The overgeneration problem and the case of semipredicatives in Russian 55
Such feature sharing can be easily implemented through multi-attachment or
multi-dominance, as in Frampton (2004) or Citko (2011), among many others.^35 This
has been applied with much useful insight to the analysis of movement. The basic idea
is that there are no autonomous copies; “movement” is instead a metaphor for multiple
occurrences, i.e. distinct nodes in the tree are linked to the same item (they “call up” or
“point” to the same address). Thus, (5) could be represented roughly as in (67):
(67) [TP1 TFIN ljubit [TP2 TINF got ovit’ s ama]]
Nadj aWhereas under the MTC account the lower copy of Nadja lacks case and the upper
one is nominative, under the multi-attachment model there is only one Nadja, which
becomes nominative by virtue of its higher occurrence, infinite [Spec,TP].
Of course, Nadja (like all syntactic items) is a collection of features. So to repre-
sent feature sharing between Nadja and sama we can show each of these items drawing
its morphological features from the same source:
(68) [TP1 TFINljubi t [TP2 TINF got ovi t’ ]]
[SAM][“NADJA”][PHI= . ] [CASE= ?]Bullets in (68) are addresses of information structures (i.e. data sets). The diagram
is meant to indicate the network of addresses pointing to the information at other
addresses, with square brackets containing feature subsets, just before Tfin values
Nadja as nominative. (The dotted arrow from Tfin indicates this.) So sam shares phi
and case features with Nadja, and whereas the former are feminine singular intrinsi-
cally, the latter can only be determined syntagmatically. So sama is co-valued with
- Feature sharing is proposed in Frampton and Gutmann (2000) and used in Pesetsky and
Torrego (2007) inter alia. An actual implementation of these two models to sharing case and
phi-features in Polish copular clauses is provided in Bondaruk (2013).