5
Introduction
Resultative constructions have been matter of many discussions in the last decades, which moved
through manifold and differentiated theorizations involving syntax, lexicon and cognitive
linguistics.
Despite being a complex field of analysis, we feel like giving a general definition of what we are
going to investigate in the following pages: when we speak of resultative constructions we allude to
secondary predications which “describe the state of an argument resulting from the action denoted
by the verb” (Boas 2003: 1). We give some preliminary examples.
(1) I dyed my grey school skirt dark red. (BNC)
(2) Guatemala mob stones driver to death. (SF Examiner, 2.5.2000)
(3) He had run himself out of breath… (BNC)
The italicized phrases are resultative phrases describing the final state/change of state of the
posteverbal Noun Phrase (NP), caused by the action denoted by the verb. A general scheme would
be:
[NP V NP XP]
Where X = A, N, or P
1
(Boas 2003: 2)
We said “secondary predication” because we have “two basic dependency relations between the
postverbal NP and the two constituents surrounding it” (Boas 2003: 2) The former (primary
predication, as in I dyed mu grey school skirt) holds between the verb and the NP, the latter holds
between the NP and the following XP (secondary predication, as between skirt and red). This
structure is illustrated by the following diagram (in Boas 2003: 2)
V NP XP
primary secondary
1
NP = Noun Phrase, V = Verb, A = Adjective, N = Noun, P = Preposition.
6
Yet, if the relationship among V, NP and XP is quite transparent in example (1) and (2), it is not so
in example (3), in which the verb run, typically intransitive, acquires a reflexive object, traditionally
not associated with it.
We may quote many other examples of resultative structures in which the relationship occurring
not only between V and NP (as in “Jim ran his feet sore”) but also between V and XP (as in “He
wiped the table dirty”) could sound quite odd. This secondary predication is therefore to be seen in
wider terms, involving not only syntax but also other linguistic issues like semantics, pragmatics,
contextual information and cognitive linguistics.
This is the reason why we have chosen to deal with resultatives in the light of constructional
theories, which are supposed to be the most suitable for giving an exhaustive description of such
structure: the claim for the existence of constructions, that is, form-meaning pairs able to provide a
new perspective of the event described by the verb, reveals to be determining for a wide-ranged
understanding of this field.
Starting in the first chapter with Adele Goldberg’s constructional approach, which considers
constructions as a range of abstract form-meaning correspondences to be applied to verbs in order
to provide new arguments, we show that the recourse to similar structures is superfluous and
imprecise when it comes to capture the full semantic and syntactic distribution of resultatives,
Once discussed these problematic aspects, we move towards a similar account, brought about by
Hans Christian Boas, but differentiating from the previous one in that it considers each verbal entry
as a sort of conventionalized mini-construction, which together with its interrelated contextual
information, is able to create what Boas calls a single “event frame”, constituting a point of
departure for the licensing of resultatives.
Despite holding Boas’ approach in greater consideration than Goldberg’s (we will explain the
underlying reasons in detail), we feel like observing that the matter of language conventionalization,
that is, the consolidation process of linguistic structures established arbitrarily inside a speakers’
community, is here too much stressed, so as to avoid the possibility of finding out, as far as it is
possible, some principles of language motivation governing the choice of one or another structure.
This is exactly what we try to do in the second chapter, dealing with a closer analysis of some
verbal items belonging to the poison verb-class in combination with the resultative phrase to death.
By moving from the list of V-PP combinations offered by Boas in his work (Boas 2003), we draw
some verbs pertaining to Levin’s class of poison verbs (Levin 1993), ulteriorly expanded by Faber
and Mairal (1999) under the name of verbs of existence class, and try to understand the nature of
their collocation with to death by means of cognitive factors: conceptual metonymy ( Peña 2010),
iconicity principle (Dirven & Verspoor 2004) and temporal extension (my theorization). As for this
7
last factor, we rely our analysis upon the splitting of our verbs into two general linguistic frames
drawn from the Internet project FrameNet: the “cause harm-frame” and the “killing-frame”
In conclusion we offer a brief (and quite superficial) contrastive outlook upon resultatives in
English and German. Through some examples taken from both language (mostly consisting of a
sentence and its translation into the other language), we state that both are able to show an equal
degree of morphosemantic compactness in different situations, and the solutions adopted from time
to time by each language to reproduce a resultative outcome can vary according to the linguistic
material a language relies upon, be it a resultative or non-resultative structure.
As for the numerous examples quoted to support our statements, most of them are drawn directly
from the essays/articles examined for our research. Only in the second chapter we provide some
sentences ad hoc taken from the web at large.
8
Chapter 1
Theoretical hints upon resultatives
1.1 Goldberg’s perspective
1.1.1 The notion of “construction” in Goldberg
We shall move our discourse from the key notion of “construction”, reworked by the American
linguist A. Goldberg, in order to get a closer outlook both on the important relationship occurring
between lexical items and semantic structures, especially as far as verb semantics is concerned, and
on the interconnection constructions show between lexicon and syntax.
In her main inquiry into grammar (Goldberg 1995) she argues that an entirely lexically-based
approach to the analysis of sentences, where the global meaning of a sentence is chiefly instantiated
by the idiosyncratic
2
verbal semantics together with its relevant argument structure, is unable to
account for the full range of English data. Taking into consideration the following sentence:
(1) He sneezed the napkin off the table.
(Goldberg 1995: 224)
Goldberg points out that there’s no way of capturing its overall meaning only by focusing on the
verb to sneeze, whose argument frame can be described uniquely as follows:
Sneezer (agent) > sneeze (action)
She notices that neither force dynamics nor movement are inferable from the basic verbal event
structure (no direct object would be expected from the verb sneeze), while it is clear that in our
sentence a caused motion process is enacted, in which an unanimated object (the napkin) is thrown
out of a table by a “sneezing agent” so as to change its first location as a result.
It is therefore necessary to make use of other structural principles which go beyond the verb
semantics and succeed in accounting for the entire process taking place, that is, “semantic structures
2
With “idiosyncratic semantics” we refer to every specific verb sense carried out by the same form, each of them
differing from the others as for its proper meaning and argument structure.
9
together with their associated formal expression (…) independent of the lexical items which
instantiate them” (Goldberg 1995: 1): this peculiar correspondences between form and meaning are
what Goldberg calls constructions. In our study we are going to deal mainly with argument
structure constructions, a subclass which is able to “provide the basic means of clausal expression
in a language” (Goldberg 1995: 3). The example considered before is an instantiation of a “Caused
Motion Construction” (X (agent) CAUSES Y (napkin) TO MOVE Z (off the table)). As suggested
by the definition itself, the introduction of a construction depicting such a process allows the verb
to gain a new argument pattern which is derived neither from the single component parts nor from
other already existing constructions (non compositionality), but it is provided by the construction
itself. If a single verb, then, appears with different syntactic patterns, these differences should be
attributed to the particular constructions it figures in.
Conative X DIRECTS ACTION AT Y Subj V Obl at
Sam kicked at Bill
Resultative X CAUSES Y TO BECOME Z Subj V Obj Xcompl
Sam kicked Bill to death
The first instantiation of kick results from the application of a “Conative Construction”, suggesting
that the action performed has not achieved its target (to hit Bill with a kick); therefore the use of a
prepositional oblique formally mirrors the lower degree of affection of the patient (Bill). The
second instantiation differs from the former one in that the action is not only fully accomplished but
causes the patient to undergo a change of state ending with his death. Here the use of a direct object
is indispensable since the affection of the patient reaches its highest degree.
Furthermore, although having insisted on the “non derivational” aspect of constructions, it should
also be added that these are not comparable to the so-called ”listemes” (see DiSciullo and Williams
1987), that is a simple list of loose grammatical rules, but they should be taken “to constitute a
highly structured lattice of interrelated information” (Goldberg 1995: 5). These interconnections
work not only between different construction types (we will discuss later how resultative
constructions are referred to as metaphorical caused motion constructions), but among constructions
sharing analogous syntactic patterns: it is hypothesized that “construction involving basic argument
structure are shown to be associated with scenes reflecting basic human experiences” (Goldberg
1995: 5), such as transferring something to someone else, or causing something to change its state.
It is not difficult to recognize in this statement a re-proposition of the linguistic iconicity principle,