Re: The Great Proto-Indo-European Thread's Sequel
Posted: Mon Oct 05, 2020 2:15 am
perhaps a root of the problem then is that a great deal of 'mainstream' historical linguistics is bad and the presentation is worse - for an unrelated example, Wurm's Papuan groupings are extremely preliminary & admitted as such, but generally presented as accepted facts on par with families that have been established through any real methodology whatsoever
once you get into the weeds of actual words, such as Italian acqua, it becomes very obvious that careful scholarship is necessary for accuracy; although dumping shit into a meat grinder is a necessary step in the process of developing knowledge, the end product is still ground shit, and extensive further processing is necessary to turn it into water
the process of turning shit into water of course necessitates a pipeline with multiple steps, from "dude weed lmao" to exceptionally careful types who probably don't even quite believe in Afroasiatic or Sino-Tibetan yet; but it kind of helps if the people at the earlier stages are at least somewhat sober, and not Ehrets or Starostins. (or worse, Ruhlens.) the Greenbergite objection that the process of proto-language reconstruction necessitates some prior idea of what to compare with what is obviously correct for combinatorial reasons, but that doesn't justify Greenbergism except as the stage that grinds shit; and in some cases, even well-established ones like Afroasiatic or even Uralic (where apparently about every root is irregular somewhere), you may not ever quite get water. certainly not even IE is entirely neogrammarian; e.g. lingua, kantwo etc.
(and some established facts about IE are likely wrong and soon to be overturned - "everyone knows" that PIE *d palatalized to TB ś, but the only example is śak 'ten', which is imo better treated as irregular. there are various other sound correspondences that remain to be worked out - iirc Orel claims both *sw > d and *sw > v)
once you get into the weeds of actual words, such as Italian acqua, it becomes very obvious that careful scholarship is necessary for accuracy; although dumping shit into a meat grinder is a necessary step in the process of developing knowledge, the end product is still ground shit, and extensive further processing is necessary to turn it into water
the process of turning shit into water of course necessitates a pipeline with multiple steps, from "dude weed lmao" to exceptionally careful types who probably don't even quite believe in Afroasiatic or Sino-Tibetan yet; but it kind of helps if the people at the earlier stages are at least somewhat sober, and not Ehrets or Starostins. (or worse, Ruhlens.) the Greenbergite objection that the process of proto-language reconstruction necessitates some prior idea of what to compare with what is obviously correct for combinatorial reasons, but that doesn't justify Greenbergism except as the stage that grinds shit; and in some cases, even well-established ones like Afroasiatic or even Uralic (where apparently about every root is irregular somewhere), you may not ever quite get water. certainly not even IE is entirely neogrammarian; e.g. lingua, kantwo etc.
(and some established facts about IE are likely wrong and soon to be overturned - "everyone knows" that PIE *d palatalized to TB ś, but the only example is śak 'ten', which is imo better treated as irregular. there are various other sound correspondences that remain to be worked out - iirc Orel claims both *sw > d and *sw > v)