I'm an amateur linguist and recently wrote a paper called "The relationship between Proto-Indo-European and Proto-Yeniseian" which mostly comprises a short history of the Yeniseian language family, some typology and a list of 74 cognates between these two families. I submitted it to a couple of journals and (obviously) none accepted it, because it's not done by linguistics standards. But I also came across this paper by Merrit Ruhlen called "The origin of the Na-Dene", which focuses on what some would call "random look-alike words" (similar to what I did), and isn't what scientists would mention as actual, elaborate linguistic proof. And yet, his paper got published in PNAS. How can I improve my paper? Do you have any advice? Here's the link to my paper: https://ling.auf.net/lingbuzz/005734 and ruhlen's: https://web.archive.org/web/20110708074324/http://www.merrittruhlen.com/files/1998c.pdf , if anyone is interested. Thanks.

  • 9
    This is a question that's answerable here, but I'd recommend rewording it: focus on "why do journals apparently not consider lists of potential cognates to be evidence of relation, when Ruhlen etc have done exactly that?" rather than why your particular paper was rejected.
    – Draconis
    Feb 20, 2021 at 2:51
  • I would think that journals have a higher standard for papers that attempt to broaden the most studied language family in the field, as opposed to papers proposing a connection between two relatively obscure families
    – b a
    Feb 20, 2021 at 19:29
  • 2
    I take it your question is supposed to be something like "are the claims in this paper scientifically valid". Suggestions for improving manuscripts or term papers are off-topic as matters of opinion and not about the substance of linguistics. I think you should revise the question to clarify what the linguistic question is.
    – user6726
    Feb 20, 2021 at 20:37
  • I've not had a chance to read either paper yet but it's worth noting that Ruhlen's paper was published 23 years ago. I dare say standards have changed somewhat since then. It also seem to've been published in an anthropology rather than linguistics journal.
    – Miztli
    Feb 20, 2021 at 22:42

3 Answers 3


From your paper:

Chance is ruled out by probability, because two unrelated language families can’t have 74 accidental resemblances.

The problem is, this simply isn't true. Here are 109 accidental resemblances between Mandarin and English.

Ruhlen makes a similar claim in his paper, but he's similarly wrong. Look at Vajda 2010 for a specific rebuttal of his methodology. In particular:

Random similarities in basic vocabulary are insufficient to demonstrate language relatedness. A list of look-alike words can be compiled, even using basic vocabulary, between any human languages.

Systematic correspondences are needed to establish relatedness (and that's what Vajda tries to do); lists of similar-looking words aren't enough.

  • Actually, after rereading, I understand that you meant that it's wrong that it's impossible for there to be 36+ chance coincidences between languages, which is (more or less) demonstrably true. The first time I read, I had thought you meant the connection between the languages is wrong, but it turns out that Vajda accepts that but not the methodology
    – b a
    Feb 20, 2021 at 19:48
  • 3
    @ba Ah, right. Yes, my intent is to say that Ruhlen is incorrect in claiming that a few dozen chance resemblances are statistically impossible; I don't know enough about the languages in question to say if his conclusion is correct or not.
    – Draconis
    Feb 20, 2021 at 19:56
  • The link would be good if it adressed each etymology individually. A few of those terms have no established etymology on the English side, a few are loan words, and in some cases alternatives could be brought up so it looks as if they didn't even try. I don't trust they are qualified to judge Chinese etymologies, and I'm rather skeptical of the whole field (the majority will be carried out in Chinese if course, with all the ideological baggage that is difficult to avoid, and the west has to rely on it to a degree. The writing system is of course a huge impediment and trife with folk etymology)
    – vectory
    Feb 21, 2021 at 22:42
  • anyway, chinese was not in question.
    – vectory
    Feb 21, 2021 at 22:42
  • 2
    @vectory I'm responding specifically to the claim that "two unrelated language families can’t have 74 accidental resemblances". English and (Mandarin) Chinese are widely accepted to belong to unrelated families.
    – Draconis
    Feb 21, 2021 at 22:51

What I'm going to say might sound a bit depressive to you, but I think from experience that your paper has no chance of being published, no matter what.

  1. The first problem is that the relationship PIE vs PY is gigantic and extremely ambitious, and it cannot be adressed in a paper. It's more a topic for a (huge) monograph.
  2. Besides, you have to discuss the other proposals like PY vs Caucasic, and again this cannot be done within the limited space of a paper.
  3. I cannot think of any Journal that would accept that kind of paper. As a rule, Journals do not like extremely innovative papers and rather like sober and uncontroversial papers.
  4. People who have worked on Yenisseian do not favor a relationship with PIE, so it is probable that peer-review of your paper will be fairly hostile.
  5. Incidentally, your paper is weak as regards contents and methodology, but I think this is not the worst issue.

If you have an account on Academia.edu, you might consider opening a session with comments to see what other people have to say about your paper.

  • Point 5 seems rather snide, as there is no methodology.
    – user31182
    Feb 21, 2021 at 19:29
  • 1
    Well, I don't want to be exaggeratedly severe. So it's not snide, rather euphemistic.
    – user23769
    Feb 22, 2021 at 4:58

Other people already commented on your difficulties regarding publication, but you probably would like to receive a feedback regarding the essence of your paper.

I am sorry to disappoint you, but from the table of personal pronouns it seems that Yenissean are not even Mitian languages, those which are characterized with 1st person singular pronouns starting with m- (as in English me) and 2nd person singular pronouns starting with t- (as in English thou).

If there is any relation between the two families, it is far beyond the Mitian node.

  • There must be more to it, many IE language families do not have a m- 1st person plural. They often do have, however, some kind of m in the 1st person verbal forms. Feb 20, 2021 at 12:54
  • 2
    @VladimirF they should not have, 1st person plural was we/ne in PIE. 1st person singular nominative was eghom, but it is a PIE innovation, the older form mi(n) was kept in oblique cases.
    – Anixx
    Feb 20, 2021 at 13:45
  • 2
    taking a mitian node as axiomatic is extremely shoddy methodologically
    – Tristan
    Feb 22, 2021 at 11:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.