Mendele: Yiddish literature and language ______________________________________________________ Contents of Vol. 3.165 December 14, 1993 1) Yiddish grammar (Ellen Prince) 1)---------------------------------------------------- Date: Mon Dec 13 15:52:32 1993 From: "Ellen F. Prince" Subject: yiddish grammar To: marnen laibow-koser i appreciate your very rational peace-making post, but would like to make a few comments in the hope of clarifying some points. >The recent debate about verb-first construction is, in fact, a >perfect example. Those who claimed that it was always incorrect were >flying in the face of established evidence, trying to remold the >data to fit the hypothesis instead of the other way around. ales heyst a 'hypothesis'. hypothesis: a tentative investigation that accounts for a set of facts and can be TESTED BY FURTHER INVESTIGATION; a theory. (american heritage dictionary, 3rd ed. emphasis mine.) as you no doubt noticed, there was no possibility of testing the claim made here by the bermans; rather, we were supposed to take a line from a 1st year language textbook as irrefutable fact. just wanted to point that out... > Perhaps because Yiddish is primarily a >spoken language, the rules seem to be looser in Yiddish than in, >say, Latin, but that doesn't mean they don't exist. However, >precisely because the rules *are* loose, we must be very careful >about saying "always" and "never". first, EVERY natural living language is 'primarily a spoken language'. a very tiny percentage of the world's natural languages also have a literary tradition, and yiddish is in fact one of these. however, having a literary tradition is really irrelevant to whether a language has a grammar. EVERY NATURAL LANGUAGE HAS A GRAMMAR. or as many grammars as it has dialects. the grammar of a language resides in the head of the person who speaks it. how else could we explain one's ability to produce and understand sentences never before uttered? a parrot memorizes a bunch of 'sentences'; people don't. rather, we memorize a bunch of words and then we somehow infer a bunch of rules/strategies/patterns/whatever for how we may combine those words to form the possible utterances of the language. we call those rules/strategies/patterns a 'grammar'. the task of the contemporary yiddishist, then, would be to figure out what those rules/strategies/patterns in a yiddish speaker's head are. several issues arise here. first, in whose head should we look? i.e. different yiddish speakers have different grammars; which one(s) should we try to investigate? in one sense, the answer is simple and non-controversial: only the yiddish of NATIVE speakers can give us insights into what the grammar in a REAL yiddish speaker's head looks like. you and i and all the others whose competence in yiddish is as semi-speakers, adult language learners, passive bilinguals, etc. cannot possibly be trusted to produce the necessary sort of data for such study, since the one thing we know about the yiddish grammars in our heads is that it is significantly different from those we would consider native speakers of yiddish. ok, so we take as data the output of native yiddish speakers, those who acquired the language at the normal time (early childhood) and in the normal way (in a yiddish speech community). (what constitutes a yiddish speech community is itself an interesting sociolinguistic question that i won't pursue.) so now for the next issue: WHICH native yiddish speaker's yiddish output should be taken as data? here the issue is socio-political, not linguistic. from a linguistic point of view, any and all native dialects are well-formed, rule-governed, equal in all (linguistic) respects. from a socio-political view, some are better than others. the point i wish to make here is that we must not confuse what a linguist would call 'bad yiddish', i.e. the yiddish of someone who's not a native speaker, with what a layperson might call 'bad yiddish', i.e. the yiddish of a native speaker of a socially stigmatized dialect. now to yiddish grammar per se. it's simply not true that it's 'looser' than, say, latin grammar. or than any other grammar, for that matter. of course, we, as modern english speakers, have very little acquaintance with latin and what we may know of it typically involves the dialect of a tiny handful of speakers (caeser, cicero, livy, virgil...), so we may think that it's very monolithic, but of course, when it was a living language, it had the same range of social and geographic dialects as any other living language, more than most no doubt given the huge region and society in which it was spoken. consider that each of the modern romance languages (french, spanish, italian, portuguese, catalan, rumanian, provencal, moldavian...) is simply the modern version of a different dialect of latin. as to the substance of yiddish grammar, a good deal is known, thanks to fairly intensive research in the past 25 years, most of it by some of the syntacticians on mendele. one very salient feature that has been at issue here is that yiddish is a so-called 'verb-second' language, meaning it is a language (like ALL germanic languages except modern english, i.e. like german, dutch, swedish, danish, norwegian, icelandic, frisian...) in which the tensed verb stands in second position in CANONICAL (i.e. typical) declarative sentences. put differently, there is exactly one 'slot' before the tensed verb and, in canonical declarative sentences, it is always filled. what is also known is that yiddish, like icelandic, has NONCANONICAL declarative sentences in which the preverbal slot is NOT filled. as is the case with ALL noncanonical sentence-types in any language, such sentences have rather special discourse functions and therefore have a restricted distribution. that is, they don't sound normal unless they're in the 'right' discourse context, since they add a certain 'meaning', and, if that meaning is not appropriate in the discourse context, the occurrence of such a sentence would sound bizarre. let me illustrate with a different language (english) and a different noncanonical sentence-type ('topicalization'): 1. q: have you seen sam lately? a1: i saw sam yesterday. a2: sam i saw yesterday. in answer to 1q, a speaker could say either 1a1, the canonical form, or 1a2, a noncanonical form called 'topicalization'. 2. q. what's new? a1.: i saw sam yesterday. a2.: #sam i saw yesterday. in answer to 2q, the canonical form in 2a1 is still fine, but the topicalization in 2a2 would be weird indeed. this is what i mean about noncanonical forms having a 'restricted distribution'. so we see that the availability of different sentence-types doesn't mean the grammar is 'looser'; on the contrary, we see that it's more complex than it would otherwise be and that native speakers have a richer set of 'rules' in their head than they would otherwise have. sorry for the long rambling; i hope this has been of some help. ellen prince ______________________________________________________ End of Mendele Vol. 3.165 Mendele has 2 rules: 1. Provide a Subject: line. 2. Sign your article. Send submissions/responses to: mendele@yalevm.ycc.yale.edu Other business: nmiller@starbase.trincoll.edu Anonymous ftp archives available on: ftp.mendele.trincoll.edu in the directory pub/mendele/files