In their Etymologische Studien zum Jiddischen (Helmut Buske, Hamburg, 2006) Erika Timm and Gustav Adolf Beckmann discuss the theory of Paul Wexler that Yiddish is a relexified Slavic language. By this, Wexler means that Yiddish started out life as a member of the Slavic family and borrowed so many words from German that it has taken on a Germanic appearance. Timm and Beckmann reject this theory and they begin their concluding remarks by asking how it is that Yiddish has a very un-Slavic (and they add, un-Turkic) uvular [R]?
That the predominant form of the Yiddish /r/ is uvular can be seen from Map 8 in the first volume of the Language and Culture Atlas of Ashkenazic Jewry (LCAAJ) (Tũbingen, Max Niemeyer, New York, the Yivo Institute for Jewish Research, 1992). Yiddish has two kinds of /r/ which are distinguished by the position of tongue contact. In the apical [r] the tongue contact is further forward in the mouth while in uvular [R] it is further back. The bulk of Yiddish speaking territory has [R] with [r] found only at the edges on the northwest, the northeast, the southwest and, a bit, in the southeast.The pattern with [R] firmly established in the middle while on the edges of the territory Yiddish agrees with co-territorial languages in having [r] stongly implies that [R] is the oldest form in Yiddish and that [r] is a new form borrowed from its neighbors.
But if [R] is the original Yiddish /r/, we have to ask where it comes from? Slavic languages can be ruled out as overwhelmingly preferring [r]. The possibility that it comes from Hebrew must be considered but there is no evidence that ancient Hebrew had an [R] and Yiddish pronunciations of Hebrew sounds often differ from the ancient Middle Eastern ones.
The obvious candidate is the Germanic languages and, at first this looks, quite obvious. Modern Germanic languages have both kinds of /r/. In Standard German the [R] is currently preferred as it is in Standard Danish and Dutch. In other Germanic languages, [r] is the standard but [R] occurs as a dialect pronunciation. In fact, Robert D. King and Stephanie Beach derive the Yiddish [R] from German in their article, "On the origins of Germanic uvular [R]: the Yiddishe evidence" in the American Journal of Germanic Linguistics and Literatures September 1998 pp. 279-290.
However, if one goes back to the handbooks from the late 19th and early 20th century one gets a different impression. The original Proto-Germanic /r/ is reconstructed as [r]. The original Old High German and Middle High German quality is also given as [r]. An apical [r] is also reconstructed for Gothic.
The evidence that [r] was once univeral in Germanic is, however, not strong. A more recent discussion in the book Proto-Germanic /r/ (Alfred Kummich, Gőppingen, 1974) by Richard M. Runge cites a considerable body of evidence and later studies that indicate that Proto-Germanic /r/ was [R] and that this value was found in at least some Old High German as well as in Gothic.
If Yiddish [R] is Germanic, than this feature takes its place along with the Yiddish tendency to stress initial syllables and to put the verb in second position as a testimony to the deep rootedness of Yiddish within the Germanic family of languages.