Recognizing Mende.

This account of a 1998 movie at the California Newsreel site tells a remarkable tale:

The story begins in the early 1930s with Lorenzo Turner, an African American linguist who cataloged more than 3000 names and words of African origin among the Gullah of coastal Georgia and South Carolina. He discovered that some Gullah could recite texts in African languages, including almost certainly the longest, a five-line song he learned from a woman living in a remote Georgia fishing village, Amelia Dawley. Although Amelia did not know the meaning of the syllables in the song, a Sierra Leonean graduate student in the U.S. recognized it as Mende, his native tongue.

These dramatic clues were taken up again in the 1980s by Joseph Opala, an American anthropologist at Sierra Leone’s Fourah Bay College. Studying Bunce Island, an 18th century British slave castle, Opala discovered that it sent many of its captives to Georgia and South Carolina where American rice planters paid a premium for experienced slaves from Africa’s “Rice Coast.” The comparative coherence of this slave community may account for the high degree of African cultural retention among the Gullah. In 1989 Opala helped organize a gala homecoming for a Gullah delegation to their long-lost African sisters and brothers documented in an earlier California Newsreel release, Family Across the Sea.

Opala joined with ethnomusicologist Cynthia Schmidt and Sierra Leonean linguist Tazieff Koroma in an arduous search to see if Amelia Dawley’s song was still remembered anywhere in Sierra Leone. Although the Mende are the largest ethnic group in Sierra Leone, Koroma recognized one word as unique to a dialect spoken only in southern Sierra Leone. On their last day in the area, Cynthia Schmidt discovered a woman, Baindu Jabati, living in the remote interior village of Senehum Ngola, who had preserved a song with strikingly similar lyrics, a dirge performed during a graveside ceremony called Tenjami or “crossing the river.” Her grandmother taught her the song because birth and death rites are women’s responsibilities in Mende culture. At the same time she made the uncanny prediction that there would be a return of lost kinsman and that Baindu would recognize them through this song.

Schmidt and Opala then went to Georgia where they found Amelia Dawley’s daughter, Mary Moran, age 69, who remembered her mother singing the song. Though transformed in plantation culture to a children’s rhyme, there was also continuity since the song was passed down by women on both sides. A reunion between Mary and Baindu had to be postponed because of a devastating rebel war in Sierra Leone which left millions homeless, including Baindu herself. Finally in 1997, Mary Moran and her family could travel and, after a painful visit to Bunce Island, were received with jubilation in Senehum Ngola. The village’s blind, 90 year old chief, Nabi Jah, organized a teijami ceremony for Mary, even though it had been in desuetude since the introduction of Christianity and Islam earlier in the century. Thus Mary’s homecoming became a catalyst for Mende people to rediscover a part of their own past. When Opala asked Nabi Jah why a Mende woman exiled two hundred years ago would have preserved this particular song, he replied that the answer was obvious. “That song would be the most valuable thing she could take. It could connect her to all her ancestors and to their continued blessings.” Then he quoted a Mende proverb, “You know who a person really is by the language they cry in.”

The Language You Cry In shows the significant benefits of multi-disciplinary research. It also is a striking example of scholars working with their informants as colleagues; the “research subjects,” African and American, were not just observed but actively recruited into researching and analyzing their own histories. Events, sometimes national in scope, were organized so that individuals and communities could make new research findings their own as part of a “usable past.” Meaning thus emerged out of the deliberate clash of present and past. As we watch Mary and Baindu reunited in a tearful rendition of this ancient song, we realize how 20th century scholarship and media technology are making their own modest contribution to preserving bonds within the African Diaspora.

The text of the song, with translation, is at the link; via MetaFilter, where there are more links.


  1. David Eddyshaw says

    some Gullah could recite texts in African languages, including almost certainly the longest, a five-line song he learned from a woman living in a remote Georgia fishing village

    Candomblé, in Brazil

    uses Yoruba as its sacred language. I have read somewhere that there were Candomblé priests who could actually speak Yoruba, but that seems to be something of an exaggeration. A lot of Yoruba terms, though.

  2. A beautiful story!

  3. I don’t know Yoruba, but I do know Candomblecistas, and they tell me they have songs in Yoruba (and what they call Kikongo, and other African langugaes, depending on their “nation”). But I think they are more or less fossilised – people are not composing new songs and they don’t know the language as such, though they know the meaning of the songs. But it would not surprise me if there are a few people who have leaned it through study: some are keen to get back to the roots of the religion and culture in Africa, have visited Nigeria, and so on.

  4. A good story indeed, and a kind I would love to hear more often.

    (and perhaps something like this should be done with other communities of descendants of slaves, north of Sahara for example. (I don’t know if such communities can be identified in Russia))

    The “milti-disciplinary” line is…er. Not a good style:) It also took an effort for me to agree that there is more than normal interaction between a researcher and informant here, but yes, there is more.

  5. Dan Milton says

    In the first line of Amelia’s song is the phrase “kambay yah” and in the first line of the translation “come together”.
    “Kumbaya” is known to be from the Gullah, but is generally regarded as the English “come by here”. Now we see what seems to to be the phrase in a Gullah song with no English that I can recognise.
    So is it English or Mende?

  6. Didn’t know that kumbaya has an English reading (but maybe I learned it before I learned enough English to understand “come by here”:-))

  7. John Cowan says

    There is no reasonable doubt that the song “Kumbaya” began life as a song in Sea Islands Creole, with these words corresponding to Standard English “come by here”. How they got into the Mande version of the song is a question.

    Supreme Court Justice Clarence Thomas grew up speaking Sea Islands Creole, and (according to him) because he was mocked for it, he got in the habit of not asking questions, which is why he doesn’t speak much on the bench.

  8. John Cowan says

    (woops, time ran out)

    “When I was 16, I was sitting as the only black kid in my class, and I had grown up speaking a kind of a dialect. It’s called Geechee. Some people call it Gullah now, and people praise it now. But they used to make fun of us back then. It’s not standard English. When I transferred to an all-white school […], I was self-conscious, like we all are. It’s like if we get pimples at 16, or we grow six inches and we’re taller than everybody else, or our feet grow or something; we get self-conscious. And the problem was that I would correct myself midsentence. I was trying to speak standard English. I was thinking in standard English but speaking another language. So I learned that — I just started developing the habit of listening. And it just got to be — I didn’t ask questions in college or law school. And I found that I could learn better just listening. And if I have a question I could ask it later.”

  9. David Eddyshaw says

    @drasvi: It’s A wa kaka that means “everyone come together” in the song (at least a wa is the command “come!”: I know almost no Mende.) I think that kambei ya le’i is the part that means “the grave is not yet.” (ya is “not” and is “yet”, at any rate.) The similarity of sound is just coincidence.

    So “kumbaya”, if from Gullah, is indeed from English lexemes, and nothing at all to do with this song.

  10. David Eddyshaw says

    Yup: searching the Mende Bible app (I can’t find a dictionary) it looks like kamba is “grave.”

  11. David Eddyshaw says

    The song actually seems quite remote from the “Modern Mende” version: while I’m perfectly prepared to believe that Amelia Dawleys’s song did indeed originate in Mende, I wonder just how much the detailed correspondence has been imagined and improved after the event.

    This is hardly surprising: anything like accurate transmission of a song in a completely unknown language (and one extremely unlike either English or Gullah, moreover) over multiple generations seems very unlikely. In particular, the rendering of the bit about graves as “kumbaya” may well have been due to it sounding like “kumbaya” not only to us but to the Gullah transmitters of the song themselves.

    I suspect an element of “Roots”-style embroidery in this story, though certainly not to a Haleyesque degree. No imaginative pseudogriots in this case.

  12. Here’s an interesting fact. The culture (and dialect) of the descendants of former slaves living along the Georgia-South Carolina coast (especially the coastal islands) goes by two different names. In Georgia, it’s called Geechee, in South Carolina, Gullah. A number of cultural organizations have both terms in their names, and there does not appear (so far as I can tell as a total outsider) to be any internal friction among the Gullah-Geechee population over which term to use.

  13. I’m spending time with cousins this weekend, including the cousin with whom I sailed into the Ace Basin, an isolated preserve in the Gullah area near Beaufort, SC. Through some mishaps, we watched our boat float away into a foggy midnight, one of us having missed his shift to watch for the incoming tide. We disagree on who.

    We spent the night worried and ready to grab the most important things if the tide submerged our shellbeach campsite. We’d counted on the missing boat as plan B. Barring a storm surge, at worst we might have been knee deep, and as it turned out, we stayed dry.

    The next morning fishermen from Beaufort found us, and on the way back found our boat, washed across a shell break and into a tidal mudflat. They remembered an urgent appointment with their wives, and I can’t blame them for not further pondering how to free the boat with us.

    A few hours later as we considered our options and waited for the next tide, two crabmen came by checking pots. We tried to explain our predicament but they indicated they only spoke Gullah.

    I’ve always wondered whether there were really people in 1990’s coastal Carolina who only spoke Gullah, or whether this was their way of avoiding having to tell two idiots they needed to solve their mudbound boat themselves.

    Eventually we realized we could use sleeping bag mattress pads as snowshoes to cross the quick-mud and grab the painter. Hours later, enough water trickled into the flat that with great effort we dragged the boat out.

    I’d been reading The Port Royal Experiment, about a Union Army attempt to get freed slaves to work for wages growing the fine Sea Island cotton, which required backbreaking work in the salty mud. I hadn’t been unsympathetic even before, but after our failed attempts to cross the mud, I felt viscerally their distaste for the work.

  14. “sleeping bag mattress pad”

    Oh. In Russian
    turisticheskiy kovrik
    turist – in USSR the word did mean a tourist, but usually it was understood as a ‘hiker’, the guy or a gal in green or brown windbreaker with a backpack (with a tent inside).
    turisticheskiy – the adjective that makes Russian write “touristic” in English (like “touristic part of the city”).
    kovyor “rug, carpet”, kovrik dim.
    Or often penka “foamie” (pena “foam”, penka diminutive, milk skin).

    And whenever I discuss hiking in English I wonder how I’m supposed to translate it.

  15. @DE, it was Dan Milton actually:)

  16. Everything I know about cotton picking (in the contex of Soviet Central Asia) makes me symapthetic. It is hard work indeed.

    Also compare

    Someone went to a math conference in Samarkand and complained that there were no students because everyone but the dean’s daughter was sent to pick cotton. During a bus tour the dean pointed at cotton fields around and proudly said: “look how Uzbek matematicians work!”. The Russian mathematicians was unsympathetic…

  17. @Stephen J: “…they have songs in Yoruba…”

    Deixa a gira girar by Os Tincoãs (1973; sometimes spelled phonetically as “Deixa a gira girá”). The song itself is in Portuguese but is packed with terms from the Afro-Brazilian religious tradition. The name of the realm of happy spirits, Aruanda, is said to be a corruption of Luanda. The names of the gods/spirits are all from Yoruba, it seems: Iansã, Xangô, Iemanjá. “Saravá” is supposedly a mispronunciation of “salve” but is usually associated with Umbanda or Candomblé. “Gira” and “girar” in this context also refer to a ritual in Candomblé or a similar tradition.

    The spoken section in the middle is different. Here it is, according to

    Zambi, rô, Zambi
    Zambi na qua tê sá
    Baquice, baquice, batabaquice de Orixá
    Kylê ibai, toté de maiungá
    Xê cumarô loxê kulundu loxê keto tá sôto
    Atotô, oba!

    I only recognize Orixá here, a god or goddess.

    Candomblé influenced the samba tradition in so many ways… Also, Vinicius de Moraes, bossa nova’s principal lyricist, was an Umbandist and freely inserted Afro-Brazilian elements into his texts.

  18. Cotton-picking is hard work.

    But if I remember, sea island cotton grew particularly well if a layer of fresh marsh mud was added to a field. The freedmen were unwilling to carry out the additional labor of gathering the mud, which would have required them to sink in and wrestle through, and that was what caused the project to fail.

    Thinking about it today, it’s probably not salty mud. Presumably it was from high enough up the tidal area that the marshes mostly filled with river water.

  19. while I’m perfectly prepared to believe that Amelia Dawleys’s song did indeed originate in Mende, I wonder just how much the detailed correspondence has been imagined and improved after the event.

    This was my reaction as well.

  20. David Eddyshaw says

    Africa’s “Rice Coast”

    Rice cultivation is hardly confined to that area of West Africa (cf proto-Oti-Volta *muɹ- “rice.”)

  21. proto-Oti-Volta *muɹ- “rice.”

    Sounds like a Mande loan (I forget the proto-Mande form – something like *malo.)

  22. David Eddyshaw says


    According to WP, “[African rice] was domesticated about 3000 years ago in the inland delta of the Upper Niger River, in what is now Mali”, so that would fit quite neatly. A date for proto-Oti-Volta is pretty much guesswork, but 3000 YBP seems reasonable.

  23. David Eddyshaw says

    In the wider “Central Gur” field, it’s múró in Chakali, which looks either cognate or loaned from a similar source; Kassem has mumunə, which looks as if it ought to be connected in some way, but is generally not very helpful.

    Koromfe mũĩ looks suspiciously like a loan from Mooré mùí, though. (Also Koromfe mũĩfe “grain of rice”, Mooré muiifu.) In Western Oti-Volta, *ɹ -> /j/ is the regular development after a short root vowel.

  24. I did not know about African rice:/ I guess again, “need to go to Ghana to try”?

    Funnily, they again try to explain domestication with Sahara pump. This too was interesting: A 2006 survey showed that a village typically cultivated 25 varieties of rice; an individual household would on average have 14 varieties and grow four per year;[10] this, however, is down from the seven to nine varieties per woman that was average in previous decades. Women, who are traditionally responsible for the seeds, trade them often over long-distance networks.

  25. David Eddyshaw says

    Twi ɛmo, Ewe mᴐli

  26. David Eddyshaw says

    Fulfulde maaroori … the word really gets around.

    (The *muɹ- etymon can hardly go back to proto-Volta-Congo, despite the enticing similarity of the Twi and Ewe forms: apart from anything else, “rice” doesn’t seem to have been around for long enough in those parts.)

  27. I read somewhere, not long ago, about a similar thing in India: enormous numbrers of (Asian) rice varieties, with small growers planting many at the same time. There’s an effort to rescue this diversity from going away.

  28. David Eddyshaw says

    I see that Karlgren’s reconstruction of “rice” for his “Archaic Chinese” was miər. I feel that a Sino-Platonic paper is long overdue on this … does Prof Mair know?

    The implications are tremendous: rice growing clearly spread from Africa to China. Black Athena, nothing.

  29. Sound symbolism.

  30. Sirius.

  31. David Eddyshaw says

    Rice is indeed from Sirius. But it seems that someone has been divulging information for which the Humans are not yet ready.

    We may need to have a word with those Dogon.

  32. Humans eat from saucers.

  33. J.W. Brewer says

    Within the context of New World plantation economies employing slave labor, working cotton was no one’s idea of an easy life but it was not quite as bad (on brutal objective metrics like life expectancy) as working sugarcane. I’m not sure how cotton compared to rice (or for that matter to tobacco). There are parts of the Caribbean where the emancipated slaves and their descendants largely refused to work sugarcane as free labor for any price — or at least any price that the bosses were willing to pay, thus leading to the importation of a replacement labor force and eventually the current ethnic/racial demographics of places like Guyana and Trinidad.

  34. I associate cotton with exploitation in Soviet context.

    As for sugarcane, our sugarcane grows in Cuba.
    Куба, отдай наш хлеб,
    Куба, возьми свой сахар…
    (to the tune of solemn Куба, любовь моя…)

  35. David Eddyshaw says

    Humans eat from saucers

    Although the natural response is horror, I feel that we are not really in a position to cast stones after the regrettable matter of To Serve Man. What can I say? Mistakes were made …

  36. Stu Clayton says

    Off-topically further to the matter of stupid German renderings of film titles: this talk of flying saucers reminded me of a amusing film about small fs and an apartment building in New York. The title was, I thought, something like Miracle on somethingth Street.

    It took me a while to track the film down. The title is “*batteries not included”, which I had never heard of. The Germans had turned this into Das Wunder in der 8. Straße.

    The land of crackers and denkers.

  37. Sounds like fun, and I was in NYC during the filming, though I don’t remember it. Wikipedia:

    Principal photography started in New York in August 1986, but location scouting began almost a year before. “Since the story called for a solitary building amidst rubble,” explained producer Ronald Schwary, “we had to find a vacant lot with burned-out buildings all around it. We finally settled on an actual building on 8th Street between Avenues C and D on New York’s Lower East Side (the building no longer stands, and was probably located on the site of the current Housing Bureau substation, or the building to the east. 40°43′27.33″N 73°58′40.49″W). Production designer Ted Haworth designed a three-sided, four-story tenement facade and oversaw its construction on a location that covered most of a city block. In the name of authenticity, he brought 50 to 60 truckloads of rubble to cover the once vacant lot. It was so remarkably realistic that the Sanitation Department came by and took away prop garbage one morning, potential customers stopped by to eat in the diner, and the business agent for the Plumber’s Local of New York visited, demanding to know why there wasn’t a permit down at City Hall for the construction.”

  38. I don’t know whether the French imitated the Germans or vice versa, but they call it Miracle sur la 8e rue (though the québécois use the literal Piles non comprises). The Romanians call it Prieteni de departe ‘Friends from afar.’

  39. I saw “batteries included” numerous times used metaphorically. I guess advetisments of software etc.
    I did not understand it.

    Not because it is difficult to understand it: now I thought about it… I guess it is what is written with or without “not” on boxes of electronic devices sold for 8.99. They must write something in Russian too, I just don’t know what.

    But when used metaphorically, I understood it as “I’m trying to write something funny and cool, and this line is idiomatic”.
    If so, “батарейки в комплект не входят” (as I said, I’m not sure it is what they usually write, but this is going to be understood… in the context of saucers, of course:)).

  40. Stu Clayton says

    Gift not included

    Ir would have been so easy and natural to render the title as Batterien Nicht Enthalten. To be fair, the translators may be subject to cognitive and legal constraints.

  41. David Marjanović says

    Definitely not with capital letters for non-nouns!

    “I am the terror that flaps in the night.
    I am the batteries that are not included.
    I am Darkwing Duck.”

  42. Stu Clayton says

    You are suggesting that carelessness is not easy and natural ?

  43. Trond Engen says

    Mali is “rice country”, obviously.

  44. Stu Clayton says

    Sparky and I know a Basenji dog whose name is Mali. Her owner is a fan of reggae, as he told me. I am confident that phonetics buffs here can reverse-engineer his pronunciation to discover the musician meant.

  45. David Eddyshaw says

    Jimmy Cliff?

  46. “Inherits its name from the Mali Empire, in turn from Mandinka or Dyula mali (“hippopotamus”).”

    (Wiktionary, Mali)

    (the quotation and the link are accurate, but I'm kidding )

  47. David Eddyshaw says


    Unfortunately, this tells you rather more about the reliability of Wiktionary than the etymology of “Mali.”

  48. Still I like it: “Obviously it’s rice!” “no, it’s a hippo”

  49. David Eddyshaw says

    Well, it is a natural confusion.

  50. John Cowan says

    This article “Hippos Threaten Rice Production” doesn’t say that hippos actually eat the rice in the rice fields they damage; for all I can tell, they may just trample it. But at any rate there is an association between hippos and rice.

  51. DE, both grow in water…

    (though, I realised that Mandinka can be interpreted as a Russian word: singulative diminutive of “vulva”, cf. trava “grass”, travinka “a blade of grass”. Also with slezá, slyózy “a tear, tears'”, slezinka – here it is just a diminutive. I’m not sure if this has to do with rice and hippos).

  52. David Marjanović says

    Hippos eat grass, rice is a grass (well, two)…

  53. David Eddyshaw says

    Moreover, the Dagbani word for “grass” is mɔri.

    Coincidence? I think not.

    (Oddly, Dagbani hasn’t got the Oti-Volta etymon for “rice”: it uses shinkaafa, nicked from Hausa. Mampruli’s the same.)

  54. Trond Engen says

    I hope it’s obvious that my ‘obviously’ did not mean “obviously” in its most obviously obvious sense.

    I have no idea about derivations, in Mande or elsewhere, but is there a chance that rice before cultivation was “hippo grass”? Or that both words are derived from something like “swamp”?

  55. David Eddyshaw says

    Apparently in Bambara, “hippo” is màli and “rice” is màlo.

  56. Meanwhile, Bambara Mali in Wiktionary (not to be confused with màli) says this:
    I think I have already seen N’Ko_script, but not Adlam_script.

  57. David Eddyshaw says

    I didn’t know of Adlam either. It looks rather pretty. It’s a pity they’ve carried over the vile Latin habit of not marking contrastive vowel length (although I see that there is at least a diacritic for it.)

    It seems that it’s meant specifically for Fulfulde, so no problem with tones.

  58. “Closing their eyes and drawing lines” reminded me abotu bark beetle.

    Sadly, I can’t find a picture (in google images) with what I usually see on twigs in Moscow region, but it is something you really want to read.
    Instead google offers much more regular and much less alien writing-like patterns.

  59. ktschwarz says

    Adlam previously at Language Hat. But if this script is specifically for Fulani, why does Wiktionary claim 𞤃𞤢𞥄𞤤𞤭 is also a written word in Bambara? Are they just confused?

  60. David Eddyshaw says

    I was wondering that myself. I suspect that they are in fact confused: though I suppose there is no law against writing Bambara in Adlam if you really want to. But this is a proper name, after all, and it seems likely that they have just carelessly taken the identical-but-for-tone Fulfulde* form as being an alternative way of writing the Bambara.

    I see that I had simply forgotten about Adlam, rather than never having come across it at all. Pity: it is rather pretty. The project of writing Ga and Dangme in pretend-hieroglyphic script that JC mentions here

    seems to reflect a similar feeling that for your language to be a real language it must have its own completely unique script. (Not unknown in Europe, either: one thinks of Irish …)

    * Pulaar, I suppose, in that part of the world. But I am an Easterner, and consequently reserve the right to call the language by its True Name.

  61. John Cowan says

    India seems to be a bastion of such scripts, though I don’t know if this reflects cultural borrowing or separate development.

  62. David Eddyshaw says

    Glagolitic is another European example. And, of course, there are Georgian and Armenian (and Caucasian Albanian.)

    In fact, in Europe it’s only the benighted Westerners who seem to feel “if it was good enough for Julius Caesar, it’s good enough for you, barbarian peasant!”

    The Islamic world also seems to be committed to the idea that there is only One True Alphabet.

    But apart from the the filioque gang and the Muslims, nobody else seems very hung up on the issue.

    Let a thousand writing systems bloom!

  63. But I think they are more or less fossilised – people are not composing new songs and they don’t know the language as such, though they know the meaning of the songs. But it would not surprise me if there are a few people who have leaned it through study: some are keen to get back to the roots of the religion and culture in Africa, have visited Nigeria, and so on.

    my (non-initiate) understanding, both from reading and my time in the capoeira world, is that the liturgical music repertoire is fairly fixed (as in many traditions). but there has been pretty regular back-and-forth travel and migration between brazil and yoruba-speaking west africa by free black brazilians since the late 19thC, as opposed to the kind of extended interruption of contact that there was in most other afro-atlantic practices. so there have certainly been active yoruba speakers in the terreiros, continuously, since before emancipation to the present.

    there is also a de-syncretist movement within afro-brazilian practice that sees itself as restoring Proper Yoruba Religion. i don’t know whether it thinks of itself as part of candomblé, as part of the larger candomblé/umbanda complex, or as an entirely separate thing, but my understanding is that there’s at least a degree of mutual hostility with the long-established terreiros. from what i’ve seen, it has its own yoruba liturgical texts/songs, but i don’t know if they’re adopted from current west african yoruba practices or original to the brazilian groups.

  64. rozele: That’s interesting. It makes me wonder whether some aspects of Brazilian Yoruba religion remained more conservative than anything nowadays practiced anywhere in the old country.

  65. i wouldn’t be surprised if there were different aspects of yoruba practice as of the 17th/18thC that are retained in each branch (candomblé, santería / regla de ocha, vodoun, west african yoruba practice, etc) but not in the others. i expect there are interesting cross-oceanic comparative studies out there!

  66. I wonder if a Yoruba speaker would recognize the lines I quoted above (from the Os Tincoãs song) as corrupted Yoruba or if it would sound like complete gibberish to her.

  67. J.W. Brewer says

    The thing about the multitudinous different Brahmic scripts in India (where the idea that having your own “unique” script shows that you have a Proper Language seems ubiquitous) is that they seem to be variations on a common theme akin to thinking that significantly different fonts of what we think of as the “same” Latin/ASCII alphabet are different scripts (so each European language could have its own Proper Font). Obviously in Former Times there was (leaving out Ireland as a single eccentric outlier) the whole Fraktur v. Antiqua thing within the Latin-scripted world, but I don’t think it got more specific than that in a way that was factionalized as opposed to merely random/elegant variation – imagine if Nynorsk advocates had insisted on a new nationalistic font different enough than what had been the Danish standard that with some practice you could distinguish it at a distance.

    Although is it also true that historically each Semitic language that had a literate class had its own script (albeit one related historically to the others with varying degrees of transparency)?

  68. John Cowan says

    Pretty much. On the Unicore [sic] mailing list, I used to refer to them as 22CWSAs (i.e. “22-character West Semitic abjads”). There was some debate whether they were really script differences or just font differences.

  69. David Eddyshaw says

    is it also true that historically each Semitic language that had a literate class had its own script?

    In fact, the so-called “Hebrew” alphabet* is specifically an Aramaic alphabet.

    The original Hebrew alphabet fell out of use after the Exile, and since then only shows up as a piece of nationalist nostalgia (e.g. on things like coinage.)

    * I refuse to use the horrid term abjad. No alphabet represents all the contrasts of the spoken language: it’s all just a question of degree.

  70. The original Hebrew alphabet fell out of use after the Exile, and since then only shows up as a piece of nationalist nostalgia

    The Samaritan script is close enough.

    Some later Hebrew manuscripts persisted in using the old Hebrew script for the tetragrammaton.

  71. January First-of-May says

    is that they seem to be variations on a common theme

    I know that the assorted para-Greek alphabets of the ancient Mediterranean certainly feel like variations on a common theme to (very-non-expert) me. Lycian, Lydian, Phrygian, Etruscan… early Latin probably goes in the same category.
    (But – looking at the Wikipedia descriptions, at least – Carian and Sidetic are very much unlike Greek. I don’t recall offhand if the Hispanic scripts are even derived from Greek at all.)

    Which does make distinguishing them feel a bit circular sometimes. Do we only call Pisidian a separate script because it’s a different language? If so, does that mean North Picene is written in North Picene script?

    (It probably doesn’t help that variations within a single “script” were often nearly as much as between “scripts”, if not more so, and Greek in particular had extreme variations between regions.)

    I suppose in more modern times Cyrillic and Coptic are examples of the same thing happening (with Old Permic as the equivalent of Carian).

  72. David Eddyshaw says

    We may be looking at this the wrong way round: it seems obvious to us that you can write a language with a script developed for a different language (even if the match may be unsatisfactory to a greater or lesser degree); but it may not have been at all obvious to most people in the past*. Cases where that has happened may be the ones needing explanation in premodern times: the explanations might lie in a higher-than-usual level of linguistic understanding in the relevant culture; or in pure historical accidents, like people being comparatively used to writing their own names in a foreign literary tongue, forming a base for perceptive individuals to start writing other vernacular words too, and eventually even whole texts.

    * I mean, it’s not obvious, say, that English /t/ should be written with the same symbol as French /t/: the sounds are, after all, quite different. It only seems obvious because we’re so used to it. And doing this requires abstraction across languages, which is quite a bit harder (and more error-prone) than identifying the phonemes of your own language, where you’re helped by the fact that all the allophones sound exactly the same to you anyway.

  73. J.W. Brewer says

    With the English/French example, it’s not particularly baffling because the English scribal class from the Norman Invasion forward was generally familiar with French (and before that had been familiar with Latin) and was thus familiar with the French spellings of French-origin loanwords, and what English sounds corresponded systematically to what French sounds (of course French pronunciation may have been rather different back then?) as loanwords had their pronunciation Anglicized.

    Presumably in the Latin-scripted parts of Europe, the usual historical path was that Latin literacy (in a fairly small slice of the population) generally preceded vernacular literacy, so early written evidence of the vernacular is more or less transcribed into a Latin-script approximation, and then eventually those approximations became standardized. The subsequent drive for uniformity, whereby English orthography lost such lovely extra glyphs as edh and thorn and wynn and yogh, is perhaps a somewhat different story.

  74. David Marjanović says

    In fact, the so-called “Hebrew” alphabet* is specifically an Aramaic alphabet.

    And so is the so-called “Arabic” alphabet – though most of the cursivization happened in the Nabatean kingdom, where Aramaic was written but not natively spoken.

    But – looking at the Wikipedia descriptions, at least – Carian and Sidetic are very much unlike Greek.

    Carian has been hypothesized to be descended from a Greek cursive that was later “remonumentalized”. A bit like Glagolitic, but more so.

    I don’t recall offhand if the Hispanic scripts are even derived from Greek at all.

    They’re straight from Phoenician, with the letters for obstruents reinterpreted as syllables and ‘ayin used not for o, but for e, Yiddish-style. …IIRC.

    Do we only call Pisidian a separate script

    I thought we didn’t?

    where you’re helped by the fact that all the allophones sound exactly the same to you anyway

    [x] and [ç] have never sounded exactly the same, or almost exactly the same, to me or apparently to a lot of other other people. They do, however, sound like variants of “the same thing” that are conditioned by surrounding sounds.

    (The picture is not quite as simple as “[x] after central & back vowels, [ç] elsewhere”. Both of them are actually ranges; which exact sound precedes matters to some degree*, and the following sound, or lack thereof, is not completely irrelevant either. The trick is that the middle of the total range, which you’d expect to occur after [ɛ] to [æ], is avoided and shifted toward [ç]. – I’m omitting [χ] from consideration because it’s absent from the kinds of German I grew up hearing. In much of Germany it is what follows all central & back vowels except [u] and maybe [o].)

    * Example I figured out when I was little: the [ç] in Milch is a little bit fronter than that in Mönch

    The subsequent drive for uniformity, whereby English orthography lost such lovely extra glyphs as edh and thorn and wynn and yogh, is perhaps a somewhat different story.

    In this case, buying type for printing from the Netherlands is apparently to blame for the loss of þ.

  75. DM: I asked you here once if German /x/ is simply as front as the preceding vowel (ich vs. Aachen vs. hoch) and you said it wasn’t that simple. Now I can’t find that comment. I wish the official LH indexer would hurry up.

  76. ktschwarz says

    Are you thinking of this thread?

    Y: David M., how would you narrowly transcribe German ach, Aachen, and hoch? Are the velars pronounced exactly the same?

    David M.: For me, yes, except for length: [axː], [ˈaːxːŋ̩], [hoːx]. But I’m sure you’re referring to more northern accents …

  77. Maybe! Or maybe we’d had another similar exchange, having both forgotten one. This happens here all the time.

  78. @David Marjanović: Thorns in English had been on the wane for centuries before the invention of movable type. That might have been what absolutely finished them off, but if the character hadn’t been moribund, printers couldn’t have gotten away with using type cases that didn’t include them.

  79. John Cowan says

    The use of y for þ might have survived, however. In his paper on the sorting of þ, Michael Everson quotes from section 6.5 of the introduction to A linguistic atlas of late mediaeval English: “By the later Middle Ages, insular practice was regionally coherent: south of a line running roughly from the Mersey to the Wash, but excluding much of East Anglia, ‘þ’ and ‘y’ were represented by the same (usually y-like) symbol.” This explains why the atlas transcribes a fairly clear MS reading mþkþll as mykyll ‘mickle, great’; the latter is familiar, the former will serve only to baffle and confuse.

  80. David Marjanović says

    Actually, at the level of detail where Mönch and Milch don’t have exactly the same [ç], ach/Aachen and hoch don’t have exactly the same [x] either; the one in hoch starts out a bit backer and rounder. But not as much as the [oː] is. It’s a subtle influence, not full assimilation.

  81. ktschwarz says

    American rice planters paid a premium for experienced slaves from Africa’s “Rice Coast.”

    Language Log has a post today about that history and the application of glottochronology to it.

Speak Your Mind