Before I go to the greatest hits, let me just say what a sorrow it is to click on the links to all those great blogs I’ve mentioned over the years and get “Site not found.” I miss them all, and I wish it were easier for people to keep their archives online so at least we could have the pleasure of reading them. Virtual lifetimes are even shorter than real-world ones.
IV: 2005-06
Muskogean and Lamb’s-quarters. A great discussion of Native American languages, foodstuffs, and plant names.
Alternate Nobels. Prompted an enjoyable discussion (kicked off by an intemperate Patrick White fan).
It’s Never That Simple. The complexities of Indonesia and the ways its history has been misunderstood.
Rutabaga/Bryukva. Much edifying discussion of root vegetables.
Dixon: The Word for Dog. A fascinating linguistic coincidence and a wide-ranging thread involving dingoes, quokkas, eggplants, and who knows what all.
The Multifarious Aubergine. An eggplant post inspired by the Dixon one.
Smoking Your Own. The fate of Bakhtin’s masterpiece.
Secret Languages of Hate. What languages do your relatives speak when they don’t want others to understand?
The Language of Command. One of those posts that changed the way I look at language and the world.


  1. Paul Clapham says

    I have to say, your blog has one of the best collections of commenters that I’ve seen. And I do read a lot of blogs.

  2. dearieme says

    A post about the languages of the indigenous Americans has just appeared here.

  3. marie-lucie says

    dearieme: This article is obviously not from a linguist, or even a scholar of some relevant discipline. It is true that Ruhlen had looked at resemblances between Na-Dene and Yeniseian (probably after reading Trombetti), but his multiple comparisons of vocabulary have almost all turned out to be wrong. The current work by Edward Vajda on the topic considers a lot more than just vocabulary, and IT has been accepted by well-known historical linguists fully familiar with the traditional methods. Vajda is also a fluent speaker of Russian and has been able to read the relevant Russian literature, as well as going to Siberia to do fieldwork among Ket speakers.
    Greenberg is well-known and still respected in linguistics for his work on typology, which is not at all the same type of research as that needed for classification and historical work. He and Ruhlen are also taking credit for more partial work done by their predecessors. For instance, in sub-classifying “Amerind”, Greenberg said something like “I basically accepted the classifications proposed by others”, even though some of those proposals have been demonstrated to be wrong and one of them was even a hoax! He did some of this in Africa too, basically building on the work of others, while (to his credit) discarding some wrong-headed results based on racial prejudices. There is nothing wrong with starting from the work of others, this is how we learn and progress, but one should recognize one’s predecessors, not denigrate them and pretend they did not do anything while accepting their results uncritically.

  4. What Paul said and marie-lucie proves.
    Ten years, eh? That’s like five centuries in Internet years. Thank you for each and every post and here’s to ten, twenty and thirty more!

  5. marie-lucie says

    correction: [Ruhlen’s] multiple comparisons of vocabulary
    In this case, “multiple comparison” is not the correct term, since Ruhlen compared a single language (Ket) with one language family (Athabaskan), but Ruhlen does agree with Greenberg’s general method and has been continuing his work in the same vein. Greenberg’s “method” of multiple comparison takes individual concepts (mostly common words such as natural objects, body parts, etc) and attempts to find equivalents of form and meaning (identifie by a very generous margin) from a wide variety of languages and families.
    That is fine as a preliminary assemblage of data (assuming that the forms are correctly recorded and analyzed, but Greenberg’s lists have an astonishingly high percentage of errors of many kinds), but even the most perfect list of words is not in itself sufficient to draw any conclusions.
    This method of vocabulary comparison does give good results for very closely related languages, such as French, Spanish and Italian, which also have common morphological features such as the complex structure of the verb: even if you don’t know that they descend from Latin, or you don’t know Latin at all, you don’t need special training in linguistics to classify these three languages as one group as opposed to, say, Russian or Sanskrit, and you can also correctly sub-classify Spanish and Portugues together. But this is much more difficult to do with more distantly related languages like English and Persian, which both belong to the Indo-European superfamily (as demonstrated by historical linguists) but not to the same lower-level families.
    As I mentioned elsewhere about “Amerind”, most of the language families of the Americas are lower-level ones on the order of Romance (Latin descendants) or Germanic or Slavie, and most of them were recognized in the mid-19th century, with a few higher-level ones such as Algic (Algonquin + two languages on the California coast) added later. Yet other higher-level groupings have been proposed (especially by the linguists Sapir and Swadesh) but not generally accepted (although I think that some of these proposals should be reconsidered), but Greenberg accepted practically all such proposals, and then topped the pyramid, or rather house of cards, with “Amerind”. From my point of view, the only thing the “Amerind” languages taken as a whole have in common is that they are neither Eskimo-Aleut or Athabaskan.

  6. Marie-Lucie: about Greenberg accepting “practically all such proposals”: to be maximally fair he himself stressed that the reality of Amerind was much more secure than the existence of any of its constituent “subgroups”.
    Also, to call his data error-ridden is frankly unfair, as Greenberg himself pointed out: most of his critics pointed to errors in his data by means of sources published after the publication of Greeenberg’s own LANGUAGE IN THE AMERICAS in 1987, or, even worse, of sources that had never been published.
    My own hunch (for whatever that is worth) is that Amerind (in the sense of a language family comprising a majority of the languages and/or language families of the New World, Na-Dene and Eskimo-Aleutian excluded) is real. And that whoever demonstrates its existence (Greenberg certainly did not) will need to take as basic building-blocks known and accepted language families of the Americas rather than fragile constructs such as the ones Greenberg worked with. And will also need to show that the common features used to demonstrate the existence of Amerind cannot be accounted for by means of (recent or ancient) language contact.

  7. marie-lucie says

    Etienne: Sorry, but “secure” and “real” are not terms I would associate with “Amerind”.
    As for the age of this alleged group (a topic brought up earlier), Johanna Nichols pointed out that its diversity was such that it would have taken perhaps 50,000 years to achieve it, a far greater length of time than the current estimate of the earliest entry into the continent.
    I think that the reason there is such a large number of acknowledged families in the Americas (amounting to much greater linguistic diversity than would be expected from the recentness of its population) is that most of the accepted relationships between the languages of these families are very shallow (the languages are very similar to each other), and that some genetic relations between families have not been recognized because they were not obvious. One linguist wrote: “A language family is either obvious, … or it is forever unknowable”. I find this attitude bizarre: there is plenty of space between knowing everything and knowing nothing, and even if the gap will always exist, scientists have to do whatever they can to reduce it.
    (similarly, in Indo-European, some languages or families took a while to be recognized, because of some features which obscured their relationships with each other and with the already classified IE languages).

  8. marie-lucie says

    (The last sentence was meant to be around the middle of the second paragraph – I lifted it out intending to replace it (with some editing) after the middle sentence, but as I could not see it in the writing window, I forgot about it.)

  9. Marie-Lucie: to be clear: I do NOT believe that Greenberg has demonstrated that Amerind is real. I am trying to point out that he himself had acknowledged much of what his critics subsequently blamed him for. The fragility of the subgroups of “Amerind” he was operating with being one.
    As for Johanna Nichols’ argument about “Amerind” being too diverse to be as young as the first settlement of the New World: if you are referring to her 1992 LANGUAGE article, I do not believe it.
    Not least, in fact, for a reason which she herself had pointed out in the article: we simply lack comparative data on the diversification rate of a language family spreading over such a huge area, where the language of the initial colonizers would quickly cover such a vast area that its daughter languages could no longer influence one another or be directly influenced by the same language(s).
    By comparison, consider the transition from Indo-European to the present-day Indo-European languages of Europe: all of these neighboring Indo-European languages have been exchanging morphemes and structures for most of their history. In like fashion present-day Arabic has spread at the expense of another Semitic language, Aramaic, which itself had spread at the expense of other Semitic languages, such as Akkadian and Hebrew…
    Now just imagine an alternate history of the Americas in which the first settlers in the Americas are speakers of Indo-European languages.
    Let’s have speakers of Koine Greek settle around Argentina, speakers of Latin settle in Central Mexico, and speakers of Proto-Germanic settle in coastal British Columbia. I do not and cannot know what such alternate Hellenic, Romance and Germanic families would have looked like 2000 years later, but I am certain of one thing: these three branches of Indo-European, completely cut off from one another, would have differed far more radically from one another in this alternate history version of the Americas than they do back here in the real world.

Speak Your Mind