From: Piotr Gasiorowski
Message: 19932
Date: 2003-03-16
----- Original Message -----
From: "Michael J Smith" <lookwhoscross-eyednow@...>
To: <cybalist@yahoogroups.com>
Sent: Sunday, March 16, 2003 8:12 PM
Subject: [tied] areal and genetic groupings
> I just wanted to get this straight. Does an areal grouping strictly refer to a grouping of IE languages with geographical proximity to one another and nothing else? Or are they IE languages that have more linguistic features in common with one another than to any other IE languages, but with no direct genetic link?
It refers to a cluster of languages that are or used to be close to one another geographically and which show common _diffused_ features (as opposed to shared innovations inherited from a common ancestor. Not all common features (even inherited ones) suggest a genetic grouping: shared archaisms never do. They only prove that a language is IE, but tell us nothing about its position within the family.
> In other words, by saying that Balto-Slavic does not form a genetic grouping with Germanic means that there was never an IE branch Balto-Slavo-Germanic that then split into Germanic and Baltic and Slavic, but that as an areal grouping Germanic could be grouped with Baltic and Slavic because it has more linguistic features in common with Baltic and Slavic than other IE languages? Or would the areal grouping be because of the close geographical relationship and exchange of loan words, etc.?
Balto-Slavic and Germanic have been in direct contact for at least about two millennia, and even in more remote times they probably belonged to an areal grouping extending along the North European Plain. They share many lexical items and several grammatical features (e.g. a double system of adjectival declensions). These grammatical innovations are analogous rather than homologous, however, and any link between them is due to areal diffusion, not to common origin. Similar phonological innovations are either trivial (e.g. the merger of *a and *o as *a) or only partly parallel (the vocalisation of syllabic sonorants as *uR in Germanic and as *iR or less frequently *uR in Balto-Slavic).
Germanic has areal links with Balto-Slavic on the one hand and with Celtic (and perhaps Italic) on the other. One characteristic feature of areal traits is that they may connect a language with two or more different groups. If you have, say, five languages (A, B, C, D and E) in an areal configuration, some traits will be shared by A, B and C, others by B, C and E, still others by C, D and E, etc. There is no hierarchical clustering of the type that characterises genetic connections.
> Oh yeah, and does Hellenic, Armenian, Phrygian, Thracian and Indo-Iranian form a valid genetic grouping (once in a prehistoric time?), or do any of these languages share a genetic link with one another?
Again, any gramatical resemblances here look like diffused features rather than inheritance from a common ancestor.
Piotr