| Abstract [eng] |
Dialect atlases (e.g., The Atlas of Polish Dialects, The Lithuanian Language Atlas) do not always cover syntax, and when they do (as in the dialect atlases of Latvian, Belarusian, Russian, and Ukrainian), the number of syntactic maps is significantly smaller than the number of maps on lexicon, phonology, or morphology. Their contents also show interesting deviations from what one finds on syntactic maps of typological atlases like WALS or GB. Even though syntax is supposed to consist of very general rules that work independently of lexical items they are applied to, the above mentioned dialect atlases (as well as The Linguistic Atlas of England, taken for comparison) actually reflect the geographic distribution of syntactic properties associated with individual lexemes (e.g., the verb ‘lack’), groups of lexemes (such as motion verbs) and, sometimes, certain grammatical morphemes (the Latvian debitive jā-, or the Slavic participle -vši). That is, instead of informing the reader on whether the object follows or precedes the verb, whether the object is marked with its own affix or shares its affix with the intransitive subject, whether general questions are introduced by means of a particle or by using alternative strategies, the syntax part of dialect atlases, more often than not, educates us about the possibilities of using diverse cases and prepositional phrases in particular functions, e.g., that of goal with motion verbs, as in Latvian ‘go in order to bring water’, where the relationship between the verb ‘go’ and the noun ‘water’ can be marked by several different combinations: iet pēc ūdens / iet pēc ūdeni / iet ūdenī / iet ūdenim pakaļ(ā) / iet pakaļ ūdenim / iet ūdens pakaļ / iet dēļ ūdens. Some of the syntactic features in the atlases under analysis accidentally happen to be areal, so that the other atlases supply structurally similar (and etymologically related) examples involving the same meaning, see Belarusian isci pa vadu / isci za vadoj and Ukrainian pišov po vodu / pišov na vodu / pišov za vodoiu / pišov vody. While it is beyond doubt that those are lexically restricted structures, the atlases do not shed much light on what particular verbs and nouns, except ‘go’ and ‘water’, can be also found with them. Sometimes similar expressions are given on other maps involving different types of nouns as objects, e.g., Belarusian ‘go in order to fetch father’ [isci] pa baćku / [isci] za baćkam or Ukrainian ‘go in order to fetch a doctor’ pišov po likarja / pišov za likarem, where the goal is animate, or Russian ‘go in order to collect nuts’ pojti po orexi / pojti v orexi, but their exact relationship remains unclear. The quantitative and qualitative peculiarities of syntactic maps in dialect atlases are easily explained by the fact that, although published towards the end of the 20th century, and some even at the beginning of the 21th century, they were mostly designed in the middle of the 20th century, when the data were also collected. At the time, modern theories of syntax were either not yet developed or had not diffused into dialect studies. In addition, dialectology’s focus is on those features that are different from the standard language and other dialects, which means that it inevitably ignores many general rules that tend to be shared by closely related idioms. Syntactic variation is often, although not necessarily always, found in cases of mismatch between the semantic role of an argument and its syntactic position, e.g., the obligatory expression of goal in the above-mentioned examples. Also, the natural interest of traditional dialectology, not unlike that of historical linguistics, lies with individual words and morphemes that can serve to establish genealogical relationship between dialects. In doing this, it discovered (already at the time of the very first dialect atlases) that every word might have its own history. Although the saying initially described phonological development, it is also true of syntactic properties, which can vary from one word to another. Syntactic maps of dialect atlases thus represent collections of lexically restricted structures bearing fascinating affinity to how the relationship between grammar and lexicon is interpreted in Construction Grammar. |
