24 Mar 2013

Analytic Proto-Bantu?

In 2007 Derek Nurse asked: Did the Proto-Bantu verb have a synthetic or an analytic structure? He came to the conclusion that it was originally analytic. Proto-Bantu is the originator of several central and south African languages including Swahili, Xhosa and Zulu.


  1. It seems to me that most reconstructed language families are proposed to have more synthetic structures, from proto-Sino-Tibetan to proto-Austroasiatic and so on and so on. It seems refreshing to find one unattested protolanguage that is posited to be analytic. Does that underlie an academic (maybe justified) bias, or is there something to that? I might simply be focusing on the wrong details, but I've seen people arguing (my teachers mostly) that Proto-Human (not even known to feasibly have existed in any meaningful sense) "must" have been very highly synthetic because so many of the (at least the popular) language families appear to be on an at-large "analytic-alizing" trend. Of course, it's a matter of degree, and I'm no expert, but to your knowledge is there anything to explain that trend?

    If I had to wager, I'd probably bet that my idea of an average degree of syntheticness/analyticalness is coloured by my being speaker of only a few western European languages with negligible experience in Chinese reinforcing the perception, and the protolanguages I've read about are typologically actually closer to average. But, you know, figured it was worth asking.

  2. It's of course a potential trap for any of us to be unintentionally biased towards certain models of language based on our narrow exposure to different possibilities. We're only human.

    However the terms "analytic" and "synthetic" are not pure states of a language. A language will fall in a scale somewhere in-between these two ideals in some way or another. Mandarin, for example, might be intrepreted by some as more analytic than English because it doesn't conjugate verb for person, number or tense with affixes as often is the case in European languages. That said, Mandarin grammar is just as rich and nuanced as English albeit with different grammatical rules. It contains aspect markers that could just as well be considered suffixes instead of free particles. Do we really have an analytic construction qu le 了 (= "went, has gone, starts to go") with separate morphemes or is it to be analyzed synthetically as qu-le? The latter analysis has some weight considering that the aspect marker is never used in isolation without a preceding verb. It depends on how you look at it. In my experience, most languages have some sort of inflection even when labeled "analytic", including English (eg. washing) and Mandarin (eg. qu-le).

  3. There is an important difference between Mandarin qù-le and English washing: You can insert words--even whole phrases--between qù and le, but not between wash- and -ing. For example, you can say 去看电影了 (qù kàn diànyǐng le) "went to see a movie", but you cannot say *"wash the dishes -ing".

    I do have a bias, though. In fact, I have a whole crackpot theory about the evolutionary origin of the difference between analytic and synthetic processes: http://sektu.blogspot.com/p/pre-human-language.html.

  4. Apologies in advance because this comment has nothing to do with this specific post... but after years of frustration with ridiculous fake Ancient Egyptian word spellings with e's haphazardly thrown in for every missing vowel (and some things not even based on Egyptian spellings at all, but GREEK!) and spending the last couple weeks hopelessly searching the internet (I don't have easy access to any academic libraries, nor would I frankly even know where to start if I did considering how pathetically little has been accomplished on the subject) for any kind of reconstruction dictionary or word list at all (let alone a decent one), I stumbled upon your blog.

    Needless to say, I just spent several hours reading through your entire archive of posts regarding reconstruction of Ancient Egyptian, and based on my relatively decent knowledge of historical and comparative linguistics (it wasn't my degree specialty, but I did do some in my spare time) and what (admittedly little) I remember from my studying Egyptian in college, your reconstructions and analyses seem by far the most sound and well thought out (in particular, I jive with your idea of keeping it simple instead of baselessly over-elaborating and over-complicating, as well as your consistent comparisons to the Coptic, of which I admittedly know nothing but what I've read on other sites regarding changes from Egyptian to Coptic) of any that I have seen.

    So I saw in one of your earlier posts that you had started keeping yourself some kind of list of reconstructions that you had come up with, but I don't see any kind of list on the blog (sorry if I'm just blind and didn't see it... but I thought I looked pretty thoroughly!), and I was wondering if you would ever consider making a list like that available to the public.

    The absolute dream for my study of Ancient Egyptian is to have a dictionary list with the hieroglyphs, transliteration, a good reconstruction with Latin spelling and IPA pronunciation, and an English translation, with bonus points for side notes showing representations of the word in other sources aiding in reconstruction (cuneiform, Coptic, other proto-language reconstructions, etc). If I had knowledge of Coptic and other ancient languages from the area and thought I could do any kind of intelligent reconstruction on my own I would (given my experience in historical linguistics), but I know my limits, and unfortunately I just don't have the skillset, Now, I can't tell from your posts if you've gotten that much information together yourself (it's an ultimate wishlist, I know) or have even done that many different word reconstructions at all, but I would certainly be interested in seeing in a concise form any and all reconstructions that you have done if you'd be open to sharing them, because as I said, I think they are very sound and well-thought out... and frankly at this point I'm desperate enough to not even care if it's a perfect or near-perfect pronunciation, as long as it's BETTER!

    (As a semi-aside, I would also be interested in any thoughts you might have about shifts in pronunciation between Old and Middle Egyptian. I am particularly interested in the Middle and Old Egyptian pronunciations for the names of gods and have come up almost entirely blank on insights into that topic online, so it would be great if you had any thoughts or possibilities to share.)

  5. Brian: "For example, you can say 去看电影了 (qù kàn diànyǐng le) 'went to see a movie', but you cannot say *'wash the dishes -ing'."

    Good point but then there is the matter of verbal le versus sentential le (see Soh, Gao. Perfective Aspect and Transition in Mandarin Chinese: An Analysis of Double –le Sentences. 2006).

    From what I read, 去看电影了 (qu kan dianying le) with sentence-final le reads more like "has gone to see a movie" (ie. remarking on the past experience of watching the movie) as opposed to 去了看电影 (qu le kan dianying) "went to see a movie" (ie. remarking on a completed event). This way, our heads won't short circuit when encountering double instances of the two types of le as in 去了看电影了 (qu le kan dianying le).

    1. Thanks for the reference to this article. I never thought of post-verbal 了 being different from sentence-final 了, but I think this must be right. My first Chinese teacher used to say that double-了 was "cute", so I try to avoid it, but I've never seen an explanation of why it is permissible at all.

  6. Sorry, perhaps concerning the sentential particle, I should ammend that to "ie. remarking on the past experience of going to watch the movie". Alas, either way I could still be wrong! Mandarin even confuses Mandarin speakers. ;o)

  7. Em: "The absolute dream for my study of Ancient Egyptian is to have a dictionary list with the hieroglyphs, transliteration, a good reconstruction with Latin spelling and IPA pronunciation, and an English translation, with bonus points for side notes showing representations of the word in other sources aiding in reconstruction"

    Definitely! Wiktionary is the closest thing I know of so far but its interface is so clumsy and hard to wade through.

    1. For just looking up dictionary terms I find wiktionary pretty much the clumsiest format out there, especially since searching for Egyptian words with symbols that aren't normal English letters is a complete pain in the ass. When I use it, I usually have to search for the English word then hope the Egyptian word is in the translations section.

      Once you're on a page it does tend to have the Coptic version of the word, which I'm guessing is where you're coming from with it being useful, but I've seen maybe two or three pages ever that actually have a reconstruction attempt on them.

      Needless to say, the fact that this is one of the closest to ideal options available says pretty much everything about the state of Egyptian linguistics. Given the number of posited Proto-languages out there for which there are numerous lists of reconstructed words, it's frankly ludicrous that scholars are either too lazy or just too un-knowledgeable to accomplish this for Egyptian, and I say this as someone who fits securely into the latter category. Which is why I was so excited when I found your reconstructions and they seemed rather sound. So I was really serious when I said I'd be totally interested in seeing any kind of list of posited reconstructions that you might have, and if you were at all willing to share them en masse that would be mega awesome.

      On a wholly related note, I feel compelled at this point to ask your opinion on theories that Cushitic languages are either related to or direct descendants of Egyptian. Personally I think it makes plenty of geographic sense and I've seen enough isolated samples to think it's at least a possibility though I obviously don't know enough (and by not enough I mean I found a note about them in mybsearches for reconstructions and then went and read a couple wikipedia pages) about any of those languages to feel in any way confident saying that those isolated similarities that I've seen pointed out indicate actual relationship and not just a lot of word borrowing between languages in close proximity. But then obviously if any of these languages ARE in fact related to Egyptian it would give anyone willing to learn one or more of them a serious leg up in reconstructing Egyptian, or at the very least some kind of proto-Egyptian language.

  8. What you say is right on the money about typing in awkward symbols. The web designer must make a system with the *user* in mind but this often doesn't happen. This is why I programmed my Etruscan dictionary to simultaneously recognize different ways of entering the same word. One can A) type the Greek letters directly; B) use 'ch', 'ph' or 'th' instead; or C) just press the extra buttons labeled with the Greek letters in the applet. It also recognizes all attested spellings of a particular Etruscan word put into my database, not just a canonical form. I think programs should always be designed to meet the user half way to do most of the grunt work for us. Easy peasy.

    Regarding Cushitic, it is treated as its own branch of the Afro-Asiatic language family, separate from Egyptian, Berber and Semitic. Some Egyptian loanwords are to be expected due to historical reasons but Cushitic is not *from* Egyptian at all.

    Proto-Cushitic though usually lies outside of my focus, ie. the Mediterranean. This is why my interests had expanded to Berber, Egyptian and Semitic since they participated directly in that elaborate sea-trade network of yore and Etruscans traded with Punic Phoenicians from Carthage. I wish everyone stopped writing yet one more dusty journal article and simply shared their ongoing thoughts and research online. I believe gradually everyone will have to treat their research in a more open-source fashion like this one way or another. This is where the Computer Science department must come in, to help facilitate research in all the other fields using intuitive interfaces that encourage collective-managed collaboration. I find it unlikely though that traditional institutions are going to be the ones to usher in this new era; it will be likely a grassroots thing. We not only need an open platform but a system that adequately incentivizes contribution and disincentivizes petty vandalism, but in order to get to that level, we have to let go of the sacred cows of the previous generation (hierarchy, bureaucracy, subordination, credentialism, statism, the false teacher-student dichotomy, the permanence of facts, etc.). When this new, more maleable organizational theory is finally put into action instead of resisted or demonized by the status quo, networks of like-minded and equal individuals will replace old hierarchical institutions whose organizational structures are too brittle to adapt, due to their characteristic power bottlenecks.

    Wolfram's Alpha has an interesting platform, but I wish:
    1) it was more intuitive to the average user,
    2) accepted input from anyone as Wikipedia does,
    3) allowed voting of all contributions like Youtube & Reddit do,
    4) enabled *users* to help design and expand the website itself,
    5) empowered collectivized site policing rather than through appointed admins,
    6) enforced total user anonymity to strip away collusion and petty personal attacks,
    7) gave users monetary incentive for good contributions, perhaps using Bitcoin and sponsors.

    That so far is the "perfect system" that I have in mind thus far. Programming all that though would be quite a wizard's act! <:S