# Selections from Tumblr

I have a Tumblr blog which I use for writing short-form things that aren’t necessarily of any lasting value. But occasionally things do end up there that might be worth reading, so I intend to make an organized list of links to Tumblr posts that might be interesting to readers of this blog every year or so. The last time I did this was in December 2015 (here on WordPress and here on Tumblr), and I have been posting on Tumblr at a higher rate since then, so the list in this post is rather long, and I’ve organized it into subsections to make it more manageable. Only posts from December 2015 are included; for earlier posts, see the earlier lists.

## Miscellaneous

### 7 responses to “Selections from Tumblr”

1. David Marjanović

Yay! So much to read! ^_^

On the horribleness of American universities, rest assured that America is the odd one out there.

From “On Latin trahere“:

NE throw < OE þrāwan, with -w- already (unlike draw and drag, which < dragan), […] Actually, it turns out the -w- is a hiatus-breaking consonant—the PGmc form was *þrēaną, which would have become*þrāaną in Northwest Germanic. I don’t know why it becomes drehen in German, but I don’t know the German vowel developments. It was still drāen in Old High German, so it looks like there was no w-epenthesis in that language.

*lightbulb moment*

In southeastern German dialects like mine, dreh- is /dra/-. I’ve been wondering why for most of my life until now. Behold the almost complete solution:

First, drehen has two meanings, transitive and intransitive, just like English turn or indeed French tourner. What if the transitive one is the causative of the intransitive one? Then we’d be looking at one of many, many other instances where a verb has merged with its own causative, and if dreh- were spelled etymologically, it would be drä-. In the dialects in question, /a/ is the umlaut of *a & (which in turn are now both /ɒ/) maybe half of the time; I have no idea of the conditioning factors.

This would imply that the word has either been lost in those western dialects that haven’t merged /æː/ into /eː/, or that whoever fixed the spelling happened not to know such a dialect. That happens sometimes; the clearest example I know of a word entering Standard German with an etymologically wrong phoneme in it is jäh “suddenly”, which is /gax/ in dialects like mine, showing that the word got into the standard from one of the Central German dialects that turned into /j/ instead of /g/.

• David Marjanović

Clarification: “one of the Central German dialects” like those of Cologne or Berlin; most have had different developments of .

Also, it’s funny how the h is wholly spurious in dreh-, but etymological in jäh.

2. David Marjanović

I’m cited in “On the Proto-Indo-European velar series”. 🙂 My sources were probably Wikipedia and/or the preface of the North Caucasian Etymological Dictionary; I’ll try to look stuff up at some point.

The difference between palatalized velars and actual palatals has confused a lot of people, which surprises me, because they’re easy to tell apart even though they tend not to occur in the same language. Russian is full of [kʲ gʲ], Polish has them too, and so do mainstream Greek ([kʲ] anyway) and some Turkish accents. Actual palatals sound much more like palatalized alveolars; in Europe they occur in Hungarian (ty gy ny, and apparently still ly in very conservative accents) and Latvian (ķ ģ ņ ļ), not to mention that [ɲ ~ ɲj] is all over Romance and [ʎ ~ ʎj] persists in Catalan, Portuguese and a few others.

I do wonder if the kentum merger is a recurring substrate effect that happened independently in Hittite, Tocharian, Greek, and once or twice in “West IE” (Germanic + Italo-Celtic) depending on where Albanian belongs.

3. David Marjanović

Concerning “On the early development of the Indo-Aryan languages”, it has apparently become mainstream that even the earliest stage of Rgvedic is not Proto-Indo-Aryan. Where Vedic has a megamerger of, like, six different clusters into kṣ, various Prakrits and later stages keep some of its sources apart, notably the voiced and voiceless ones.

The idea exists that Vedic is actually ancestral to Dardic and only Dardic. I have no idea.

Catalan, Portuguese and a few others

Notably Italian.

4. David Marjanović

“On cognacy of ‘food’ and ‘corpse'”: German Aas – “carrion” < “food”.

• David Marjanović

(The meaning “food” is extinct, the meaning “carrion” rather technical.)

5. David Marjanović

“On ostrich“:

The French word is supposed to have been formed from classical Latin avis ‘bird’ + strūthiō ‘ostrich’. This looks like a head-initial compound (in English it would probably have to be ostrichbird), but maybe head-initial compounds were common in Old French, I don’t know (it wouldn’t be surprising since it shifted from adjective-noun to noun-adjective order).

Are you sure about adjective-noun order? Latin had noun-adjective order as a default, though the slightest amount of emphasis was enough to override the default, and poetry didn’t care about word order at all as long as the prepositions didn’t become postpositions.

The few compounds found in modern French are all head-initial, and the pattern is mildly productive.