The "most X" language of family X

[dɮ the phoneme]

When learning about a language family, I often get the impression that some languages are more characteristic of that family than others, because they have a particularly high proportion of features that make that family unique. This is, of course, not exactly a scientific notion, it's purely a subjective judgement. But I'm curious what people's judgements are on this topic.

I got thinking about this because it occurred to me that Sanskrit was, in some sense, the "most IE" of the IE languages. Or perhaps the "most old-IE". It has so many of the features that stand out to me as so characteristic of IE: the vowel gradations, the mess of a verbal system, the syllabic consonants, the conjugations, etc. This is not to say that it's the most representative IE language, and certainly not among modern IE. But it gives one, rather, a uniquely concentrated dose of the "IE essence". Similarly with Arabic for Semitic, I think.

So, I'm curious if anyone who has studied another language families has a sense of which languages are the most characteristic of that family? Is there, for instance, a "most Algonquian" language? "Most Bantu"? Etc.

[zompist]

I got thinking about this because it occurred to me that Sanskrit was, in some sense, the "most IE" of the IE languages. Or perhaps the "most old-IE". It has so many of the features that stand out to me as so characteristic of IE: the vowel gradations, the mess of a verbal system, the syllabic consonants, the conjugations, etc. This is not to say that it's the most representative IE language, and certainly not among modern IE. But it gives one, rather, a uniquely concentrated dose of the "IE essence". Similarly with Arabic for Semitic, I think.

This is a bit circular, since the traditional PIE reconstruction is heavily influenced by Sanskrit-- way too much so, according to Lehmann. There was an attitude that things could decay but not be innovated, so anything in Sanskrit got tossed into PIE. The discovery of Hittite should have upended this, but some IE scholars still sideline Hittite.

[dɮ the phoneme]

This is a bit circular, since the traditional PIE reconstruction is heavily influenced by Sanskrit-- way too much so, according to Lehmann. There was an attitude that things could decay but not be innovated, so anything in Sanskrit got tossed into PIE. The discovery of Hittite should have upended this, but some IE scholars still sideline Hittite.

Interesting. I knew early reconstructions were Sanskrit-centric, but I didn't realize people argued that it was still true. What features of the PIE reconstruction are inaccurate, according to Lehmann?

[zompist]

Interesting. I knew early reconstructions were Sanskrit-centric, but I didn't realize people argued that it was still true. What features of the PIE reconstruction are inaccurate, according to Lehmann?

Oh, most of the morphology. I'd have to re-read Theoretical Base of Indo-European Linguistics to report accurately, but he thinks most of the cases and some of the verbal paradigms are late, shared only by a core of Greek + Indo-Iranian. (I forget what else is in there.) He also thinks there were only 3 sets of stops, but I think most modern IE-ists agree on that one.

[Creyeditor]

Swahili is NOT the most Bantu, it is probably among the least Bantu. Maybe Chichewa or Kinande are in the top five if you exclude Nguni languages.

Iau and Skou propably among the most Papuan, even though this is of course more of an areal thing.

[bradrn]

Iau and Skou propably among the most Papuan, even though this is of course more of an areal thing.

‘Most Papuan’ doesn’t really make sense to me given the typological diversity of that area. (Iau isn’t even the most Lakes–Plains, let alone the most Papuan). Furthermore we have a natural bias towards better-described languages, given that many large families have only one or two languages which are really well-described. That being said, I’d suggest that better candidates might be Kalam for TNG and Wutung for Skouic.

[alice]

Maybe "Most Foobaric language" means "has fewest features in common with non-Foobaric languages"?

[Kuchigakatai]

I vote for Portuguese as "most Romance" of the Romance languages. The phonetic differences between European and Brazilian Portuguese even come up as useful in terms of forming an "average".

[Creyeditor]

Iau and Skou propably among the most Papuan, even though this is of course more of an areal thing.

‘Most Papuan’ doesn’t really make sense to me given the typological diversity of that area. (Iau isn’t even the most Lakes–Plains, let alone the most Papuan). Furthermore we have a natural bias towards better-described languages, given that many large families have only one or two languages which are really well-described. That being said, I’d suggest that better candidates might be Kalam for TNG and Wutung for Skouic.

I definitely agree with you on Skouic

[Moose-tache]

Okinawan is the most Japonic of the Japonic languages, as it's speakers are the most likely to be centenarians.

[FlamyobatRudki]

But what language is most klingon?
Then we splice the language with klingon hol and we make paramount pay for this language…… a sufficient amount for our entire bloodline to never have to work again! == Qapla!

[abahot]

I got thinking about this because it occurred to me that Sanskrit was, in some sense, the "most IE" of the IE languages.

I know I'm a bit late to the party, but I think this is just because Sanskrit is one of the most archaic. It has a lot of old features, but I'm willing to bet that out of all IE dialects spoken around the same timeframe (1500 BC), Sanskrit was not particularly conservative. If I had to guess, I would say something pre-Celtic would seem the "most IE" out of the dialects contemporary with Vedic Sanskrit.