The only trouble with internet forums is that it is hard to resolve disputes like this the way they should be resolved—with a fierce argument over a bottle of good scotch.

Somehow just having the argument isn’t the same.

Second, with all due deference to the right of specialists to develop specialized vocabulary, construing “related” to exclude such pairs is a piss-poor choice* that invites confusion, since it conflicts violently with the common meanings of the word ("1. Being connected; associated. 2. Connected by kinship, common origin, or marriage.” AHD, emphasis mine.) Saying father and Vater can be called related but mill and muileann cannot seems like saying that the Wilton brothers are not related to each other but Washoe the chimp and I are.

Third, springing the specialist’s restricted sense into the discussion in the way you did is like telling a programmer who has remarked that he has a lot of work to do that he’s wrong because “work” is force times distance and working at a keyboard involves only small forces acting over small distances. Or like Dr. Doowop’s objection to “delusion” here, except that it’s plausible that he was unaware of the non-specialist sense of “delusion”; “related” is too common a word for you to be unaware of the common senses.

Oh, I quite agree, and in general I try to use commonly understood vocabulary rather than obscure specialist terms (and “phonetic respellings” rather than IPA, which is gibberish to most nonspecialists).  My problem here is that I didn’t realize my use of “cognate” was so counterintuitive to most people; now that I know, I will try to avoid it.

The case of father is one where we can state that it is true with extreme confidence, similar to that with which a physicist talks about relativity, but this is not the case with all PIE roots.

Of course not, that’s why I picked a root for which it can be stated with confidence.  I’m not trying to get into a detailed discussion of the ins and outs of comparative reconstruction, I’m trying to establish a very basic principle, the difference between inheritance and borrowing.  And Dr. T, with all due respect, the connection between mill and muileann is not “somewhere in between.” The fact that both words have undergone phonetic change since they were borrowed is neither here nor there; the important fact is that they were borrowed in the first instance.  No amount of phonetic change or analogical reconstruction can change that fact.  Neither word is inherited, and that’s what I’m trying to get firmly entered into people’s consciousness.

My problem here is that I didn’t realize my use of “cognate” was so counterintuitive to most people;

It was “related”, not “cognate”, that caused all the trouble.

That’s what I meant.  My mind is going fast.  (I don’t think putting on a tie yesterday did my brain any good.  Cuts off circulation to the head.)

languagehat - 22 March 2008 05:18 AM

Oh dear.

On the “related” front:

To use the “Joe and I” analogy, it is like saying “Yes, Joe and I have the same grandfather, but we’re not related.”

But these words (mill, muileann) don’t have the same grandfather!  They’re borrowed from the same word!  Why is it so hard to see this distinction?

I don’t think anyone here is confused about the distinction.  The confusion lies in finding a concise and clear way to characterize it.  We are trying to do it by analogy, with mixed success.  Going back to the “Joe and I” analogy, the step of a word being borrowed is (to me) like adoption.  Joe and I have the same grandfather, but Joe’s father was adopted by someone else entirely.  Are Joe and I related?  Perhaps not in every sense, but certainly in some.  The analogy is unclear, hence the ensuing whackiness.

Dave Wilton - 23 March 2008 07:09 AM

The burden of proof is always on the person making the positive assertion, not on others to disprove. This is true in every single academic and professional discipline. If you make a statement, you have to provide sufficient evidence to convince reasonable people that it is true.

Fair enough; however in the case of the name Adonis, the reference I provided is the Etymologicum Magnum. Hat dismissed the reference by claiming that it is of no “specific value” because it is a medieval work. The burden of proof on it being of “value” despite being dated is certainly not mine .

JimWilton - 23 March 2008 12:23 PM

The only trouble with internet forums is that it is hard to resolve disputes like this the way they should be resolved—with a fierce argument over a bottle of good scotch.

I’ll drink to that!

The fact that Joe and I share a common grandfather is of no scientific importance, although it is firmly estabished that we are descendants of a hypothetical fellow called Stanley who marauded somewhere in the steppes between Europe and India several millennia ago - which makes this relationship of theirs not too interesting…

I have often seen scholarly variations of this argument in a few posts above.

Useful and elegant as the PIE conventional wisdom is, we must not lose perspective of the fact that it is just a working hypothesis which elegantly explains common features in several branches of language. Actually, “common PIE roots” are in fact a reconstructed hypothetical pre-historical “Esperanto,” and many people sometimes take for granted that there indeed was such a people speaking that language - I am however not aware of any historical evidence that such a people ever existed, please correct me if I’m wrong.

I too often find PIE roots very interesting. However, it seems to me that many of us often avoid the challenge (and responsibility) of attributing the origin / etymology of a word to a specific historical language, but instead conveniently allude to a retroactively reconstructed PIE root. Its a rather facile approach which may be dubbed reductio ad PIE.

I don’t have a problem with new etymologies, but I would like to know (1) what the old ones are (if any) and (2) why the new one is more compelling. Take the proper names Adrian. Hadrian, Adrienne, etc. Saying that it is “Calm, like the Adriatic sea. From a-, negative prefix, and dro, to act” or “Firm. From a-, negative prefix, and dro, to act.” doesn’t convince me. (I’ll leave to one side that the female version of the name has the same origin but a different meaning. Why are females calm, but males firm? And what about that problematic h? If there is an initial h in the Greek original, then it can’t be the negative prefix a-, because a word-initial h in Greek is usually from an s (cf. helios ‘sun’ Latin sol, Sanskrit surya, halus ‘salt’, Latin sal) or occasionally a w (hudor ‘water’). It just might very well be non-etymological, but it would be nice in any new etymology to address it. What is the commonly accepted etymology of Hadrian. That it’s from the Roman emperor Hadrian’s name. His ancestors were from Hatria or Hadria (Atri) in Picenum (Abruzzo). This Hatria is not the only city on the Adriatic with this name. There is another Hatria (Atria), an Etruscan city, farther to the north in Venetia (Veneto). It is this city that the common consensus holds is the origin of the name of the Adriatic Sea. Interestingly enough, the Romans called this sea the mare Superum (upper sea), but the Greeks called it the pelagos ‘sea’ (or sometimes thelatta ‘sea’, kolpos ‘gulf’) Adrias (or Ariatikos). I would also like to see what the Greeks called the other seas (or gulfs) that the Mediterranean was divided into: e.g., the Tyrrhenian Sea (named after the Etruscans), the Corinthian Sea (named after Corinth), the Sicilian (Sikelias) Sea (named after the island of Sicily), and the Ionian Sea. This last sea’s spelling is problematic in Greek. Is the o an omikron or an omega? Ancient texts have both (see Herodotus, Polybius, and Strabo). There is an origin myth about the Ionian Sea being named after the goddess Io, but then there were also Ionian Greeks. I think it far likelier that a gulf or small sea would be named after a city, province/country, or a people before positing a compound which is unattested in an other IE language. As far as proper names go, one would have to explain why those names did not come from the famous Roman emperor or the two 4th century martyrs who bore his name. But back to what a possible etymology of Etruscan or Picene hatria might be. Some connect it to an Illyrian root adru ‘water’ which may be connected to Pokorny’s etymology (in IEW 4. *ad(u)-, *ad-ro- ‘current (of water)’). Could this root be connected with PIE root for water whence English water, Latin unda ‘wave’, and Greek hudor ‘water’ (IEW 78ff.). In the end, I do not state that Pavlos’ etymology is wrong and the commonly accepted one right, but I find the latter more believable than the former.

Thanks jheem for your well thought response, I learned a lot by reading it!

The Latin version of Greek words whose first letter has a daseia (rough breathing), do in fact begin with an S, H, V or W as you mention. ΥΠΕΡ becomes HYPER or SUPER, ΕΣΠΕΡΙΑ becomes VESPER, etc.

This makes me quite stumped with the etymology of town Adria:

* If Adria is related to ΥΔΩΡ that would explain the H in Hadrian, but it would fly in the face of the fact that ΑΔΡΙΑΝΟΣ does not have a rough breathing in Greek.

* On the other hand, the explanation I propose in my web site (a-, dro) does not explain to the H in Hadrian.

The plot thickens, back to the drawing board!