David Eddyshaw has been for some years now praising the work of Kofi Yakpo, e.g., here in 2019:
Talking of English-lexifier Atlantic creoles (we were, you know, you just didn’t notice), a kind person got me Kofi Yakpo’s A Grammar of Pichi for my birthday. [Fernando Po creole.] Best account I’ve yet come across of one of those languages. At 500+ entirely unpadded pages, it’s a sort of counterargument in itself to John McWhorter’s standing views about creole exceptionalism. It’s got lexical tone! It’s got a distinct narrative perfective! What more could you want?
Now, thanks to a Facebook post by Slavomír Čéplö (bulbul), I have learned about his background; it’s quite a story:
Twice in his life, Kofi Yakpo has made a name for himself as a linguist: Once as a rapper in the German hip-hop band, Advanced Chemistry, where his stage name was “Linguist”. The 1992 single “Fremd im eigenen Land” (Foreigner in my own country) made the band famous whilst his academic career only began shortly before his 40th birthday.
Since 2013, Yakpo has been teaching linguistics at the University of Hong Kong und conducting research into Afro-Caribbean Creole languages: languages that develop when two or more languages converge and form a new one. This kind of hybridisation emerged during the colonial era, the linguistics professor explains, often under duress. And although there are nearly 200 million speakers of Creole languages worldwide, unlike European languages, so far, they have often not been studied sufficiently.
The fact that he has this second career as a researcher at all, says Yakpo, is not just a result of his huge interest in languages but also because of his hip-hop outlook. “As hip-hoppers, our attitude was: I am large. We were always brimming with confidence.” At the time of our video conference, Yakpo is in Nairobi, Kenya, where he is exploring the linguistic variants of Swahili.
The turning point, that is, the moment he decided to focus exclusively on research, came in 2008. He was working for Thilo Hoppe, then a member of the Green Party in the Bundestag, as a political advisor on food security in Africa. “I was living a sort of double life,” he says and laughs. One part was his work at the Bundestag, the other his research in linguistics. Alongside that, he completed his doctorate: the first complete grammar of Pichi, an Afro-Caribbean Creole language. When it was finished, he showed it to colleagues at the Bundestag, Yakpo relates. “One person leafed through and said: I don’t understand it at all. That’s totally different from what you do here. Either you’re an impostor or you have a split personality.” That was when Yakpo realised that he had to plump for linguistics, his true passion. […]
Yakpo was born in Holzminden in Lower Saxony but spent his childhood in Ghana. German is his native tongue. In Ghana he came into contact with his father’s language, Ewe, for the first time. When he was ten, his family moved back to Germany, to Heidelberg, where he taught himself Ewe grammar using an old schoolbook from his parents’ bookshelves. He admits to having been a language nerd. “My first passion was Latin.” Soon he started French and, at 15, was borrowing books on languages from the university library: grammars of Fijian, Tok Pisin, the national language of Papua New Guinea, or Yoruba, one of the three principal languages of Nigeria. “I was under the illusion I could learn any language in no time,” says Yakpo. After doing his civilian service, he enrolled at the University of Cologne to study linguistics. […]
Whilst working as a Humboldt Research Fellow in Berlin, he recently explored another path: “I wanted to know how languages develop when they are not standardised by written forms or state authorities.” After all, strict language standardisation is ultimately a European concept. In Europe, language is seen as something that is not supposed to change. “In West Africa, it’s different,” says Yakpo. “Variation is quite normal.”
There’s much more at the link; I have to say I’m glad he made the career change. I suspect there are a lot more good rappers than there are good linguists doing that kind of work.
I looked at his publication list on Zenoodo. So much good stuff! Here’s one, Creole prosodic systems are areal, not simple:
Take that, McWhorter!
French, too? I once read it’s happened in a few English-based creoles, but not in any French-based ones because French stress is too regular and unimportant to be interpreted as tone. Apparently somebody did manage that feat.
Anyway, here’s a direct link to the paper; all Frontiers journals are in open access. Present company might also enjoy the rest of the Research Topic ( = paper-free special issue) “Simple and simplified languages”, though none of the other papers seem to be about creoles.
as a rapper in the German hip-hop band, Advanced Chemistry
Trumps even Geoffrey Pullum.
It’s not only stress that can get reinterpreted in terms of tone in creolisation; tone itself can get reinterpreted in terms of a substrate tone system:
https://scholarworks.iu.edu/dspace/bitstream/handle/2022/21201/Tone%20Splitting_Gwandara%20ethnohistory.pdf?sequence=1
I suspect that Linguist the rapper adopted his name in the Ghanaian sense of okyeame “chief’s spokesman, herald.” The Kusaal equivalent nɔdi’es is used in the Bible translation for “prophet.”
I question David E.’s ventures into music criticism. But so others can make up their mind:
1. Here’s a clip of the 1992 number that was apparently the (in Germany) breakthrough for Herr Doktor Yakpo’s Teutonic rap act: https://www.youtube.com/watch?v=TJKJKb2XjOs
2. Here’s a clip of a 1967 performance (on German tv, as it happens) by the ensemble that included the young Geoff Pullum (on Hammond organ and backing vocals – he gets comparatively little camera time but does get some). https://www.youtube.com/watch?v=TJKJKb2XjOs
The similarity is of course that in both instances you have a buncha Europeans earnestly trying to imitate or adapt or appropriate a vernacular black American music style that was popular at the time. #1 is definitely wordier lyrically and may be superior lyrically if you like that sort of thing, but #2 is more interesting rhythmically, with more swing and groove, not to mention melody and harmony. So unless you think that the cultural function of music is to be minimalist background to a polemical-lyrical-content delivery system, I think the advantage goes to #2. (This is not merely a genre thing – some of the late ’80’s U.S. hip hop acts that Advanced Chemistry claimed to be influenced by were of much greater interest qua music.)
Does McWhorter expect creoles not to have tones?
@JWB: Both links are to Fleetwood Mac. Is that the new version of Rick-rolling?
Hans: No, a cut and paste mishap I did not detect before posting (I had intentionally posted that FM link elsewhere on the internet). Thanks for alerting me to it.
The links intended above were:
https://www.youtube.com/watch?v=HhsEcPiRKrU = Advanced Chemistry (feat. K. Yakpo) in ’92
https://www.youtube.com/watch?v=jrzLuweNMg8 = Geno Washington & the Ram Jam Band (feat. G.K. Pullum) in ’67
the cultural function of music is to be minimalist background to a polemical-lyrical-content delivery system
But of course.
You’re not secretly some sort of running-dog imperialist bourgeois, are you? I mean, I know that on the Internet nobody knows you’re a running dog …
Also, hip-hop is cooler than soul. It just is.
Does McWhorter expect creoles not to have tones?
He certainly used to, but I think he’s added some epicycles more recently, to deal with the fact that they obviously can have tones: they have tone, but only in grammatical particles, and/or only in vocabulary borrowed from the substrate, etc. Between that and adjusting the boundary between “real” creoles and “semi” creoles you can perhaps save the phenomena, with enough effort.
EDIT: Actually, I think I’m traducing McWhorter, on reflection: his nice Saramaccan grammar certainly has lots about tone and tone sandhi IIRC (unfortunately I’m hundreds of miles away from my own copy at present.) It’s probably more to do with alleged simplicity of tone systems (though I must say that Hausa has a tone system about as simple as you can get and still be a tone system.)
The Bundestag part bothers me. It’s a very different job, but when a political advisor on food security is a serious researcher (perhaps not a linguist), it must not be too bad…
I remembered where I saw someone grumbling at McWhorther’s tone. here p.32. But it’s 2003…
Somewhat more fresh comment by McWhorter on languagelog: “To wit: I hypothesize, based on how such matters have come out in cases worldwide, that if Thai plantations had been worked by Chinese and Hmong slaves, the resultant creole would still not have used tones to distinguish words or encode grammar.” (LL)
The Nubi examples in my first link must depend on whether pitch accent and tone are considered the same…
Thanks, drasvi. Maybe I wasn’t traducing McW after all. I can’t remember how he squares this particular circle in Saramaccan, unfortunately. He does seem to have an impressive knack for explaining away data that don’t fit his preconceptions about creoles.
Incidentally, just before that McW says
Sadly, he gives no examples. I can’t think of any myself. He perhaps means the Zulu-based creole mentioned in the other link you give. On the other hand, I can think of several counterexamples. I note that he says “often shed using the tones”; “often” is such a useful word …
In my first link:
McWhorter,
1998 ‘Identifying the creole prototype: vindicating a typological class’
2000 ‘Defining “creole” as a synchronic term’
2003 The power of Babel. A natural history of language
I’ll take a look, but it is not fair to expect his initial proposal for the “creole prototype” to be fully accurate. He must have updated his views.
The thing is, his “Creole Prototype” is entirely empirical, and thus changes with additional data, but pretends to come from some fundamental reasoning.
There is an issue with our sample:( Not enough Khoi-San speakers who shifted to Tlingit.
And I don’t expect anyone to make successful predictions about languages from “fundamental reasoning”:( To get some things right is already a feat.
if a language is descended from a pidgin and is young it will
Epicycle.
Also false – quite egregiously so.
I think McWhorter’s problem is that deep down he thinks lexical tones are in some way “unnatural”, and that tone languages are doing something very strange and cross-linguistically very “marked.” To say the least, this view is not supported by the data …
I suspect this is just Eurocentricism (ironically enough.)
There are arguably-tonal languages (Norwegian, IIRC?) even in Europe AFAIK.
(And I suspect that SAE movable-stress setups, where the stress position can distinguish words and/or encode grammar, would have been classified as tonal by a linguist from some place where tonal languages are the default; for all I know it might well be a Eurocentric view to not consider them as such.)
But yes, most of Africa has tones, most of Asia has tones, IIRC much of the Americas has tones… dunno about New Guinea, hadn’t looked it up. Can’t think of any tonal languages in Australia.
They do seem to be uncommon (absent?) in Australia, but then Australia is in some ways one huge Sprachbund when it comes to phonology.
None of the New Guinea languages I know anything about is tonal; though my knowledge of that region is extremely slight, it certainly doesn’t seem to be anything like Southeast Asia or Africa.
Still, though the feature looks pretty (macro)regional, it can hardly be claimed to be either a cross-linguistic rarity or confined to only a few language families. Moreover, previously non-tonal languages have often (if I may be so McWhortery) independently developed lexical tone. (So how hard can it be, really?)
McW has a habit of drawing his examples from regions where lexical tone is nigh-universal, which makes his contention that lexical tone is a highly marked feature even more peculiar. If he really has lots of specific examples of African creoles lacking tone which are not in fact based on inadequate analyses of their actual tonal structure I will be happy to eat my words (but I’m not holding my breath.)
There are tonal languages in New Guinea. In fact there’s every weird phonological system you can imagine in New Guinea.
I am both pleased and relieved to hear it.
(However, tones are not weird. Let us be clear about that …)
I stand corrected. “There’s every phonological system you can imagine in New Guinea.”
I’m not sure why it would be “ironic” for McWhorter* to suffer from “Eurocentrism.” He’s an American. To think that the extent to which American culture (and thus the worldviews of randomly-selected Americans) is largely descended from European culture would depend to any significant degree on the skin color of the particular American in question would be a gross error, just as an empirical matter.
*As I may have said before, I’ve never met the man but we are the same age, grew up less than 30 miles apart, and have various mutual friends and acquaintances. Way back when we were both 13 years old I wrote a little research report for school about the variousness of the languages used in Surinam, as we then spelled it. But then he grew up to be paid actual cash money to do fieldwork on certain of the languages of Suriname (as it had by then become) and I didn’t.
I was thinking of McWhorter’s linguistic focus, not his skin colour. It seems ironic to me that a creolist who sees creoles everywhere and reckons that all of them are creole-y in similar ways, should fall into such a daft error as linguistic Eurocentrism.
On the other hand, I may be wrong, in which case the irony evaporates.
“is largely descended from European culture”
“descended” as opposed to “Europeans earnestly trying to imitate or adapt or appropriate” :-E
@drasvi: American culture is largely descended from European culture yet has also innovated in various ways which still-echt-Europeans may or may not be able to imitate w/o appearing clueless or naff.
@David E.: I apologize (-ise?) for having perhaps misconstrued your sense of irony. OTOH, the thing about creoles globally (at least the ones creolists tend to study) is that they pretty much invariably have a European-origin language as the lexifier, so why shouldn’t a creolist take Europe as the obvious point of departure?
Back to popular music: it may be quite difficult due to the plethora of candidates to determine which popular-in-Britain-in-the-Seventies song was the *most* bourgeois, but this number came to mind when reading David E.’s comment above, not least because, as chronological coincidence (or is it?) would have it, it was released as a single (in a shorter edit than this LP version) the very same month (December ’75) that the Sex Pistols played their first gig and is perhaps an excellent illustration of the various features of the ancien regime that the angry young radicals of the day found unsatisfactory: https://www.youtube.com/watch?v=hTdikCon128
they pretty much invariably have a European-origin language as the lexifier
No; KiNubi (one of several Arabic-lexifier creoles) and Gwandara (a Hausa creole) immediately come to mind, along with Sango (Northern Ngbandi) and Lingala and other Congo examples. To say nothing of Chinook Jargon or Hiri Motu. Though I concede that a disproportionate amount of work has been done on the Euro-lexifier kinds. I suspect that that is part of the problem.
Wiktionary Adj. Creole
1. Designating a creolized language. [from 18th c.]
creolize
1. (linguistics) To cause a pidgin language rapidly expanding in vocabulary and grammatical rules to become ultimately a creole.
2. To render an imported object ‘localised’; to produce variations which give an object a regional flavour.
Have you ever caused a pidgin language rapidly expanding in vocabulary and grammatical rules to become ultimately a creole?
It’s silly indeed. Creolize is only used in the passive.
(P.S. It’s too bad you can’t say in English “almost only”. It would be a fine and useful expression.)
expanding a little on DE:
seems to me that it would take a ton of evidence to ascribe the similarities among the european-language-lexified creoles to anything about language contact as such, rather than to the specific social landscapes of european colonialism in which they emerged. so i’m inclined to doubt there’s much reason to trust work on creoles that’s primarily based on that constellation of lects to have any conclusions that hold elsewhere.
But we need some definition of “creole”.
And we do start from a prototype. The langauges of creoles in European colonies (maybe using “prototype” for McWhorter’s Platonic creole was an unfortunate choice).
The next stage is our desire to use the word in more general sense. There are certain properties of the original creoles that made linguists want to use the word so, that inspired them. At this stage we and our set of properties depend on our original creoles. And the set is subjective. It is like intorducing a word “masha” to designate all girls “similar” to a certain girl Masha. Such a notion may seem totally arbitrary to anyone who does not know Masha or who is impressed by a different quality of Masha than what impressed me. In other words, my “masha” bears an imprint of my and Masha’s individualities.
We are not satisfied with this and want to define “creole” in a more objective way. If we base our definition on one elementary property (historical, lingusitical, e.g. “simplicity”, sociolinguistical) we already may run into problems. Our historical definition (a creole is a language that arose in a certain way) can be based on false understanding of history of our original creoles.
And we don’t want a class of all languages defined by a simple property. We hope to discover a bundle of correlated properties. So we form hypotheses in the form : “all creoles [languages with a property A] also have a property B” (so the bundle exist and is not smaller than the set of creoles) and “there is not easily definable property C that entails A and also is correlated with B” (so the “natural” bundle is not larger and our desciption can’t be simplified).
All of this not easy. But if we base our definition on a complex property (say: an outcome of massive language shift and a simple language) we have more problems. What if this specific combination is not any more interesting than any other random combination?
So what I want to say: I don’t know what is creole, but I think we require several things (that include language shift and change) and there are several more suspicious expectations like “having descended from a pidgin” (which is based on a hypothesis about the original creoles) or “is simple”. We based the definition on a complex property and we faced all these issues. And as we trying to solve everything and demonstrate that our creole is a thing, we of course keep looking back on our prototypes: languages of creoles in european colonies. It is natural.
I don’t know if Gwandara is a creole. If Newman is right, it is “a recent off-shot of Hausa that underwent a massive contact-induced langauge change” – is this always a creole?
I read McWhorter about tones. It does make me want to ironise:/
It does seem that for him tones are a language’s second floor for some reason (I mean Russian second floor). Derived rather then primitive. I don’t understand why. His three features are “1. Inflectional affixation 2. Tone 3. Derivational noncompositionality”. 1 and 3 are seen (not “by him”, by many) as parts of some cycle. Presumably the assumption is that pidgins are made up of “words” and touch this cycle on a specific point.
Tones for him too arise from something else. But then:
Ouch!!
But he has been studing Mandarin since then. I would love to see what he says about tones now.
Why should Mandarin change his opinion? It’s not a creole.
His thinking seems to be “there are languages without tone, ergo tone is not a necessary part of language, ergo they’re not part of UG, ergo they always must be a later-developed superstructure”.
There are several assumptions in that reasoning that don’t necessarily hold water, including of course assuming UG.
Because DE seems to be right: for him tone is marked. His argument is: there are languages with and without tone ergo tone is unnecessary [an unnecessary complication that pidgins don’t need]. So yes, markedness. Maybe good knowlege of a tonal language will change something….
I learned UG 10 years ago when I drank with students of mechanico-mathematical faculty of Moscow State University. They were throwing their lecture notes on “numerical methods” into the fire (used them for kindling) and said it is “UG” – unýloye govnó.
“there are languages without tone, ergo tone is not a necessary part of language, ergo they’re not part of UG, ergo they always must be a later-developed superstructure”.
@Hans, sorry, I repeated what you wrote:-/ Yes, he beleives that tones develop;
(McWhorter, J. 2000. ‘Defining “creole” as a synchronic term’, in Neumann-Holzschuh and Schneider eds., Degrees of restructuring in Creole languages, Benjamins, 2000, 85–123.)
Fully-established lexical tone (not just tone “encoding morphosyntactic distinctions”) is securely reconstructable to Proto-Volta-Congo. (I know because I can provide regular Proto-Bantu/Proto-Oti-Volta tone correspondences.) There is no reason to suppose that it is the result of phonological erosion (apart from a priori assumptions about how languages must be.) I wish McW would leave African linguistics alone. He has no idea what he’s talking about in this domain.
I may have expressed myself unclearly. I did not mean to say that there are no extant (or historically-attested) creoles with a non-European lexifier, just that most creolistics scholarship of which I am (as very much a non-specialist …) aware tends to treat the European-lexifier variety as the prototypical/paradigm example. There are of course problems with that approach as David E. notes, in that you won’t know whether an observation/pattern you find in those creoles will generalize to creoles with non-European lexifiers unless you actually do the empirical work to investigate.
To rozele’s point, I think the European-lexifier ones arose in a range of different social/historical circumstances (some involving more direct subjugation via conquest and/or enslavement, others more in the context of trading-for-mutual-advantage), and I’m not sure that the non-European-lexifier ones arose in circumstances outside that range.
Now, even if one accepts that primordial human language didn’t have tones and that tones historically always evolved from erosion / neutralization of other phonological features (I assume DE would not accept even that claim), it doesn’t follow that pidgins cannot have tone, except if you believe that each pidgin structurally returns to that primordial state. I know that some generativist creolists held views close to that belief, I don’t know whether McW does. If you don’t accept that, there are at least two ways pidgins can have lexical tones – by inheriting them from lexifier languages and by reinterpreting other features of lexifier languages as tone distinctions. That gives you at least lexical tone; whether a pidgin also can have morphological tone would be parallel to the question whether pidgins do inherit or develop other types of morphology.
Now, even if one accepts that primordial human language didn’t have tones and that tones historically always evolved from erosion / neutralization of other phonological features (I assume DE would not accept even that claim)
And why should he? It’s the kind of thing that would only occur to a speaker of a language without tones. Presumably if historical linguistics had first arisen among speakers of Chinese or Kusaal, it would have been assumed as a matter of course that primordial human language had tones.
there are languages without tone, ergo tone is not a necessary part of language, ergo they’re not part of UG
As there are languages without any stop consonants at any particular point of articulation you care to specify, languages without any nasals, languages without any fricatives, and languages without any liquids or approximants, it follows that no consonant is part of UG. There might be two vowels, though, so our forebears may have spoken in binary.
@DE, you misundestood: he means the grammatical role is due to erosion in Bantu.
The grammatical role is important for him: a pidgin is unlikely to inherit it.
I found an explanation of why he speaks about tones.
@de
There are many people today who conduct entire conversations in grunts. Is it possible these people are members of a Secret Society dedicated to the preservation of the proto-human language?
What non-creole language has neither inflectional morphology nor lexically or grammatically contrastive tone?
Khmer. Ask me another … (though Phnom Penh dialects have apparently developed tonal distinctions secondarily.)
Confucius also might have reacted negatively to the idea that his native tongue was a creole (though I have a feeling that this has actually been suggested by McWhorter-like aficionados of circular arguments.)
@PP:
Ug. Ug ug. Ug.
More than that, I am not at liberty to disclose.
He could tell you, but then he’d have to bash you over the head with a club.
he means the grammatical role is due to erosion in Bantu
This strikes me as every bit as problematic an idea as the idea that all lexical tone arose from erosion. Sure, there are cases where tones have been left “floating” after their segmental support has disappeared: indeed this is a very common tonal phenomenon, especially in Africa, and it’s well understood.
But many kinds of grammatical tone cannot possibly be explained in such a way. Kusaal verbs (for example) undergo complete tone neutralisation in most kinds of non-subordinate clause; Dogon nouns undergo complete tone neutralisation before modifiers … examples abound.
Declaring that such phenomena are the result of unspecified processes of “erosion”, via no attested tonal mechanism, is … unconvincing …
I really don’t like this whiff of facts being selected and adjusted to fit the theory. It brings the game into disrepute.
@DE, it’s not fair: I only quoted the very beginning of his quest. Khmer appears later, see here:
https://epdf.tips/defining-creole.html
He actually explained why he speaks about tones.
The question of a meaningful measure of complexity totally makes sense. Intuitively, some languages are “more complex” in the sense: “harder for a learner” (but it is hard to speak of “benefits” of being a native speaker of an easy or hard langauge). The idea that we should not explore this because speakers of simple languages will be deeply humiliated by speaking a simple language is mad. We all have butts, let us be ashamed of our butts if we need to be ashamed of something.
But tones… Well, it seems the only reason they are here is
(1) so we could rule out Chinese
(2) he knows some theory of tones where they are the “second floor”. He finds them an impressive exmaple of a second floor.
I can’t object to the “impressive” part (if it impresses him, then it is a fact that they are impressive) but intuitively a similar disctinction “some languages are inflecting and some are alalytical” seems more fundamental than “some have tone and some do not have tone”.
As for (1) it reminds me about flat-nailed bipeds.
Thanks for that, drasvi. I see that I was indeed interpreting that bit out of context, as you rightly say.
I remain unimpressed by McWhorter’s Prototype, though impressed by the ingenuity of the many and various dodges he has come up with to protect the idea against definitive refutation. Essentially, no counterexample can ever really count. What is it with the UG People?
I must admit that he is at least interestingly wrong.
And why should he? It’s the kind of thing that would only occur to a speaker of a language without tones.
Exactly. I was just trying to show that even if you assume that, it has no bearing on pidgins except if you hold some additional unfounded assumptions.
As there are languages without any stop consonants at any particular point of articulation you care to specify, languages without any nasals, languages without any fricatives, and languages without any liquids or approximants, it follows that no consonant is part of UG.
That’s exactly the absurd result McW’s kind of thinking would lead to, yes.
There might be two vowels, though, so our forebears may have spoken in binary.
But here you’re not radical enough; as Ubykh has shown that languages only need one vowel phoneme, that surely was the case in the primordial language. So it wasn’t “UG, UG, UG”, it was “aah, aah, aah”.
Presumably if historical linguistics had first arisen among speakers of Chinese or Kusaal, it would have been assumed as a matter of course that primordial human language had tones.
I’ve heard of some actual European (and/or white North American) linguists suggesting that (essentially) Khoisan languages are too complicated to have derived all their weird stuff, so surely the primordial human language was Khoisan-like, with tones and clicks, and then most everything that wasn’t Khoisan happened to lose their clicks and (presumably) tones. On the face of it this doesn’t sound like a noticeably worse argument than McWhorter’s.
How plausible a generalization is it that creoles are in general phonologically “simpler” than non-creoles — not in the sense of all creoles being on one side of some bright line and all non-creoles on the other but as a sort of general statistical tendency? Presence rather than absence of phonological tone would be ceteris paribus evidence of comparative complexity, just like a large consonant inventory, or phonotactics that permit consonant clusters that are comparatively cross-linguistically rare etc.
Sorry for Russian. It is from Комедия книги (A könyv komédiája) by István Ráth-Végh. A collection of curiousities about books that I read as a child. Here he is retelling this:
https://archive.org/details/johannispetrieri00eric/page/1/mode/2up (in Latin with weird Greek). Namely that Adam invented Ω when he looked aound and said Ω and then he smelled Eve and invented ὗ.
But I am too tired to check the Latin original.
(the same source also recommends: Reimmann J. Fr. Versuch einer Einleitung in die Historiam Literariam Antediluvianam. Halle, 1727.)
@JWB:
I think that, as a general tendency, it’s quite true; come to that, I don’t have any real beef with (what I take to be) McWhorter’s basic idea that there is often a sort of lowest-common-denominator thing going on there. It’s his overgeneralisations that bug me. (Along with a striking tendency to handwave or to cite predigested secondary sources with their own theoretical axes to grind, rather than reliable primary sources.)
I wonder what he and Everett think of each other.
Because you know, information can be conveyed by tone alone and stops can be used interchangeably… “ornamental to nuanced” in McWh.’s words:)
information can be conveyed by tone alone and stops can be used interchangeably
It’s striking that in at least the Oti-Volta languages, word tone is more stable than word segmental structure, which seems very counterintuitive to a speaker of a non-tone language, but is actually very common in tone languages (it’s at the back of many of the “floating tone” phenomena I mentioned above, and indeed this motivated the whole “autosegmental” approach to tone.)
A concrete example is the combining forms of Kusaal nouns, used before adjectives and bound pronouns like demonstratives. In principle, these are bare stem forms, but in many cases where bare stems would be ambiguous they have been remodelled after the singular stem-plus-class-suffix form; with certain stem types and noun classes this has actually become regular. But the remodelling never affects tone, and the effects can’t be described in terms of floating tones, either, because they involve the addition of segmental material, not its loss: it’s purely that tone is more stable than segmental shape.
All this despite the fact that lexical tone carries quite a low functional load in Kusaal; the absence of tone marking in the orthography causes no real problems to the relatively few people used to reading Kusaal at all.
[Though I suspect it explains one or two cases in the Bible version where combining forms seem to have been coordinated, a thing my informants didn’t permit: in all such cases I’ve identified, the combining form and singular are segmentally identical, and I think that they’ve simply been confused, so that people lost track of the intended construction; because of the lack of tone marking, the error doesn’t jump out at you unless you’re a foreigner grimly trying to parse it.]
I wonder if it is easier or harder for speakers of a lnaguage like Chinese to come up with a whistle langauge…
I am clicking random links in search of definitions of “creole”. One of them The Handbook of Pidgin and Creole Studies, Introduction:
I don’t remember anyone putting Lenin first and four years earlier next:-/
What’s wrong with them? They worship him.
I wonder if it is easier or harder for speakers of a language like Chinese to come up with a whistle language…
We discussed the kindred phenomenon of Drum Language previously:
https://languagehat.com/drum-language/
I’m sorry to say that I know nothing about it myself. I know that there exists a Dagomba drum language, and the Dagomba drummers are in fact the official royal annalists.* Dagbani is like its close relative Kusaal in that it has lexical and grammatical tone but they don’t carry a heavy functional load, so presumably communication employs similar strategies to those described for Yoruba in that thread.
Although Yoruba has a more complex tone system than Dagbani**, I’m pretty certain that you couldn’t follow arbitrary non-conventionalised Yoruba speech on the basis of tone alone. (My grandfather had a Yoruba talking drum from his time in Nigeria, and used to demonstrate its rendering of Yoruba phrases, but I’ve no idea how accurate he was, if at all. I was about ten years old at the time …)
* Because the Dagomba drummer clan are much the best organised among the Mossi-Dagomba peoples, this has given the traditional history of the region a rather odd Dagombacentric slant. The Mamprussi kingdom is actually senior to all the rest.
** Kusaal, like Yoruba, has three basic tones, but the actual underlying complexity of the system is virtually the same as Dagbani, which has the locally favoured high/low terracing system with emic downsteps; the Yoruba system is distinctly more complex. Kusaal’s at-most-very-distantly-related (Mande) immediate neighbour to the north, Bisa, has also managed to create a three-tone system out of an original two, though by a different route. But Yoruba seems to have been born that way …
What do you mean by “emic downsteps”? I understand the terms only separately.
What’s wrong with them? They worship him.
Bioprogram bollocks. That’s what’s wrong with them.
I was sorry to realise that McWhorter, in most respects a fine and upstanding Real Linguist who actually studies actual languages instead of discovering the Truth by introspection, has been touched by the Chomsky Plague. Explains a lot. Alas.
The above quotation is not McWhorter….
Since philosophy is just a pretend projection of personality, I’ll go on a limb and guess about McWhorter (and I apologize if I do him injustice. I don’t know him.) I get the impression that McWhorter is at heart a counter-leftist reactionary. His Doing Our Own Thing is a book about the “degradation of language” (this, from a linguist), and about how that darn Rap is making kids today stupid or something. These kind of opinions coming from a Black man are hot stuff among conservative think tanks like the Manhattan Institute (where he was stationed for years) and their consumers.
If left-wing Mufwene and DeGraff stand for restoring the image of Black creoles, underlain by the liberal axiom of equality among people and cultures, then opposite stands McWhorter, who I feel gets his jollies out of arguing that some languages are, in fact, lesser than others.
It’s one of those weird exceptions that you’d never guess. The impossibility of très beaucoup is another.
Russian = Chinese = US second floor = western European first floor.
Kusaal, whose tone system goes straight back to Proto-Volta-Congo at the very least? Sure. Chinese, though? IIRC, there was at least one Chinese philologist in the 19th century who said Old Chinese lacked tones, as it turns out it did – tone developed between Early and Late Middle Chinese, between the 7th and the 11th century IIRC, out of syllable-final consonants, followed by a split of the whole system by reinterpretation of the voice contrast of syllable-initial consonants.
Yes, through loss of [r].
Ubykh has three. Abaza has only two, but that’s the limit so far.
On the morphophonemic level, there are one-vowel languages (and Ubykh might qualify as one), but not on any other.
What do you mean by “emic downsteps”?
Sorry. I’m too accustomed to this stuff to remember my manners …
The typical sort of tone system that turns up in West Africa is what’s called a “terracing” register system. Hausa is a good and pretty simple example: the tone-bearing unit is the syllable, and the basic tones are just High and Low (there are some falling tones caused by merging of HL on one syllable, too, but they are plainly secondary; Hausa has no rising tones.)
Characteristic of these systems is that a H following a L is pronounced at a lower pitch, so that the overall pitch of H tones within a breath group steadily falls: this is “automatic” downstep, and it’s fully predictable and thus not meaningful. In Hausa, that’s basically all there is to it.
However, in Dagbani, which likewise has a two-tone register system with automatic downstep, there is a further complication. Low tones are often lost because the vowel of the syllable they fall on has been elided. When this happens, a following H is still downstepped: but now this is no longer “automatic” and predictable, but “emic.”
For example, “stranger, guest” is sánà, from *sáánà (Dagbani has a peculiar rule of shortening of original long vowels in originally open syllables.)
But the plural *sáánìbá has lost the second-syllable vowel, but the final á is still downstepped after the lost L tone: sáám!bá (where ! represents emic downstep.)
Kusaal (in case anyone is interested) has reworked the system by turning H! sequences into an extra-High tone; it’s simplest just to call this High, with the original H “demoted” to Mid: thus Kusaal sāan “stranger”, plural sáam.
The above quotation is not McWhorter….
I realise that: I was basing my more-in-sorrow-than-in-anger comment on McW’s writings elsewhere. He does seem to have been exposed to the virus, poor fellow.
I can’t say I’ve read everything or even a majority of things McWhorter has written, but perhaps I’ve read different pieces of his ouevre than David M. has and I would say that he has written plenty about the legitimacy and interestingness of AAVE as a language variety. That is not necessarily inconsistent with being a reactionary when it comes to politics of course. Any good reactionary should support traditional local dialect variation against Esperanto and/or some semi-artificial standardized/elite dialect being promoted by the Illuminati shills at the Ministry of Education as the Only Way for vulgar social climbers to obtain cultural capital and join the Illuminati themselves.
In one of his books he writes quite interestingly, and with a reasonable amount of phonological detail, about what he dubs (unfortunately to my mind – it has the semi-cleverness that is the Achilles’ heel of popularizing writers on technical topics) the “Blaccent,” i.e. the specific phonological tells that will reveal to an AmEng speaker with a good ear that the speaker is probably black (not even Southern white) even when speaking perfectly standard AmEng with no identifiably AAVEish syntactic or lexical peculiarities.
I get the impression that McWhorter is at heart a counter-leftist reactionary.
For what it’s worth, I have the same impression.
His later political writings are more of the tedious same.
To be fair, What Happened to English?, His paper on Norse contact precipitating the simplification of English morphology, is good.
Now I can’t remember if it’s in the same book or not (my bookshelves are currently an extremely disordered mess), but McWhorter is one of comparatively few scholars I’m aware of to have done work on long-term historical shifts in AAVE. There have been some within my own lifetime — Baugh’s “Black Street Speech” (published not too long before I read it in a class circa 1986 but based on fieldwork done 15 or so years previously) talked up one or two syntactic constructions that were completely absent from the idiolects of the (now middle-aged) black kids in my own generational cohort that I grew up with.
But McWhorter was struck by the notion (obvious to anyone if you think about it) that the highly stereotypical “Comical Negro Dialect” that you can find in plenty of written American sources from the 19th century through let’s say just before WW2 contains a number of features that were pretty much totally absent from the actually-existing AAVE of the Sixties-turning-into-Seventies that first attracted modern published scholarly study. Is that because it was a pejorative outsider’s idea of how Negroes spoke that was not based on good empirical data because the outsiders frankly didn’t care that much about accuracy, or had the empirical data itself changed over time? That’s to my mind an interesting question susceptible (if but only if you can find the right data/evidence) of empirical inquiry. And lo and behold McWhorter found some early 20th century recordings of born-in-the-19th-century black speakers, whose speech indeed had some (not all) of the features embedded in those old comical/pejorative representations which their grandchildren’s speech had lost.
All of this may frankly be in some sense a less useful deployment of scholarly attention and time than documenting and analyzing the creole(s) of Suriname, but to the extent the question of whether AAVE and other distinctively black speech varieties are something other than deficient/error-ridden standard English, which is a question with some broader political-etc. consequences in the U.S., it is my view that McWhorter on balance has on that issue been on the side of good scholarship. Which I would hope Hattery-denizens would agree is thus the side of the angels.
Just do not confuse a “reactionary” in the sense “feels more comfortable in a racist society” with “reactionary” in the sense “disagrees with someone”.
By “reactionary” I mean something like “contrarian”.
@Y & David Marjanović: Actually, I think you can say, “almost only,” in idiomatic English; what you generally cannot do is write it. The reason is that to use “almost only,” there has to be a very strong stress on “almost” (probably including a short pause after the word and maybe gestural intensification as well). Spoken aloud, the following seems fine to me
However, it doesn’t seem like it can be written in normal prose, because it reads wrong without the very strong stress.
The OED (in an entry that is supposed to be fully up to date) lists the generalized sense of reactionary:
only under the adjective form of the word. For the noun, they only give the sociopolitical sense. My subjective impression is that the generalized sense is indeed quite a bit more commonly intended when the adjective is used than the noun; however, the corresponding generalized sense of the noun does also exist.
On the morphophonemic level, there are one-vowel languages (and Ubykh might qualify as one), but not on any other.
Don’t spoil my jokes by insisting on exactitude.
as long as we’re here…
@Y, people who chronically disagree [with opinions of a group N] are not the same as people, who, in their hearts, have goals [opposite to what N proclaim].
I grumble at feminism because of my own rather strong feminist inclinations. (Just as an example). Similarly I grumble at some science fiction books… but never romance fiction.
The question “do languages differ with respect to their complexity?” seems interesting: for typology, for applied lingustics and for any innocent langauge learner who tried two languages and found one of them much easier.
I heard somewhere (in some linguistics lecture a few months ago) that there’s been an attempt of a compound ranking of a bunch of languages as easier or harder to learn by various criteria of their complexity, and it turned out that Chinese scored in the top few easiest, apparently mostly because of its very simple grammar. Not sure if there’s actually a source for this, though.
whose speech indeed had some (not all) of the features embedded in those old comical/pejorative representations which their grandchildren’s speech had lost
Such was the case for one of the two dialects of English that my father commanded (the other being, as I suppose, ordinary Philadelphian of the early 20C): it was singularly more like Stage Oirish than anyone from Ireland would be willing nowadays to admit.
@JfoM:
Well, Dagbani is pretty easy if you know Kusaal …
An objective measure of language complexity would have to transcend all the familiarity/unfamiliarity stuff that in reality makes Mandarin pretty difficult for most human beings on the planet.
I think there actually have been efforts in that direction, though I know no details. Devising fairly objective measures of the complexity of phonology and morphology would be relatively tractable, I suspect; the real difficulty would be with syntax, not only because it is not easy to see where to start but also because the syntax of most languages has so far been very poorly described.
In lexicon, there is a tendency to suppose that fewer individual lexemes straightforwardly implies greater simplicity. This ain’t necessarily so, as anyone who knows both Latin and Greek can tell you.
And the idea that any form of Chinese has “very simple” syntax strikes me as a pure artefact of ignorance. Come to that, Kofi Yakpo’s Pichi work, which I praised above, suggests strongly to me that the idea that creoles have radically simple grammars is based partly on focusing exclusively on morphology and partly on relying on woefully inadequate descriptions, often (I suspect) inadequate for the very reason that their writers just assumed there wasn’t a lot there to need describing, because everybody knows creoles are simple, right?
I don’t think that there can really be any doubt that languages really do differ quite considerably in intrinsic complexity, mind; still, measuring such complexity across all grammatical domains, not just relatively easy ones like phonology and morphology, strikes me as very hard.
It’s also worth saying that learnability and complexity are not the same thing, even factoring out the learner’s own language background. Some languages hit you at the outset with major difficulties before you can say or understand much of anything; others sadistically keep the real complexity in reserve, to zap you with later.
Even thinking up a good metric for phonological complexity seems harder the more I think about it. A language may have a relatively modest total number of segments and suprasegmentals, but highly complex internal and external sandhi rules, for example: how do you begin to come up with a meaningful objective number for the complexity of such rules, especially given that the precise rules vary hugely between languages, and that there are virtually always several different but perfectly good ways of analysing the system even within a single language?
How do you weight the complicated system whereby Kusaal words drop final short vowels in most contexts but keep them in others, and keep them, but with systematic mergers, in yet other contexts? Outside Kusaal and its two closest relatives, Nabit and Talni, I know of few systems which are even similar, and none quite as complicated: do you just give Kusaal a bonus point for that, given that cross-linguistically there are so few good points of comparison for it? It certainly is one of the major difficulties for a foreigner trying to learn the language …
Although accounts of the language (other than mine) handwave it all vaguely as part of the morphology, this is not so: it’s all external sandhi, and not even morphophonemics.
@David E.: yeah, weighting different aspects is difficult-to-impossible to do non-arbitrarily so that’s where this tends to fall apart. Even if you can say with a high degree of confidence that language A is phonologically more complex than language B but is also morphologically less complex than language B, how do you determine which of the two is more complex on net in some sort of objective/quantified way? Easier to fall back on the Intro Linguistics Myth (or perhaps Noble Lie?) that eh, they’re all pretty much the same overall, just with variation in which parts are more complex v. less complex.
It’s dawned on me that Mandarin has automatic downstep.
It also has one of the easiest tone systems if you aren’t used to tones and lack absolute pitch: the tones are distinguished by contour more than by pitch. Compare and contrast Cantonese; it’s got around six tones, of which three are level – and packed in the lower half of the pitch range.
Conservatives want the status quo, reactionaries want the status quo ante.
@JWB:
I wouldn’t go so far as to abandon the idea completely: I’m pretty certain, for example, that apart from people who already speak a Berber language, pretty much everybody on the planet would find it easier to learn Timbuktu Songhay than to learn Tuareg. I suppose it’s possible that Koyra Chiini has hidden syntactic foibles that would ultimately lead to the Songhay tortoise overtaking the Berber hare, but I don’t have any evidence to support the idea, and it seems frankly unlikely.
But I suppose that, as you imply, the fundamental problem is that it’s a delusion to suppose that language complexity could ever be reduced to a single number.
Right, which is why I don’t see why it’s so bad to fall back on “the Intro Linguistics Myth (or perhaps Noble Lie?) that eh, they’re all pretty much the same overall, just with variation in which parts are more complex v. less complex.” It seems to me close enough for government work, and a useful corrective to the usual simplistic notions that there are “easy” and “hard” languages.
@DE, JWB, I thought about the afroementioned complications (I can only add “they’re all pretty much the same overall, just with variation in which parts are more complex v. less complex” does not work, because the same reasoning is applicable to any given part: all parts are equally complex, just with variation within the part. Let’s go fractal) and I agree with DE, but…
… don’t these complication means that it is a GOOD question?
A good question is not one to which you have ready answer that you can write down in your Book, put a tick in your list and forget for there are no implications for anything. Good questions behave exactly like this: raise more questions.
lack absolute pitch
I don’t think that really comes into it. Unlike my wife and daughter, whose abilities in this regard strike me as positively uncanny, I have no notion of absolute pitch, but that has never been a problem for me in hearing language tone distinctions. (You certainly know, without my mentioning it, that no language makes use of absolute pitch – for obvious reasons.)
I think I probably am better than average at hearing pitch differences, though. I’ve always been more prone to notice out-of-tuneness than most, including (oddly) my wife and daughter. And I do know a quite eminent linguist who can’t hear linguistic lexical tone at all, though he presumably would have been able to if he’d grown up in Ghana.
Well, all right.
Is one ideal circle simpler than two ideal circles? If yes, why, in what sense?
(I honestly don’t know what is two ideal circles: being an ideal circle she (why not she) does not inhabit any specific space…)
Is one circle on a sheet of paper simpler than two? If yes, why, in what sense? (I haven’t specified if they have the same radius).
How about abandoning words “simple” and “complex” for good, if they lead to such complications? And what about all other words? Do we need them?
Is one circle on a sheet of paper simpler than two? If yes, why, in what sense?
Yes, of course. You need to specify their spatial relationship to one another, even if they have the same radius.
If you’re talking about the Platonic Ideal Circle in your first question, “two” is impossible by definition, so the question is meaningless.
How about abandoning words “simple” and “complex” for good, if they lead to such complications? And what about all other words? Do we need them?
Ug.
Words fail us when pushed beyond their limits. That hardly makes them useless. You might as well say that a car is useless because it can’t fly.
“government work” – can relate it to languages by noting that Russian вертел на хую and English “to screw” are based on the same underlying metaphor….
FSI and DLI both have language difficulty scales (which is not what motivates my underlying metaphors: diplomates and the military are good at languages, they suck at diplomacy and warfare).
And the idea that any form of Chinese has “very simple” syntax strikes me as a pure artefact of ignorance.
Hmm? Victor Mair claims Mandarin was one of the easiest languages for him to learn. (And he commands quite a few.) What makes it diabolical in most teaching settings is that they try to teach you the written form alongside the spoken.
I suspect that at the time, Prof Mair had the strongest motivation to communicate in Mandarin: reader, he married her.
Anthony Burgess claims Malay is easy to start with, but will take a lifetime to become competent.
Unlike my wife and daughter, whose abilities in this regard strike me as positively uncanny, I have no notion of absolute pitch,
Pity them! Absolute pitch is a curse. Later in life it disintegrates or (if you’re lucky) ceases to be; but while disintegrating, it makes everything sound out of tune — especially if you’ve accustomised to concert pitch.
FSI, hours to ILR/FSI level 3:
Category I: 23-24 weeks (575-600 hours) Languages closely related to English
Category II: 30 weeks (750 hours) Languages similar to English
Category III: 36 weeks (900 hours) Languages with linguistic and/or cultural differences from English
Category IV: 44 weeks (1100 hours) Languages with significan linguistic and/or cultural differences from English
Category V: 88 weeks (2200 hours) Languages which are exceptionally difficult for native English speakers
Names are total bullshit.
Your L1 (and also previous experience with other langauges) is a factor. But don’t claim it is the only factor.
Spanish is “closely related”, German is “similar”, Indonesian has “differnences” and all Europe but Romance or Germanic has “significan differences”.
Surely “Arabic, Cantonese, Mandarin, *Japanese, Korean” in group V has nothing to do with writing systems and Arabic diglossia (especially Japanese writing system has nothing to do with the fact that it is asterisked as harder than “super-hard” (another term from their site))
DLI:
Category I languages, 26-week courses, include Spanish, French, Italian and Portuguese.
Category II, 35 weeks, includes German and Indonesian
Category III, 48 weeks, includes Dari, Persian Farsi, Russian, Uzbek, Hindi, Urdu, Hebrew, Thai, Serbian Croatian, Tagalog, Turkish, Sorani and Kurmanji
Category IV, 64 weeks, includes Arabic, Chinese Mandarin, Korean, Japanese and Pashto
Interesting.
FSI 24 Spanish 44 Russian and DLI 26 Spanish 48 Russian are more or less proportional.
But… DLI has more weeks even though
– DLI as I understand, aims at ILR 2 while FSI aims for ILR 3.
– DLI courses must be more intensive than FSI’s 25-hour weeks (even with self-study which is not included in these 25). Somewhere they write “full-time immersion classes” whatever that means.
Soviet courses for military interpreters were really intensive.
FSI 44 Russian, 88 Arabic vs. DLI 48 Russian 64 Arabic are not proportional at all.
“Serbian Croatian”:)
he married her
This was actually suggested to me by Kusaasi colleagues as an effective route to improving my Kusaal, but my argument that this would be entirely in line with local cultural norms [a man has a positive responsibility to do his part by supporting as many wives as he can afford] cut surprisingly little ice with my existing non-Kusaasi wife.
I am inclined to think that if Prof Mair says that Mandarin is particularly easy, it’s probably in fact extremely difficult. I also am the proud owner of a copy of Y R Chao’s Grammar of Spoken Chinese, the notable thickness of which suggests to me that Chao may not have found his native language quite as simple as that talented fellow Mair did.
Spanish. Es una lengua muy fácil – until you’re suddenly supposed to keep three past tenses apart.
Very good point (I was literally falling asleep). I was (probably) thinking of the paper that found an allele related to perfect pitch to be noticeably more common in East/Southeast Asia and (IIRC at least) West Africa than elsewhere, where tone languages are considerably more common than elsewhere, but of course you’re right that absolute pitch ought to be rather a hindrance to learning phonemic tone.
FWIW, I haven’t even begun to learn enough tone or pitch-accent languages to do statistics with (uh… n = 1), but I seem to have grasped the Mandarin tones faster than my coursemates and also seem to be noticeably good at hearing out-of-tuneness. But it could also be general autistic precision or some leftover effect of having been bilingual when I was little or whatever. (No idea if I ever acquired the FYLOSC pitch accent, though.)
Ni oui, ni non, bien au contraire ! My limited experience is that practically all the difficulties you’d expect from a language are absent: no inflectional morphology, a minimal politeness system in the grammar, the word order mimics SVO most of the time, and while things like aspects exist they’re more part of the lexicon than of the grammar and largely optional. What’s trickier is that you have to rethink some things completely if SAE is all you know: instead of a relative pronoun, you use a word that turns everything that precedes it into an attribute (…a lot like Basque -ko, apparently), and “and” is handled in like 20 different ways (but the most common one is simple omission).
The limiting factor is the writing system. Either you have an amazing memory, or you get lots and lots and lots of practice, or you forget most characters the same day you learned them – and Prof. Mair has written plenty about “character amnesia” among native speakers.
an allele related to perfect pitch to be noticeably more common in East/Southeast Asia and (IIRC at least) West Africa than elsewhere, where tone languages are considerably more common than elsewhere, but of course you’re right that absolute pitch ought to be rather a hindrance to learning phonemic tone.
A musician’s take, and repeating that same meme — it’s all to do with plasticity of the brain of a child listening to language. (Allegedly: he’s a musician not a phoneticist; props for mentioning Sviatoslav Richter and Oscar Peterson in the same video.)
I do recall reading somewhere an account by a student of Mandarin and Japanese where he talked about how much harder Japanese was, once you got past the initial stages; he’d got to the point of being able to converse fluently and idiomatically in Mandarin after a couple of years of fairly heavy exposure, but he despaired of ever achieving the same thing in Japanese. (The writing system is worse too, but his complaint was all about spoken language.)
“until you’re suddenly supposed to keep three past tenses apart.” – this is difficulty of preforming a certain task for adult learners….
We all are familiar with various kinds of, say, genitive in Latin. Partitivus etc. In Russian I use imperfective in experiential meaning, in habitual/repetetive meaning, in continuous meaning.
Where all of this comes from?
1. Do these “meanings” exist in speakers’ heads?
2. Or do we confuse them (“I don’t know which of two is meant here, maybe neither maybe both”) and existence of “meanings” only makes sense in comparison to a foreign language?
3. Are they inventions of insane classificators based on what those believed must be “logically” different senses?
If the answer is (1) we have a problem. We must measure complexity of our feelings about langauge:/
If the answer is (2) we again have a problem. The whole concept of grammar depends on TWO langauges, not one.
And for (3), unless we ignore the approach to a langauge based on context (where a “native speaker” can predict what form is used where – and the goal of anyone describing a language is also predicting this) we run into nightmare: the set of all possible contexts.
The whole concept of grammar depends on TWO languages, not one.
It might do if languages had no connexion with meaning, I suppose. But in fact, it is perfectly possible to construct a grammar of one’s own language without reference to any other. The Greeks did (and our grammatical tradition still uses much of their framework.) So did Panini and his predecessors.
There are actually formal techniques for working out the grammatical categories of an unfamiliar language, without depending on intermediate-language translation. From Kapitonov’s nice Kunbarlang grammar, for example:
I think it’s mostly a mixture of 2) and 3) – these distinctions were born by grammarians trying to teach Classical Greek and Latin. And besides for learning, these distinctions make most sense as shorthand in language comparison – e.g. when stating that Russian has a (marginal) separate form for the partitive genitive while (say) Polish hasn’t.
@DE: I’d say that in order for a tradition of grammatical analysis to arise, you need if not foreign language speakers trying to learn a different language, then at least competing lects of the same language, which creates the impetus for deciding what the “correct” usage is, resulting at one point in descriptions of the correct usage. Historically, that usually was a (mostly written) prestige lect that needed to be taught because it wasn’t the native lect for many speakers.
That might well account for the motivation of the earliest grammarians (Panini and his tradition is all about preserving the purity of the sacred tongue from colloquial corruption, for a start) but I don’t think it accounts for the actual grammatical analyses themselves. In the case of the Greeks, for example, the terminology and categories that were created apply pretty much equally to all forms of Greek at that time: indeed, that seems to be the whole point of them: I suspect that Dionysius Thrax, like many a much less sophisticated thinker about language since, naively supposed that the structures of his own mother tongue were just “logical” universal linguistic givens.
https://en.wikipedia.org/wiki/Dionysius_Thrax#T%C3%A9khn%C4%93_grammatik%C4%93
“The Greeks did”
Yes, and they ruin the row “Sumerian grammars, Panini, Sibawayh, [and Dietmar Elyashevich Rosenthal who wrote Russian style guides]”.
All right, Panini, but you can explain Panini: Sanskrit both has a VERY special status and is not too native. They learned it.
But Greeks!?!
I suppose you might say that Chomsky constructed a grammar of his native language without reference to any other*, but then he doesn’t believe that there are any others.
I must also own to some minor niggling doubts about the accuracy and reliability of his grammar …
* Well, I suppose I would have to concede that he was comparing English with abstract mathematical structures which are often called “languages” by (a somewhat forced) analogy …
@DE, well, yes. Maybe he is a genius.
@Hans, for me it is an illustration of why such questions (“can we meaningfully define complexity?” in this case) are productive. It made me think of those semantical distinctions and I simply enjoy it.
Subjectively, sometimes it is easy for me to tell what I mean.
Don’t “I [ever] read Lord of the Rings”, “I [often] read LotR”, “I read LotR [and went to bed]” feel like different things? Existence of different forms (like Russian two forms for verb of movement) makes one think that speakers could already feel difference before the formal distinction. And speaking about verbs of movement, adverbs like “бегал in circles”, “бегал to and fro” maybe don’t cover бегал as opposed to бежал wholly (e.g. when бегал means sport, jogging), but relative ease of finding them might be related to existence of semantical cathegory. Same for “ever” (in English in questions).
But there are, in turn, situations when I simply don’t know what English cathegory I mean in a particular case.
It happens with definiteness.*
—
And if “grammar” only exists for TWO langauges, it si an epistemological disaster. Or fun. Or something:)
—
* I mean, I sometimes can produce two English translations, one with obvious “a” and one with obvious ‘the”, but have no slightest idea which one is meant in Russian.
I also don’t know if I mean a cup or what when in Russian I say “vessel” and the whole idea of requiring all Wikipedias to have an article “[a] girlfriend” is strange. You can be someone’s girl.
You can’t be just “a girlfriend”… or you can. In ENglish:)
That might well account for the motivation of the earliest grammarians (Panini and his tradition is all about preserving the purity of the sacred tongue from colloquial corruption, for a start) but I don’t think it accounts for the actual grammatical analyses themselves
Just to make that clear, I don’t think that either. It started off as peevery, so to speak, but it turned into serious analysis. And in Ancient Greek there was the added point that there were several varieties that were used for different literary genres, so grammatical analysis had to be able to account for them all.
Here the goal was to teach Classical (in several dialects as Hans mentioned) and Homeric Greek to native speakers of Hellenistic koiné. All six chapters, actually, seem to be about the correct recitation and understanding of ancient revered texts, even though they weren’t revered to the extreme extent of the Vedas.
So the whole concept of grammar originated in peevery …
Yet may we transcend our base origin, Brothers and Sisters in the Great Common Syntactic Task! I see a glorious Dawn of Descriptivism, not afar off, but already amongst us for those with eyes to see! Excelsior!
#Authorship though.
“Though initially rebuffed by scholars of the caliber of Pfeiffer and Hartmut Erbse, Di Benedetto’s argument today has found general acceptance among specialists.” says WP referring to De Jonge
Oh! So most of the text was intended to teach Classical (in several dialects) and Homeric Greek to native speakers of Late Roman-era Alexandrine Greek across a gap of 700–1000 years.
Greek was taught to speakers of other languages – we know it well from the history of Rome – so it is plausible that there was an influx of grammatical ideas from teachers and learners.
The actual set of extant texts does not give you a precise understanding of where the ideas originated: even your genius author was the first to confide an idea to paper (and we never can be sure) it says nothign about where she heard it, who she discussed it with, what was the background that made the idea natural (like special relativity was natural 100 years ago) and how exactly she came up with it if it is her idea.
Sibawayh was not the only Arabic grammarian.
Greek was taught to speakers of other languages – we know it well from the history of Rome
I’m not sure how much of this was formal teaching in the classroom sense that we modern Westerners think of when we think of language learning. Upper-class Romans by the late Republic were raised pretty much bilingual in Latin and Greek*, and the sort of lower-class yobs (slaves, even) that Paul wrote to at Rome probably mostly didn’t learn their Greek at school. It was the urban lingua franca of the richer half of the Empire, after all.
I imagine that formal Greek teaching was indeed largely concerned with instructing pupils who already spoke Greek how to understand (and imitate) the classics.
* If Suetonius is to be believed, Julius Caesar’s last words were καὶ σύ τέκνον; presumably to M Brutus, son of his erstwhile mistress Servilia. Even if this is not true, it must mean something that people thought it plausible that Caesar might have spoken Greek even in articulo mortis.
“I imagine that formal Greek teaching was indeed largely concerned with instructing pupils who already spoke Greek how to understand (and imitate) the classics.”
@DE, but how well Roman pupils knew spoken Greek? And did they always acquire it as toddlers “from overheard conversations”?
Also even before Rome, (1) Greek was associated with impressive cultural and literary tradition (2) Greek was foreing to many people who nevertheless came in contact with it (3) the Mediterranean world knew schooling (in different forms: for children and adults, nobility and professionals).
Even language teaching, because of diglossia (Sumerian among Akkadian speakers, Akkadian writing among Semites, various forms of Egyptian).
Does not this create a market for formal Greek teachers for rich people who want to converse about philosophy, but speak something else?
Not just Rome, let it be Ptolemaic Egypt (or even India): were not there rich families who just were not Greek?
Medieval (and later) Christian and Muslim worlds knew language education, why BC dudes would be different?
@DE: … that Caesar might have spoken Greek even in articulo mortis.
His penultimate prayer is said to have been:
In articulo mortis caelitus mihi vires.
Deo adjuvante non timendum in perpetuum.
Dirige nos Domine ad augusta per angusta.
Sic itur ad Graecam excelsior
@drasvi:
Cleopatra was the first (and, of course, last) of the Ptolemies who could speak Egyptian. The sort of people in Ptolemaic Egypt who afford to pay for language teaching were Greek-speaking already. In lower Egypt, there were also a lot of Greek/Egyptian bilinguals, as the vocabulary of Coptic demonstrates.
And as I say, it’s a mistake to project modern Western ideas about foreign-language learning into other times – or places.
An example of the normal state of affairs:
My chief outpatient nurse in Ghana was an L1 Kusaal speaker. He was also fluent in English, Hausa, Mooré and Twi, and knew a fair bit of Bisa. The only one of these he’d learnt at school.was English, and that was not really by any formal teaching but because it was the working language there and he was sent there early enough to pick it up in the way children still can. The only language he could read and write was English. (His impressive set of languages was unusual locally only in that he knew some Bisa: local Bisa people use Kusaal to communicate with non-Bisa, few of whom bother to learn any Bisa.)
David Eddyshaw: If Suetonius is to be believed…
That’s never a good start to a proposition. However, there is no reason to believe that the dictator’s last words were not in Greek, or at least that they were not so reported by the Liberators, who would be the only ones in a position to know. “The die is cast,” was probably in Greek (and may have been a specific literary allusion). Casca, probably the third most important of the conspirators, is supposed to have shouted to the others for help in Greek, after he delivered the first, rather ineffectual, blow of the attack.
That Brutus was actually Caesar’s son seems to have been an ancient meme, although perhaps one originating in the first or second century C. E., long after everyone personally acquainted with those involved was dead. Plutarch, for instance, insinuates that Caesar and Servillia were lovers as teens, probably to support the idea of Brutus being plausibly Caesar’s son, even though Brutus was only about fifteen years younger. In reality, the affair between Caesar and Brutus’s mother probably only began during or shortly before the period of the First Triumvirate, by which time Servillia was probably a widow twice over.
It has been suggested that the reason Caesar’s last line in Shakespeare is in Latin is to represent his “actual” last words to Brutus being in Greek. Shakespeare’s play may also demonstrate a divergence in how the principals involved in the assassination were perceived in different European cultures, more than a thousand years after the fact. A few centuries earlier, Dante had placed Brutus and Cassius at the very lowest point in hell, as betrayers. In contrast, Shakespeare’s play makes Brutus a tragic hero, and Cassius is a fully rounded character, although so is, for example, Marc Anthony. (Shakespeare also makes Casca the third most important of the killers, although he is a less interesting character.)
“The sort of people in Ptolemaic Egypt who afford to pay for language teaching were Greek-speaking already.”
I know about presence of Greek speakers in Egypt, but I have no reason to assume absence of other elites….
Actually I have a good reason to assume their presence when Alexander came, and the subsequent (and possibly preceding) growth of Greek-speaking population was not instantaneous.
“The only language he could read and write was English.” – this peculiarity is associated with schooling. (I do not mean, of course that school is the only way to learn to “read and write”. Just “associated” and just in his case)
“And as I say, it’s a mistake to project modern Western ideas about foreign-language learning into other times – or places.”
I understand you here but people don’t normally project schools into there. What you mean is that there are other modes of language exchange than what school offers. They can be so successful, that in certain communities speaking 5 languages is normal, while here it is like “wow, you can read in English!”
Yes, there is no reason to think that those other modes were not in action in the Antiquity (though educated Classical Greeks do not seem too polyglotic:)). But what is the reason to think that teaching was absent?
These two claims: 1. presence of what is uncommon among educated Muscovites and 2. absence of what is common among educated Muscovites are not identical.
The important component here is literary prestige.
These two claims: 1. presence of what is uncommon among educated Muscovites and 2. absence of what is common among educated Muscovites are not identical.
That’s right – insofar as I can make sense out of that mind-bending pair of contraries – but 2. does imply 1, as can be seen by the following mind-bending argument:
Let C(X) denote all things commonly found among a group of people X, and let M and G be groups of people. If C(M) is absent from G, then it must be that G != M, and whatever is common to that group of people, C(G), is uncommon in group M.
The absence in G of what is common among educated Muscovites implies that what is common in G is uncommon among educated Muscovites. 2. implies 1.
It may be that nothing is uncommon among educated Muscovites M, and that the members of a group G have nothing in common. In this case the implication holds trivially.
If such thought-processes are common among educated Muscovites then I, for one, would flee their company.
I would in fact flee any company that operated with “presence”, “absence”, “common” and “uncommon” in such a way.
“and let M be a group” > “and let M and G be groups of people”
Yes, I wondered if it was intentional.
“Aquí se queda la clara, la entrañable transparencia, de tu querida presencia…“
@drasvi: Even if three is formal teaching, it doesn’t have to include grammatical analysis and explanation of why one usage is correct and another is wrong – in many traditional school settings, the method consisted of rote learning the correct usage and hitting pupils with a stick when they got it wrong.
I’ve finally got round to reading Creole prosodic systems are areal, not simple; excellent stuff. Quite apart from demolishing McWhorter’s selective misanalyses and special pleading, its description of how stress can get reinterpreted as tone (and vice versa) is in itself very interesting (and he cites actual cases, not “often” this or “often” that.)
Kusaal basically carries over Hausa tone as is, but the loans all seem to antedate an internal sandhi process of rightward H tone spreading for which there is some other evidence of lateness: it may only have been completed as recently as as the 1970’s.
English stress, on the other hand, is consistently mapped to high tone, except in words which were probably transmitted via Hausa; moreover, this high tone isn’t subject to normal internal tone sandhi. Thus, lɔr “car, lorry” has high tone, and its plural lɔya has the tones high-low; if the word were not an English loan it should have been mid-high. (The fact that the word has been completely adapted to the noun class system segmentally but remains irregular tonally is also characteristic of register tone languages: tone is consistently more stable than segmental form, weird as that seems to speakers of non-tonal languages.)
I don’t see a theory of tonality in McWhorter’s publications (haven’t read all, though) and I don’t think he can be the main target of “Creole prosodic systems”: there is not much to target.
@Hans, yes, but I did not say that language learning and teaching always results in a grammatical theory committed to paper:)
I just think it is one context (perhaps among many) where such ideas come to mind with relative ease.
And why do we have Arab and (medieval) Latin grammarians? I am not saying “because of learning”, but why?
You can recite the scripture without grammar…
As for formal teaching: I am not sure what exactly the term means (actually I did not hear “formal” used this way in Russian when I was a student, but I heard it many times in English). What we need is a community where people discuss language – or one very smart person who developed a system on her own.
Teaching and learning is one context where people think about language and discuss language.
I don’t see a theory of tonality in McWhorter’s publications
Well, that’s kind of the point, is it not?
If you’re going to make sweeping statements about creoles and tone (which McW does), it might be an idea to actually know something about tone systems (which McW does not, or at least, not nearly as much as Kofi Yakpo does.)
in many traditional school settings, the method consisted of rote learning the correct usage and hitting pupils with a stick when they got it wrong
This is unfortunately not a bad description of a great many traditional Qur’anic schools in West Africa (for all that there is also an impressive tradition of actual literacy in Arabic in Muslim West Africa.)
But is it so bad a method for language learning?
I think those guys who buy (or download) a Pumsleur’s or Glossika course and then listen to and repeat 10 000 short Russian phrases with translations when they are driving or cooking or whatever are going to be quite efficient. (not very enjoyable method, though).
I read learners’ essays and totally loved an Argentinian girl: she wrote short stories in her very basic Russian and she specifically used those elements of langauge that “advanced” (that is, those who receive compliments “I thought you are a native speaker!” from native speakers) learners avoid. Importantly the conjunction “a”.
And she felt native. I mean, she was making millions errors (she was a beginner) but she felt like your classmate in the first grade:)
(‘girl’ above meant young lady, she was not a child).
Yet Pimsleur (used to cost 1000, now just $575…) and cheaper Glossika is what various polyglotes from the Internet recommend. I am not an Internet polyglote, I like that girl’s approach, but it works for them.
@drasvi – you post a lot and you have many interesting ideas, but it’s often difficult for me to keep track of what you’re actually arguing for or against.
Hans, I was responding to this (DE):
This is unfortunately not a bad description of a great many traditional Qur’anic schools (emphasis mine)
I think language learning is about rote memorisation anyway.
Pimsleur is a fashionable and expensive “Method” praised by many Internet polyglots (an Internet polyglot is a genre on its own). Recordings of numerous phrases in the target language and their translations (also in audio). You are supposed to listen to it in your free time and repeat the phrases after the recording.
Rote memorisation.
What is “unfortunate” is that the students may end up memorising the entire Qur’an (which, to be fair, often is the actual principal objective) but not actually really knowing Arabic. It seems a missed opportunity …
Language learning surely entails a lot of rote memorisation (although “rote” is not really the right term for my own experience of how it works) but it is by no means “about” rote memorisation; any more than chess is “about” memorising openings.* And no amount of memorisation alone can make you a fluent language speaker or a capable chess player.
* I thought about putting this as “any more than cricket is about net practice” but on reflection decided that this analogy might be a bit culture-bound …
I think language learning is about rote memorisation anyway.
Not true at all; most language learning is via immersion. My youngest brother learned Spanish by watching hours and hours of TV in Argentina, and plenty of people learn English that way. Not to mention all the people who just get thrown into a monolingual environment and sink or swim. Your view of language learning is classroom-centric.
@LH, DE, you misunderstood me: I mean that natural process that we call “picking it up”, not the classroom.
“I heard how a native speaker used this word in this context”
=>
“I used it in a similar context myself” (possibly without a clear understanding of what it means)
=>
“after having heard and used it many times I developed a feeling”
This is imitation. But I agree with you – it possibly also involves creativity.
That’s certainly not what’s usually called rote memorization.
True. But still we parrot each other. What I meant is that memorization of texts is akin to what happens during normal communication.
It is not really too difficult to remember a poem. Sounds like a (1) cheap/efficient (2) pleasurable way to expand your vocabulary. I never did that on purpose, but I suspect the several Italian songs that I know by heart (1000+ unique words) must have positive effect on my ability to comprehend what they say on Italian TV (which in turn will have positive effect on the amount of time I will need to learn to speak Italian).
Of course “sticks” absolutely exclude “pleasurable”. For most people…
And of course it is not enough.