Documentation

Linglib.Studies.Svenonius2004

Svenonius (2004): Slavic Prefixes Inside and Outside VP #

[Sve04] splits Slavic verbal prefixes into two classes: lexical prefixes are R-heads inside VP (resultative, particle-like, connecting to [dD95]'s treatment of Germanic verb particles), while superlexical prefixes are Asp-heads outside VP (aspectual operators). The same prefix realises either class — his §1 exx. (1a)/(1c) use Russian za- both ways, on the shared fragment morph Russian.Verbs.za. The classification is contested rather than consensus: [Rom04] documents diagnostic mismatches, and Tatevosov's later work posits an intermediate class.

The syntactic height cut named here is the same one Minimalist.AspFlavor (outer vs inner aspect, Travis/Cinque) carries for Sinitic split-aspect; the types are kept separate because the commitments differ (lexical = resultative R-head, not merely inner-Asp host).

Data flow: attested examples live in Data.Examples.Svenonius2004 (generated from JSON); prefix morphs and stems are fragment entries; an Analysis pairs an example with a stem and a classified prefix sequence — the classification is this paper's analytical act, applied to shared data, never a fragment field.

Main definitions #

Main results #

Aspectual subtypes of the superlexical class — the labels recurring in [Sve04] §4 (his Bulgarian ordering (57)) and in [Ist04]'s prefix-by-prefix taxonomy. A selection, not a closed set: excessive, terminative, and perdurative also occur in the literature.

Instances For
    @[instance_reducible]
    Equations

    [Sve04]'s lexical / superlexical split as a single ADT — the superlexical case carries its subtype.

    Instances For

      A classified prefix sequence (surface order, outermost first) is well-stacked when no lexical prefix appears outside a superlexical one — [Sve04] §1: "the superlexical prefix always appears outside the lexical prefix".

      Equations
      Instances For

        An analysis of an attested example: the fragment stem it is built on and its prefix sequence (outermost first), each prefix a fragment morph paired with this paper's class assignment.

        Instances For

          The word-formation tree: the prefix morphs folded, innermost-last, over the stem root.

          Equations
          • One or more equations did not get rendered due to their size.
          Instances For

            Every analysis tree is concatenative: prefixation only.

            The example's hyphen-segmented text equals the analysis' decomposition into prefix forms plus stem. Applicable to citation-form examples (not sentence examples or inflected tokens).

            Equations
            Instances For
              @[instance_reducible]
              Equations
              • One or more equations did not get rendered due to their size.

              The paper's analyses #

              Russian from §1 and §4.1; the Polish (4c) and the Bulgarian stacks (3a), (3e) as he classifies them (for (3a) [Ist04]'s own finer taxonomy diverges — her attenuative po- — see Studies/Istratkova2004.lean).

              (1a) za-brosil — lexical spatial za- on perfective brositj.

              Equations
              Instances For

                (1b) za-brosil 'gave up' — the same lexical za-, idiomatic.

                Equations
                Instances For

                  (1c) za-brosal — superlexical inceptive za- on imperfective brosatj: the minimal pair with a1a on the same fragment morph.

                  Equations
                  • One or more equations did not get rendered due to their size.
                  Instances For

                    (4a) po-vy-brasyvatj — superlexical distributive po- outside lexical vy- on the secondary-imperfective stem.

                    Equations
                    • One or more equations did not get rendered due to their size.
                    Instances For

                      (58) discussion: za-kuritj — superlexical inceptive za-.

                      Equations
                      • One or more equations did not get rendered due to their size.
                      Instances For

                        (58) discussion: po-čitatj — superlexical attenuative po-.

                        Equations
                        • One or more equations did not get rendered due to their size.
                        Instances For

                          (4c) po-w-chodzili — superlexical distributive po- outside lexical w- (Polish counterpart of a4a).

                          Equations
                          • One or more equations did not get rendered due to their size.
                          Instances For

                            (3a) po-na-razkaža — his gloss DLMT-CMLT on perfective razkaža.

                            Equations
                            • One or more equations did not get rendered due to their size.
                            Instances For

                              (3e) iz-po-na-pre-razkaža — his gloss CMPL-DSTR-CMLT-RPET, the deepest stack he cites from [Ist04].

                              Equations
                              • One or more equations did not get rendered due to their size.
                              Instances For

                                All analyses of this study.

                                Equations
                                • One or more equations did not get rendered due to their size.
                                Instances For

                                  Results #

                                  Every analysis is well-stacked.

                                  The attested ungrammatical reversal (4b) vy-po-brasyvatj (recorded as the alternatives of (4a)) violates well-stackedness: it would put the lexical prefix outside the superlexical one.

                                  Likewise the Polish reversal (4d) w-po-chodzili.

                                  Diagnostic (56c) (§4.1) over the Russian analyses: a superlexically prefixed verb is built on an imperfective stem. Russian-scoped: his Bulgarian stacks attach to the quantized perfective razkaža ([Ist04]), so (56c) is not a Bulgarian diagnostic.

                                  For the citation-form examples, the paper's hyphen segmentation equals the analysis' decomposition. (Sentence examples (1a)-(1c) and the inflected Polish token (4c) are excluded.)