org/resources/publications/entry/9184
I think there are examples where speakers of A can understand speakers of B but not vice versa.
I’m guessing (somewhat guided by David Eddyshaw’s descriptions above) that such an inequality might happen in cases where A had undergone a commonly-occurring sound merger (and/or reduction) that B had not: speakers of A would understand that both B sounds are varieties of their one A sound a lot easier than the speakers of B could figure out which B sound is being implied by the A sound in each case.
If I had to guess a pair that would be a good example if this hypothesis is correct: Danish and (either kind of) Norwegian respectively.
As for problematic triangle-inequality cases, they can happen when B had been in close contact with both A and C, but A and C had not otherwise been in much contact with each other. I’ve heard this of Russian/Ukrainian/Polish; not sure if it’s actually true. There are probably better examples.
I did in November:
I’ve probably said it a few more times since 2010, but I can’t find that now.
Anyway, rozele relativized that experience in the same thread.
@David Eddyshaw, I think it must have occured to everyone here that:
The distance between Italy and France is 0. The distance between France and Spain is also 0.
0+0=0. Then the distance between Italy and Spain is also 0. Hurrah!
What I mean: people try to measure differences between langauges in many ways, by mutual intelligibility too.
Sometimes they call it “distances” and imagine langauges as points dispersed in an imaginary space.
One issue is obvious: when the you-to-me distance is not the same as me-to-you distance – can we call it a “distance”? And which one?! One can try to ignore this problem by using the average instead ((me-to-you + you-to-me)/2).
But then there is that triange inequality. If the distance between you and me is 4000 miles and we both are only 1 step away from Y – it is crazy. Can’t I go to Y first and then from Y to you, reaching you in two steps and call it a distance?
A lot of the commoner cases are probably largely questions of prestige. Speakers of low-status dialects, if not positively bilingual in their own dialect and the prestige dialect, often have at least a good passive knowledge of the Posh-sprache.
This is basically the position with Bron vis-a-vis Twi. It may explain some of the Kusaal situation, too: although Agolle Kusaal is now very much the more high-profile dialect in Ghana, the first attempts by missionaries to evangelise the Kusaasi used Mooré and then moved on to Toende Kusaal, which was the basis of the earliest proper grammatical studies of Kusaal (by the Indefatigable White Father André Prost, and then by David and Nancy Spratt.) There are a few loanwords from Toende into Agolle Kusaal which reflect the former prestige of the Toende dialect, like faangid /fã:gɪd/ “saviour” from Toende fããgit instead of the echt-Agolle faand /fã:d/.* The original 1976 New Testament in (Agolle) Kusaal has e.g. malek “angel” and aaruŋ “boat”, both Toende forms, now replaced by the proper Agolle maliak and anrʋŋ.
* The survival of this form almost certainly has something to do with the fact that faand means “robber.” It’s actually the selfsame etymon, the agent noun of the verb faaen /fãɪ̃/ “snatch, snatch back, rescue”, and my chief informant confirmed specifically that faand had both meanings in his idiolect.
it is of course logical (also cf. “captivating”) and I like it.
it is of course logical
True: it all depends on who is robbing whom:
This little babe, so few days old,
Is come to rifle Satan’s fold …
Or: one man’s terrorist is another man’s freedom fighter.
One can, however, see why the pious translators were bothered by the saviour/robber conflation …
To be somewhat less subversive, it is actually possible that two originally distinct verbs have fallen together here: faaen /fãɪ̃/ is the regular Agolle outcome of Proto-Western-Oti-Volta *fããgɪ, and in WOV *gɪ can be a “reversive” derivational suffix; moreover, there is also a root-stem verb fan /fã/ “grab.” So it is conceivable there was once a verb specifically meaning “ungrab” involved somewhere in among all this. In Agolle Kusaal, the regular loss of *g word-internally after long low vowels has caused verb forms of this pattern to fall completely together with CVV-stems everywhere except in the perfective and gerund, including in agent nouns.
The triangle inequality fails for Danish, Swedish and Norwegian, as previously covered — the Norwegians handle both Danish and Swedish pretty well and vice versa, while Danish and Swedish is like dogs and cats.
This clearly has something to do with the fact that Norway has two (not quite) official languages, one a lightly phonologically adapted Danish and the other in a dialect continuum with Swedish. I don’t know if native speakers of Norwegian pick between features they command when talking to either Danes or Swedes, at least at the unconscious level where you always adapt.
Lars: This clearly has something to do with the fact that Norway has two (not quite) official languages, one a lightly phonologically adapted Danish and the other in a dialect continuum with Swedish.
Yes. Or with the resulting wide acceptance for dialect diversity in the public space and exposure to variation. This exposure is probably diminishing, though, with the younger generations overwhelmingly informed and entertained by English-language media.
I wouldn’t say that it’s Nynorsk specifically that is in dialect continuum with Swedish. The districts whose dialects approach Swedish almost uniformly prefer Bokmål, and most of them always have. Nynorsk was standardized with some specifically Western features, while Bokmål has been open for Eastern grammar and vocabulary.
I don’t know if native speakers of Norwegian pick between features they command when talking to either Danes or Swedes, at least at the unconscious level where you always adapt.
All of the above, I think, and with a great deal of individual variation. I have a young colleague who likes to speak and write plain Swedish with Swedes and plain Danish with Danes. My approach has been to talk as I usually do, but slow down or repeat with a different wording if I sense that I need to.
A fine theme for this strange National Day. Hurra for søttende mai!
Gratulerer med 17. mai!
Danish and Swedish is like dogs and cats
What about the cross-border cop series Bron/Broen where the Danes speak Danish and the Swedes speak Swedish yet no grisly murder goes unsolved. Don’t tell me this mutual comprehension is part of the fiction?
In Atlantic Crossing, a TV series about the Norwegian royal family in WWII that is currently airing in the U.S. on PBS, the (actor playing the) king speaks Danish, the crown prince speaks Norwegian, his wife speaks Swedish, and nobody looks “Whaaat?” at anybody. This extends also to conversations outside the family.
What is truly wonderful is that the English subtitles actually tell you who’s using which language at the start of each conversation.
Looking down the list of cast, I think they may even have used actors with roughly the right dialect for the government ministers.
(I haven’t seen the series, thinking that I would be annoyed by the reported veloc&laxity with history, but I should perhaps reconsider.)
My wife and I watched the first several episodes because we each have familial connections with Norway and wanted to know more about the history involved (without going to the trouble of reading a book about it); the first few, about the German invasion, were very enjoyable — it was great hearing the Norwegian, and some of the acting was quite good — but once the action moved to the States it became an absurd soap opera with only a tenuous connection to reality (Crown Princess Märtha inspired FDR to create Lend Lease? Eleanor Roosevelt was a conservative trying to keep FDR from doing good, progressive things? seriously??) so we stopped watching. I assume the part set in Norway is more accurate, because the series was created by a Norwegian company and Norwegians would have been quick to revolt if the show’s treatment of that part of the story was as cavalier.
Just on the now-conventional alphabetical (rather than non-) order that I take it is the primary subject of Flanders’ book, I was delighted to come across the following tidbit in wikipedia: “As a spa city, Aachen has the right to name itself Bad Aachen, but chooses not to, so it remains on the top of alphabetical lists.”
Perhaps Aachen doesn’t want to be confused with Pynchon’s fictional spa town, Bad Karma.
Incidentally, I’m the author of a book whose index begins with Aachen and ends with Zeus.
@Rodger C.: Points docked for not figuring out how to work zymurgy into your narrative. (By free-association, I note that it seems like a long time since zythophile commented in these here parts.)
an absurd soap opera with only a tenuous connection to reality
It’s a literary form called a secret history, the fictional analogue of The Secret History of the Mongols or Procopius’s Anecdotes, aka Ἀπόκρυφη Ἱστορία. (The latter was lost for more than a millennium, though its existence was known, until it turned up in the Vatican Library and was published for the first time in 1623, so it remained pretty secret.)
The linked WP article lists about 50 secret histories (spoilers!);
1) those that purport to tell what almost happened before it was covered up, like The Day of the Jackal (de Gaulle almost assassinated);
2) those that purport to tell what happened but was hidden or forgotten, like The Three Musketeers (the assassination of the Duke of Buckingham by John Felton);
3) those that, within a fictional world, purport to tell within that world what was hidden or forgotten for the sake of maintaining continuity (one kind of retcon), like “The Problem of the Empty House” or the four main Lensman books, each of which save the first gives the secret history of the previous book.