La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

Challenging discrete approaches to secondary-predicate constructions [Elektronische Ressource] : towards a non-discrete account of resultative, depictive and qualifying constructions in English / vorgelegt von Holger Saurenbach

De
368 pages
Challenging Discrete Approaches to Secondary-Predicate Constructions Towards a Non-discrete Account of Resultative, Depictive and Qualifying Constructions in English Inaugural-Dissertation zur Erlangung der Doktorwürde der Philosophischen Fakultät IV (Sprach- und Literaturwissenschaften) der Universität Regensburg vorgelegt von Holger Saurenbach aus Amberg Regensburg, Februar 2004 Erstgutachter: Prof. Dr. Edgar W. Schneider Zweitgutachterin: Prof. Dr. Roswitha Fischer ACKNOWLEDGEMENTS iiiAcknowledgements *This project was inspired by Edgar W. Schneider's 1997 article "As as 'is': is as 'is'", which points out many of the questions raised by secondary-predicate constructions in English and has paved the way for my research into the subtleties and complexities of this intriguing aspect of English grammar. Prof. Schneider's motivating remarks on my Zulassungsarbeit and on two talks given in his research collo-quium have encouraged me to broaden the theoretical and empirical basis of my work, which has fi-nally resulted in the present dissertation. As is frequently the case with such large-scale projects, work on this dissertation has quickly taken on a life of its own.
Voir plus Voir moins




Challenging Discrete Approaches to
Secondary-Predicate Constructions

Towards a Non-discrete Account of Resultative, Depictive and
Qualifying Constructions in English




Inaugural-Dissertation
zur Erlangung der Doktorwürde
der Philosophischen Fakultät IV
(Sprach- und Literaturwissenschaften)
der
Universität Regensburg



vorgelegt von
Holger Saurenbach
aus Amberg





Regensburg, Februar 2004



Erstgutachter: Prof. Dr. Edgar W. Schneider
Zweitgutachterin: Prof. Dr. Roswitha Fischer ACKNOWLEDGEMENTS iii
Acknowledgements
*This project was inspired by Edgar W. Schneider's 1997 article "As as 'is': is as 'is'", which points out
many of the questions raised by secondary-predicate constructions in English and has paved the way
for my research into the subtleties and complexities of this intriguing aspect of English grammar. Prof.
Schneider's motivating remarks on my Zulassungsarbeit and on two talks given in his research collo-
quium have encouraged me to broaden the theoretical and empirical basis of my work, which has fi-
nally resulted in the present dissertation.
As is frequently the case with such large-scale projects, work on this dissertation has quickly
taken on a life of its own. While my Zulassungsarbeit constituted a critique of small-clause analysis
within a descriptive grammatical paradigm, the present book has become a much more comprehensive
critique of discrete grammatical approaches in general and an attempt to tackle secondary-predicate
constructions from the new angle of non-discrete syntax.
A project of this size is not possible without the help of other people. Julia Hofmann (University
of Oxford, England), Marguerita O'Neill (Institute of Technology Tralee, Ireland), Voicu Popescu
(Purdue University, USA) and Christoph Saurenbach (London School of Economics, England) have
my gratitude for helping me conduct evaluation tests with native speakers of English. I would also like
to thank Adrienne Bambach, Scott Brunstetter, Roger Miller and Ben Mui for giving their views on
the relative acceptability or unacceptability of quite a large number of English sentences. I am particu-
larly grateful to Augustus Cavanna, Jamie Kohen and Alison Thielecke, Lektoren at the University of
Regensburg, who volunteered for extended interviews on the meaning of functionally related sen-
tences.
My thanks also go to the generative grammarian Kleanthes K. Grohmann for answering some of
my questions on SC-theory and modern generative theory, as well as to the syntacticians Ray
Jackendoff and Adele Goldberg, who sent me the final draft of their article on the Resultative Con-
struction (due to appear in Language) and discussed their respective views on the role of constructions
with me.
I am also grateful to Thomas Hoffmann (University of Regensburg), who has read earlier drafts
of this dissertation and whose insightful comments have greatly improved the final version. Although
we do not always share the same opinions on syntactic theory, the lively debates we have had on vari-
ous syntactic problems have provided the present work with a lot of stimulating input. I also wish to
thank my wife Ursula for meticulously proofreading the manuscript and suggesting numerous stylistic
improvements. Needless to say, none of those mentioned here are responsible for any remaining mis-
takes, for which my computer is entirely to blame.
Thanks also to those people who were continually suprised that anyone could have such a big in-
terest in 'small' clauses. I must admit that I was surprised myself that my interest in secondary-
predicate constructions in particular and non-discrete syntax in general has not waned in the nearly
two years I have devoted to the Zulassungsarbeit and this dissertation, but has actually grown into
something like fascination. Even though it is still early days, I am sure that the non-discrete perspec-
tive will have a lot to contribute to enhancing our understanding of the complexities of syntax and
semantics.

* Work on this dissertation has been made possible by a Bavarian state scholarship (Gesetz zur Förderung des
wissenschaftlichen und künstlerischen Nachwuchses) between September 2002 and February 2004.
TABLE OF CONTENTS iv
Table of Contents
1. Introduction........................................................................................................................................ 1
2. Data and methods 6
2.1 Data based on linguistic intuition.................................................................................................. 6
2.2 Data based on electronic corpora .................................................................................................. 7
2.3 A short outline of the statistical methods used.............................................................................. 9
2.4 Limits of corpus studies ................................................................................................................ 11
PART I: CHALLENGING SYNTACTICALLY DISCRETE APPROACHES TO ENGLISH
SECONDARY-PREDICATE CONSTRUCTIONS............................................................................ 13
3. A critique of syntactic discreteness................................................................................................... 14
3.1 Distributional mismatches as the stumbling block of discrete syntax........................................... 14
3.2 Constraints and representational levels — a way out for discrete syntax? ................................... 19
3.3 The non-discreteness paradigm of Construction grammar ........................................................... 23
3.4 The non-discrete perspective of the neurosciences....................................................................... 26
4. Challenging complex-transitive analyses of secondary-predicate constructions.......................... 30
4.1 Treatment of the [NP V NP XP]-pattern in descriptive grammars ................................................ 301 2
4.1.1 The models: transitive and copula clauses..... 30
4.1.2 The derivative structure: complex-transitive clauses............................................................ 35
4.1.3 Analytical problems of the complex-transitive complementation analysis .......................... 38
4.1.4 Problems of classifying complex-transitive complementation patterns................................ 45
4.1.5 Discarding the distinction between objects and predicative complements — a way out?.... 48
4.2 Treatment of the [NP V NP to be XP]-pattern in descriptive grammars ....................................... 501 2
4.3 Treatment of the [NP V NP as XP]-............................................ 581 2
5. Challenging small-clause analyses of secondary-predicate constructions .................................... 67
5.1 From Subject-to-Object Raising to Small-Clause Theory ............................................................ 67
5.2 Attempts to prove the existence of a black hole: are small clauses syntactic entities?................. 75
5.2.1 A critical look at constituency tests ...................................................................................... 75
5.2.2 A critical look at subject tests ............................................................................................... 86
5.3 Attempts to identify a black hole: what is the categorial status of small clauses?........................ 100
5.3.1 Some preliminaries: the categorial component of generative grammar................................ 100
5.3.2 Small clauses as quasi-clauses: is an SC a pure lexical projection? ..................................... 105
5.3.3 Small clauses as reduced clauses: do SCs contain functional projections? .......................... 111
5.3.4 Small clauses as IPs: do SCs contain an empty copula?........................................................ 115
5.3.5 Small cls CPs: are full and small clauses identical?................................................... 118
5.3.6 Small clauses as complex clauses: are some SCs larger than full clauses? .......................... 124
5.4 Problems of classifying small-clause patterns .............................................................................. 127
6. Challenging complex-predicate analyses of secondary-predicate constructions 133
6.1 Complex-predicate analyses in early descriptive and generative studies...................................... 133
6.2 Compyses in modern generative grammar ....................................................... 134
6.3 Compyses in Categorial Grammar.................................................................... 138
6.4 Evidence for complex predicates and attempts to cope with distributional mismatches .............. 140
7. Challenging Predication-theory analyses of secondary-predicate constructions ......................... 146
7.1 Semantic definition of predicative relations in Predication theory ............................................... 146
7.2 Treatment of depictive patterns in Predication theory .................................................................. 150
7.3 Treatment of resultative and qualifying patterns in Predication theory ........................................ 155
PART II: TOWARDS A NON-DISCRETE ACCOUNT OF THE RESULTATIVE,
DEPICTIVE AND QUALIFYING CONSTRUCTIONS IN ENGLISH 159
8. A critique of intersystemic and epistemological discreteness......................................................... 160
8.1 A critique of intersystemic linguistic discreteness........................................................................ 160
8.1.1 The functionalist criticism of the modular architecture of grammar .................................... 160
8.1.2 Construction-independent linking rules — a way out for intersystemic discreteness?......... 164
8.1.3 Symbolic relations as construction-specific linking rules..................................................... 167
TABLE OF CONTENTS v
8.2 A critique of epistemological discreteness.................................................................................... 172
8.2.1 The cognitive criticism of objectivist semantics................................................................... 172
8.2.2 Comparing constructions on functional maps....................................................................... 174
9. A force-dynamic account of the Resultative Construction............................................................. 177
9.1 Introducing the model of force dynamics ..................................................................................... 177
9.2 Constraints on the constructional slots of the Resultative Construction....................................... 181
9.2.1 The VERB-slot....................................................................................................................... 181
9.2.2 The SUBJECT and OBJECT-slots ............................................................................................ 190
9.2.3 The RESULTATIVE-slot.......................................................................................................... 196
9.3 Mapping out the functional space of resultative and related constructions .................................. 206
9.3.1 Functional map I: relative prominence of initiator and endpoint.......................................... 206
9.3.1.1 The Passive Resultative Construction ........................................................................ 206
9.3.1.2 The Autonomous Resultative Construction ............................................................... 208
9.3.1.3 The Excessive Resultative Construction .................................................................... 213
9.3.1.4 Overview of functional map I .................................................................................... 221
9.3.2 Functional map II: Resultative and Motion Constructions ................................................... 222
9.3.3 Functional map III: Constructions focusing on the manner or result of an activity.............. 234
10. A figure/ground account of the Depictive Construction 241
10.1 Zeroing in on the semantics of the Depictive Construction ........................................................ 241
10.2 Conditions on the slots of the Depictive Construction 247
10.3 Mapping out the functional space of depictive and related constructions .................................. 252
10.3.1 Attributive and predicative constructions ........................................................................... 252
10.3.2 Functional map of depictive figure/ground constructions .................................................. 254
11. A mental-space account of the Qualifying Construction .............................................................. 259
11.1 The Qualifying Construction and functionally related patterns 259
11.2 Semantic differences between groups of mental verbs ............................................................... 266
11.2.1 Group I: prototypical association with that-clauses and NP inf XP ..................................... 266
11.2.2 Group II: prototypical association with NP inf XP and NP XP .............................................. 271
11.2.3 Groups III and IV: prototypical association with NP XP and NP as XP ................................ 274
11.2.4 Extensions from the core: less prototypical qualifying verbs ............................................. 276
11.3 Formalising the semantic differences: a mental-space account .................................................. 279
11.4 A semantic explanation of syntactic and stylistic differences between qualifying patterns ....... 290
11.4.1 Stylistic differences............................................................................................................. 290
11.4.1.1 The data.................................................................................................................... 290
11.4.1.2 A mental-space explanation ..................................................................................... 295
11.4.2 Syntactic complexity........................................................................................................... 296
11.4.2.1 The data........................ 296
11.4.2.2 A mental-space explanation........... 299
11.4.3 Tense and aspect ................................................................................................................. 301
11.4.3.1 The data........................ 301
11.4.3.2 A mental-space explanation........... 306
310 11.4.4 Extraction of NP ........................................................................................................2
11.4.4.1 The data........................ 310
11.4.4.2 A mental-space explanation ..................................................................................... 313
11.4.5 Voice................................................................................................................................... 314
11.4.5.1 The data.................................................................................................................... 314
11.4.5.2 A mental-space explanation........... 316
322 11.4.6 Categorial realisation of XP .........................................................................................
11.4.6.1 The data........................ 322
11.4.6.2 A mental-space explanation........... 324
11.4.7 Overview of syntactic and stylistic differences .................................................................. 340
12. Conclusions ....................................................................................................................................... 342
References................ 348

INTRODUCTION 1
1. Introduction
In a series of articles in the early 1970s, the syntactician Georgia M. Green pointed to an in-
triguing verbal complementation pattern that had been largely ignored by the linguistic com-
1munity up to that point (1970, 1972, 1973). The pattern follows the formula given in (1) —
where the X ranges over the variables noun (N), adjective (A), and preposition (P) — and can
be illustrated by the sentences in (2). One interesting property of this construction is the fact
2that the second postverbal phrase XP is predicated of the first postverbal phrase NP . Without 2
intending to subscribe to any particular theory at the moment, I will apply the widely-used
term 'secondary predicate' to the predicative phrase XP and will call syntactic patterns includ-
ing such a phrase 'secondary-predicate constructions'.
(1) NP V NP XP 1 2
(2) a. Mary considers John a fool.
b. Mary considers him silly.
c. rs him beneath contempt.
What Green found even more noteworthy about (1) is the ability of this structure to code at
least three different semantic relationships (1970: 275-7, 1973: 262-7). In (3a), John drank his
coffee and the coffee was hot at the same time, while in (3b) the table became clean only as a
result of the waitress's wiping it. It is more difficult to give a paraphrase for (3c) — in fact,
Green suggested that a "Linguistic Hero Medal" should be awarded to anyone who could
come up with a good solution (1970: 270). What is intuitively clear at least is that John is a
fool only in Mary's subjective view of reality.
3(3) a. John drank his coffee hot.
b. The waitress wiped the table clean.
c. Mary considers John a fool.
Green hoped that future progress in syntactic theory would find a possibility to capture the
predicative relation between the two postverbal phrases and to explain how the same syntactic
formula can convey such diverse meaning relationships (1970: 279).
The research done on secondary-predicate constructions in the wake of Green's articles
has not borne out this optimism, though. The common denominator syntacticians have been
able to reach on these constructions in the last three decades is excruciatingly small and does

1 The subscripts on the NPs do not have any theoretical significance but merely serve as notational conveniences
to tell them apart.
2 Whenever I want to highlight the predicative relationship between NP and XP, I adopt the convention of itali-2
cising the secondary predicate and underlining its predication subject.
3 Since one tends to read the present-tense version of this sentence (John drinks his coffee hot) in its habitual
sense, illustrations of dynamic sentences will, as a rule, be given in the past tense, in which the habitual/iterative
reading is less dominant.
INTRODUCTION 2
not go much beyond the few tentative suggestions already advanced by Green. It is not a mat-
ter of much dispute, for example, that the secondary predicate cannot be sensibly analysed as
a postnominal modifier in most cases. In contrast to the AP in The waitress scrubbed some-
thing very dirty, the secondary predicate does not seem to be a constituent of the postverbal
NP in (3b), as a few standard syntactic tests quickly reveal. It is, for instance, not possible to
transpose the NP XP-string within the sentence (4a) (Green 1970: 273, 1973: 260), nor to 2
paraphrase the secondary predicate with a restrictive relative clause (4b) (Green 1970: 273,
1973: 259). Neither does the pronominalisation of the postverbal NP affect the secondary
predicate (4c) (Rothstein 2000: 244-5). This behaviour contrasts sharply with that of post-
modified NPs, as the primed sentences illustrate.
4(4) a. *The table clean was wiped.
a'. Something very dirty was scrubbed.
b. !The waitress wiped the table that was clean.
b'. The waitress scrubbed something that was very dirty.
c. The waitress wiped it clean.
c'. crubbed it (*very dirty).
Another potential analysis that can be safely dismissed from the start is the treatment of NP 2
XP-strings as some sort of idiomatic expressions. As Green already noted, none of the words
in sentences such as those presented in (3) has any unusual meaning that it does not have in
other, non-idiomatic expressions (1970: 217, 1973: 258).
Beyond this meagre and unspectacular consensus, almost every other opinion ventured
on secondary-predicate constructions has occasioned — to say the least — lively debate.
There are basically four competing lines of analysis that have gained some currency in the
syntactic literature; these analyses differ both in their hypotheses as to how many syntactic
constituents must be assumed for the construction and in their views on which constituents
belong closer together syntactically.
The traditional analysis, which is provided in most descriptive grammars, suggests that
verbs such as consider select three syntactic complements — a subject, a direct object, and
what is here called a secondary predicate (5) (e.g. Biber et al. 1999: 130; Huddleston and Pul-
lum 2002: 53; Quirk et al. 1985: 53):
(5)
[Mary] considers [John] [a fool].


4 An asterisk characterises a sentence as ungrammatical; various degrees of acceptability are indicated by one or
two question marks. An exclamation mark implies that the sentence is grammatical but does not communicate
the intended meaning.
INTRODUCTION 3
While a ternary analysis does not cause major difficulties with verbs such as give, which are
standardly assumed to require a subject, an indirect object and a direct object (cf. [John] gave
[Mary] [a cake]), the situation is less straightforward with consider because the proposal
sketched above sidesteps the issue why a fool is predicated of John and thus fails to account
for one of the most distinctive characteristics of secondary-predicate constructions.
Due to this serious analytical deficiency of the traditional account, a radically different
analysis has quickly gained ground in the syntactic literature from the early 1980s onward.
Since the relationship between the two postverbal elements is one of predication, and predica-
tion is first of all a property of clauses, John and a fool could be assumed to constitute the
subject and predicate of a verbless (or 'small') clause (e.g. Haegeman and Guéron 1999: 108-
9; Stowell 1983: 297-9). In this view, consider is a binary verb, taking Mary and the proposi-
tion John a fool as its complements:
(6)
[Mary] considers [John a fool].

While the analysis given in (6) solves the problem of the additional subject/predicate relation-
ship quite neatly, it works on the assumption that the postverbal NP XP-string forms a syntac-2
tic unit — an assumption that seems to be counter-intuitive because this string does not con-
tain any of the formal characteristics standardly associated with a subordinate clause, such as
a complementiser before or a verb between the two postverbal phrases. Proponents of a
clausal analysis must therefore employ considerable syntactic ingenuity to prove that the NP 2
XP-string does in fact act as a clausal constituent.
In view of these difficulties, another binary analysis has been put forward in the literature
devoted to secondary-predicate constructions. According to this proposal, the proper parsing
of the sentence is that given in (7), where there is a discontinuous predicate considers...a fool,
of which John is the direct object (e.g. Hoeksema 1991: 666; Larson 1988: 349):
(7)
[Mary] considers-a fool [John].

Since a fool is specified as part of a complex predicate selecting NP as its direct object, the 2
predicative relation between the two postverbal phrases does not pose the same problems here
as in other analyses. What must be entered on the debit side of this hypothesis, however, is the
fact that it cannot easily account for the surface position of the direct object between the two
parts of the complex predicate.
INTRODUCTION 4
The various shortcomings of the clausal and complex-predicate analyses have led some
syntacticians to argue for a more sophisticated ternary analysis instead. These proposals form
a rather heterogeneous set and can be subsumed under the label of 'Predication theory', a term
used by Williams for his pioneering work in this direction (1980, 1983). Some predication
theorists maintain that the secondary predicate is not a complement of the main verb, but is
licensed by the syntactic complex consisting of the verb and its two arguments through a spe-
cial grammatical mechanism.
(8)
[[Mary] considers [John]] [a fool].


All of the approaches outlined above are primarily concerned with the correct parsing of the
syntactic formula given in (1). No matter which analysis one prefers, therefore, Green's cen-
tral question as to how the same syntactic pattern can code three distinct semantic relation-
ships has not even been touched on. While some syntacticians have remained remarkably si-
lent about this issue, others have made various kinds of suggestions. The most popular proce-
dure is to assign different syntactic structures to the sentences given in (3) along the lines of
some complement/adjunct distinction. Such an approach is, however, subject to as wildly di-
vergent opinions as is the proper analysis of the syntactic blueprint in (1) itself. The research
done on the relationship between the syntax and the semantics of secondary-predicate con-
structions has thus not been able to substantially flesh out Green's early speculative sugges-
tions:
[S]ome interplay between target-structure conspiracies, syntactic properties of lexical items,
real-world possibilities and nonobvious aspects of the meaning of lexical items serves to adjust
the number of readings possible for any given sentence of this form. (1973: 268)
The present dissertation makes the rather strong claim that none of the analyses sketched in
(5-8) can be sustained, and that the different semantics of secondary-predicate constructions
cannot be explained by a complement/adjunct analysis or, for that matter, any other purely
syntactic analysis. Following the spirit of some promising recent publications on syntactic
theory (particularly Croft 2001 and Goldberg 1995), I aim to show that the quandary syntacti-
cians find themselves in when tackling NP XP-constructions results from a number of errone-2
ous assumptions that lie at the core of contemporary syntactic theories working within what I
call the 'discreteness' paradigm. Chapter 3 intends to reveal the largely implicit presupposi-
tions discrete syntactic theories rely on, and to pit them against the fresh perspective offered
by 'non-discrete' grammar, which suggests itself as a viable alternative.
INTRODUCTION 5
The rest of part I is given over to a detailed criticism of the four discrete analyses of sec-
ondary-predicate constructions outlined above. Chapters 4 and 5 contain a thorough critique
of the traditional ternary analysis and the small-clause account, respectively, while chapters 6
and 7 are devoted to relatively brief discussions of the more marginal complex-predicate and
predication theories. After the 'challenge'-part of this book has illustrated that each of these
analyses runs into major conceptual difficulties that are largely unresolvable with the machin-
ery provided by mainstream syntactic theories, part II approaches the syntactic and semantic
problems posed by secondary-predicate constructions from the theoretical angle of non-
discrete grammar and shows that this framework opens promising new avenues of inquiry.
Before we delve into the details of secondary-predicate constructions, a few remarks on
the empirical basis of my data are in order.