![1. Syntax − The State of a Controversial Art 3
(1965) onwards that the syntactic component of the grammar is “generative” (we will
come to this term below), while the other components are “interpretative”, i.e. work on
structures provided by syntax. This position has been challenged by a variety of frame-
works. HPSG (Pollard and Sag 1994; cf. S. Müller, this volume) assumes that syntactic
and semantic properties receive the same formal representation, and interact in agreement
and anaphoric binding (the reader is referred to Wagner’s contribution to this handbook
to discuss the relation between syntax and phonology). Pustejovsky’s generative lexicon
(Pustejovsky 1995) raises the question of how syntactic categorization is influenced by
lexical semantics to account for coercion processes.
These are just two proposals that challenge a model of syntax, in which syntax is
taken to differ fundamentally even from the linguistic components with which it shares
an interface (cf. also Mycock, this volume).
3. Sentence grammar differs from extra-sentential
(or sub-sentential) modes of combination
Riemsdijk (1984: 2) states that “[the structure of the sentence] is determined by rules
and principles sui generis”. As often when contemplating questions about syntax, the
statement seems to be clear and intuitive at the outset, but invites further considerations
after second thoughts. What we can assume is that there are sentence internal operations
(rules and principles), whose application to units that consist of sentences would be
nonsensical, i.e. would not lead to plausible results. Case marking might be a case in
point, as well as reference to syntactic arguments or grammatical relations as discourse
relations. But the composition of sentences may also rely on rules, which do not exclu-
sively apply within a sentence. Pronominal agreement can serve as an illustration. Con-
sider the examples in (1).
(1) a. John told Paulai that shei would have to apply for the dean position on Mon-
day the latest.
b. John spoke with Paulai about the dean position. Shei understood that she
would have to apply on Monday the latest.
To begin with, we are talking about one sentence in (1a), but two sentences in (1b). And
yet, the relationship between the object of tell/speak to the subject of the embedded
sentence and the subject of the second sentence, respectively, seems to be similar. In
particular, we see agreement in gender, person, and number. Going further, we may
assume that the agreement is a consequence of Principle B of Binding Theory (Chomsky
1981; Pollard and Sag 1994) in (1a). Since identification of the indices of antecedent and
pronoun implies agreement of antecedent and pronoun in number, person, and gender,
the observed agreement can be deduced from a condition on syntactic elements within
a clause. In (1b), however, the agreement may come about because the discourse referent
of the antecedent is identified with the discourse referent of the pronoun, as this is
usually done in Discourse Representation Theory (Kamp and Reyle 1993). Since they
are identical, and since discourse referents either cannot be formally distinguished from](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-21-320.jpg)
![I. Introduction: Syntax in Linguistics
4
indices (as in the analysis of Pollard and Sag 1994) or bear the very same features, once
again agreement emerges because two indices are identified, and hence the features that
make up the indices are identified as well. At this point, we may ask whether index
identification is a syntactic rule sui generis. It would be possible to argue that index
identification started as rule of sentence grammar and spread into the rule system for
discourses. If this interpretation is assumed, we may consider van Riemsdijk’s view as
uncontroversial. The statement would have to be translated to the effect that rules of
sentence grammar might form the basis for rules outside of sentence grammar, but that
the reverse is not likely to happen. The reader may decide whether this characterization
still allows the maintenance of the position that the structure of the sentence is formed
by rules sui generis.
We would like to note that there is a second interpretation of Riemsdijk’s statement.
Even if we take it that the rules within a sentence are different from the rules outside of
it, we may still ask whether the rules are restricted in their application to atomic units
or whether the sentence is the smallest unit to which specific rules apply. This question
translates into the relationship between syntax and morphology, or more generally into
the relationship between syntax and the lexicon (for the latter, cf. Alexiadou, this vol-
ume). One position has been initiated by lexical integrity (Bresnan and Mchombo 1995).
According to the lexical integrity hypothesis, morphological processes and syntactic
rules form two separate domains of linguistic rule systems. Interestingly, the lexical
integrity hypothesis has not only been attacked from the position of syntactic access to
sub-syntactic units, but also by lexical rules that seem to affect syntax and morphology
(for a recent vindication of lexical rules against constructional analyses, cf. Müller and
Wechsler 2014). Another position gives up the distinction between morphology and syn-
tax (cf. Harley, this volume).
Can morphology for instance be taken to be word syntax? This question requires
careful investigation as well, and sometimes, we find surprising evidence. Consider the
relationship between syntactic combinations and morphological composition. If morpho-
logical composition crucially differs from syntactic modes of combination, the former
should not be able to satisfy requirements imposed by the latter. But this does not seen
to be correct. It is well known that morphological operations − such as composition and
derivation − affect the valency of lexemes, but one sometimes even finds that modifica-
tion is mirrored in morphology. As an illustration consider mass terms in German that
may only appear together with an indefinite determiner if they are internally modified,
as illustrated with Stahl (‘steel’) in (2).
[German]
(2) a. Dies
this
ist
is
ein
a
Stahl
steel
*(von
of
besonderer
particular
Güte).
quality
‘This is a high-quality steel.’
b. Dies
this
ist
is
ein
a
Gütestahl.
quality.steel
‘This is a high-quality steel.’
The combination with an indefinite determiner is ungrammatical without internal modifi-
cation in (2a), and similarly in (2b). But in (2b), the modifier is realized as part of a
compound. If (morphological) composition and (syntactic) modification are aspects of](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-22-320.jpg)
![I. Introduction: Syntax in Linguistics
8
it seems much more plausible to assume that complementation is instantiated by a single
schema, which yields uniformly binary branching structures. Schema 1 and Schema 2
as well as the purported schema for German are instances of complementation, but
complementation is a universal mode of combination (cf. below for a qualification), and
as such requires a formal representation within the model.
It turns out that a schema more general than Schema 1 and Schema 2 in English, as
well as more general than the schema employed for German cannot be easily defined.
Still, the combination of a head with its argument(s) seems to be a plausible candidate
for a universal rule, which may even apply to free word order languages (Evans and
Levinson 2009: 441), i.e. languages that are not apparently confined by phrasal bound-
aries. Latin is a familiar instance of a free word order language, as opposed to languages
employing free order of phrases. Evans and Levinson (2009: 441) illustrate for Latin
that tree-like structures may not be appropriate to describe the relations between the verb
and its subject, or between an adnominal modifier and the head noun of the subject, as
illustrated in (3).
[Classical Latin]
(3) ultima
last.NOM
Cumaei
Cumae.GEN
venit
come.3SG.PST
iam
now
carminis
song.GEN
aetas
age.NOM
‘The last age of the Cumaen song has now arrived.’
Evans and Levinson (2009: 441)
We can observe that the adjective ultima is severed from the noun aetas, and the relation-
ship between this subject (last age) and the verb venit is interrupted by an adverbial and
the postnominal genitive carminis Cumaei. An analysis in terms of non-branching trees
is presumably not easily provided for (3). What we can say, however, is that endocentric-
ity plays a crucial role in this example, just as it does in structures that can be described
more easily in tree structures: aetas is the head of the subject, and ultima is its adjectival
modifier, and the featural make-up (nominative) must be related to the head, and not to
the postnominal complement of the head. One can say that the discontinuity of ultima
and aetas becomes a discontinuity only because they are conceived as a unit at some
level of representation, that they together fulfil the subject argument slot of the verb,
and as such that they together are subject to complementation.
While there may not be evidence for continuous phrases, there is clearly evidence
that features target the ‘right’ elements, which must be achieved somehow, and is usually
achieved by invoking endocentricity.
So while Pollard and Sag’s Immediate Dominance Schemata translate into more usual
phrase structures for German and English, this does not have to be the case, and struc-
tures akin to Dependency Grammar (cf. Osborne, this volume) may be the result. Still,
we would like to maintain for Latin that the head-complement-schema is at work, and
more generally that either the schema or the concept of endocentricity count as a univer-
sal. The universality of complementation is not entirely undisputed. Koenig and Michel-
son (2012) discuss Oneida and suggests that the head-complement-schema is not em-
ployed in this language. Even if this conclusion turns out to be correct, we would still
require a formal concept of complementation to begin with.
Van Riemsdijk (1984: 2) also takes structure dependence to be an uncontroversial
tenet of syntactic theory, but ignores the fact that structure dependency invites the ques-
tion which kind of structure is required. It might well be that not all languages of the](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-26-320.jpg)
![1. Syntax − The State of a Controversial Art 9
world can be analysed by employing tree-like structures − in fact, it strikes us that tree-
like structures are more often used as a lingua franca of dialects of syntactic theories
than as basic building blocks of the theory, consider Categorial Grammar (Baldridge and
Hoyt, this volume) and Minimalism (Richards, this volume) in this respect − but it still
seems plausible to assume that syntax adduces structure, be it for the simple reason that
semantic interpretation works more nicely, if structure is adduced. The question whether
tree-like structures or dependency structures (or functional structures, as employed in
LFG, cf. Butt and King, this volume) or something entirely different will be appropriate
can only be determined by making strong claims that stand the test against the diversity
of natural languages. Linguists criticising syntactic universals often adhere to diversity
as a means in itself, but it strikes us that only the interplay between claims embedded
in syntactic theories and the diversity of natural languages may eventually allow linguis-
tics to become a mature science. With regard to the complementation schema discussed
above, we would like to assume that language specific complementation schemata are
formally represented as instances of a more abstract, and possibly universal complemen-
tation schema, which forms an instance of endocentricity (with a modification schema
as another instance of endocentricity, and further instances as well). Couched within the
respective theory, the relation between the universal and the language-specific schemata
should be represented as simply as possible (if it turns out that the appropriate definition
may only refer to dependencies, tree-like structures employed in German or English may
be viewed as epiphenomena). Ideally, language-specific schemata should be represented
as more specific instances of universal schemata; the specific schemata are still abstract
(for instance, they do not make reference to particular categories), and yet they are much
more concrete than their universal super-types.
Imposing a concept like complementation requires its formal representation within
syntactic theory. Hence, it becomes a prerequisite for determining the adequate model-
ling in the theory, and a touchstone for the theory itself. The quest for formal universals
has thus an immediate bearing on determining the competition of syntactic theories, and
clearly cannot be established outside the realm of a theory (which may lead to the
confusion that is e.g. expressed in Evans and Levinson (2009: 436−437)).
With this background, let us enter unto a discussion of what we see as misunderstood
universal categorizations. We assume that most (if not all) linguists would agree that a
term like relative clause is much less a universal category than a pure means of descrip-
tion. It is less clear whether a relative clause is an instance of a more general operation,
where a phrasal modifier is related to a modified element through a marker (the marker
itself may be present or not, and may be more or less abstract). Gil (2001: 107−108)
tries to show that clausal association to nominals in Hokkien Chinese, as illustrated in
(4), should not be confused with the concept of a relative clause in English or German.
[Hokkien Chinese]
(4) a44(>44)
-beŋ24
Ah Beng
bue53
buy
e24>22
ASSOC
ph
eŋ24>22
-ko53
apple
‘apples that Ah Beng bought’
(Gil 2001: 107)
Because the attributive marker e24
(the superscript indicates a tone; x > y indicates tonal
sandhi) is not only used in Hokkien to establish a relationship between a clause and a
nominal, but is also required to establish relations between adjectives and nouns, PPs](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-27-320.jpg)
![I. Introduction: Syntax in Linguistics
10
and nouns, and further adnominal elements and nouns, Gil (2001: 108) concludes that
the concept relative clause should not be used in Hokkien grammar. While this may be
correct, it also seems trivial. However, it seems plausible to assume that the attributive
construction in Hokkien and relative clauses in other languages share a common core in
that a nominal element is formally linked to another phrase, and establishes a restrictive
interpretation through the combination. A relative marker (not to be confused with the
relative pronoun, and hence possibly without phonetic signature) in English or German
establishes what is established by e24
in Hokkien. If we abstract over this particular
property and indicate that we are not calling a relative clause a universal, we might be
able to identify a possibly universal mode of combination (cf. for instance, Kayne 1994
and subsequent work).
With regard to the question raised in this section, and the current state of affairs
regarding its answer, it seems to us that syntactic research should focus on identifying
or rejecting more abstract universals as part of syntactic theories (as e.g. complementa-
tion). In addition, it seems that at least some rather basic universals (such as employing
noun and verb as categories) have survived recent iconoclastic attempts to demolish
them (once again, the reader is referred to the language sketch of Northern Straits Salish
in Czaykowska-Higgins and Leonard, this volume). So, the nature of linguistic universals
is controversial, but it is a fruitful controversy.
5. What is the object of study?
After Saussure’s introduction of the distinction between langue, parole and faculté de
langage, every syntactic theory must position itself with regard to its actual object of
study. Saussure made it clear that the object of study is mesmerizingly difficult: “Quel
est l’objet à la fois intégral et concret de la linguistique? La question est particulièrement
difficile” (Engler 1968: 123). Van Riemsdijk (1984: 1) states that “we regard the object
which we study as a real object, a mental organ as Chomsky has put it. From this it also
follows that what we study is the grammar, not the language.” Van Riemsdijk concludes
that language acquisition is the central problem of syntactic theory. This position has
been attacked from various perspectives − not the least one being that there seems to a
conflation of a variety of concepts. For instance, there can be real objects without being
related to a mental organ. In their introduction to one of the strictest attempts to provide
a formal analysis of natural language, Gazdar et al. (1985: 5) state that “[o]ur linguistic
theory is not a theory of how a child abstracts from the surrounding hubbub of linguistic
and non-linguistic noises enough evidence to gain a mental grasp of a natural language.”
Other syntacticians attack the idea that grammar comes first and that language is derived
from grammar. At least some versions of Construction Grammar (cf. Fried, this volume)
assume that the grammar is induced from language data. It seems uncontroversial, how-
ever, that language in general is a bio-cultural hybrid, partly the product of biological
evolution, but also partly the product of our own making, but this characterization applies
to human beings in general, and may thus not be a specific property of language, or a
property only in the slightly circular sense that it is human language we are dealing with.
The objects of syntactic study have been delimited by means of various distinctions
that prove to be problematic in themselves. There is the celebrated distinction between](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-28-320.jpg)
![1. Syntax − The State of a Controversial Art 11
competence and performance, as introduced in Chomsky (1965), and later supplanted by
a distinction between internal and external language in Chomsky (1986). Here, van
Riemsdisk (1984: 3) remarks that “everyone lives by it”. We are not so sure that this is
actually correct. The latter distinction has been utilized to separate theories into those
that address internal and those that address external language. But even if we followed
this direction and assumed that Gazdar et al. (1985) investigated external language,
while Chomsky (1986) was concerned with internal language, we cannot be sure that
the separation actually yields different objects, as it is purely noumenal. Drawing such
an arbitrary distinction will lead to confusion at best, and will weaken the field as a
whole at worst. In any case, there are many syntactic models that are agnostic to the
distinction, as well as to the distinction between competence and performance.
Similarly, the object under syntactic study was delimited by the distinction between
core and periphery within the Theory of Principles and Parameters. If one goes through
the discussions of the past 20 years, however, one easily detects that today’s periphery
might become tomorrow’s core and vice versa. Müller (2013) discusses Exceptional
Case Marking (ECM, more familiar in traditional grammar under the rubric ‘accusative-
cum-infinitive construction’) as an illustration for this tendency: Chomsky (1981: 70)
assigned ECM to the periphery, but 15 years later, ECM became a building block of
case assignment in specifier-head-configurations.
The morale to be drawn here is threefold. Firstly, it seems to us that the distinctions
(competence/performance, core/periphery, etc.) have been useful in the past since it al-
lowed syntactic research to start somewhere without being constantly endangered by
requests to also account for other, presumably related phenomena. In this sense, early
syntactic theories had to be reductionist, and introduced distinctions to this end. Syntax
in its current guise should give up reductionism.
Secondly, it seems that these distinctions have failed to provide an instrument to
gauge whether syntactic models make headway or not. If we take a fixed concept of a
core that has to be described or explained by syntactic theory, then we can assume that
a model should be abandoned (or at least very critically scrutinized) that does not provide
analyses for phenomena that unequivocally belong to a “core”. Progress could be charac-
terized by accounting for an agreed-upon set of phenomena, and could be gauged by
consecutively extending the core that is covered in this way, until it becomes clear that
further extensions would lead to analyses of phenomena that form part of the periphery
because they emerge as cultural artefacts that have no biological basis. It would also
allow syntactic analysis to establish a canon of analysed phenomena (where a canon is
understood in the sense of engineering and natural science, i.e. as foundational part that
has already been successfully covered), and to proceed from this canon to areas of
syntactic research that still await detailed scrutiny. Yet, syntactic theories often have not
developed in this way. As a striking illustration consider the treatment of negative place-
ment in inverted English sentences.
(5) a. Can John not leave?
b. Will John not leave?
Almost 25 years after their celebrated analysis in Syntactic Structures (Chomsky 1957),
Lasnik (2000: 151) admits that “[t]here are about a dozen really fancy ways in the
literature to keep [the sentences in (5)] from violating the [Head Movement Constraint].](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-29-320.jpg)
![2. Syntactic Constructions 15
Koenig, Jean-Pierre, and Karin Michelson
2012 The (non)universality of syntactic selection and functional application. In: Christopher
Piñon (ed.), Empirical Issues in Syntax and Semantics 9: 185−205. [http://www.cssp.
cnrs.fr/eiss9/]
Müller, Stefan, and Stephen Wechsler
2014 Lexical approaches to argument structure. To appear in Theoretical Linguistics 40.
Pollard, Carl, and Ivan A. Sag
1994 Head-driven Phrase Structure Grammar. Chicago: University of Chicago Press.
Riemsdijk, Henk van
1984 Introductory remarks. In: Wim de Geest and Yvan Putseys (eds.), Sentential Complemen-
tation, 1−9. Dordrecht: Foris Publications.
Schütze, Carson
1996 The Empirical Base of Linguistics: Grammaticality Judgments and Linguistic Methodol-
ogy. University of Chicago Press.
Tibor Kiss, Bochum (Germany)
Artemis Alexiadou, Stuttgart (Germany)
2. Syntactic Constructions
1. Introduction
2. Examples of constructions
3. Constructions and syntactic theory
4. Conclusion
5. References (selected)
Abstract
The term construction is widely used descriptively in discussing grammar, and is still
used informally in most theoretical work for characteristic formal patterns of syntactic
categories or features, usually associated with a meaning and/or function. Modern lin-
guistic theories employ a range of formal devices to produce or characterize surface
constructions; they may be rules, or schemata, or constraints. It is usually assumed that
competence in a language consists largely of these formal devices together with a lexi-
con; the constructions themselves are epiphenomenal. As such, constructions are an
abstraction over the data which linguistic theory must analyze; insight in syntax is
achieved through discovering generalizations across constructions.
1. Introduction
The term construction is ubiquitous in contemporary syntactic literature, being used
informally to refer to linguistic expressions in a variety of ways. The term also has a
technical sense in the theory of Construction Grammar, as detailed in Chapter 28.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-33-320.jpg)
![I. Introduction: Syntax in Linguistics
20
ments in which arguments are, descriptively speaking, shared across the two clauses (see
Chapters 14 and 38 on Control).
(8) a. Ian wants to stay at the house. (control)
b. Ian seemed to stay at the house. (raising)
c. We believed Ian to stay at the house. (ECM)
English provides many more examples of constructions involving various configurations
of arguments, and other languages provide yet more. In some cases, the availability of
a construction may be tied to the availability of a lexical item; for example, it may be
that if a language has an ECM construction if and only if it has a verb with the selectional
properties exhibited by English believe in (8c). However, in other cases, the availability
of a construction in a given language does not seem to be connected to lexical items.
An example is the resultative construction illustrated in (7a), which many languages
lack, despite having verbs and adjectives which are otherwise like the ones used in the
English resultative construction (Snyder 2001).
2.2. Unbounded dependencies
There is a range of constructions which have been analyzed as involving displaced con-
stituents, or filler-gap dependencies. These include (in the order in which they appear in
[9]): wh-questions, embedded wh-questions, clefts, pseudoclefts, relative clauses, com-
parative constructions, and (contrastive) topicalization (for some discussion of some of
these and related constructions, see Chapter 12 on V2, Chapter 13 on Discourse Configu-
rationality, and especially Chapter 21 on Relative Clauses and Correlatives).
(9) a. Which book did you read?
b. I asked which book you read.
c. It was this book that I read.
d. What I read was this book.
e. The book that I read was long.
f. She read a longer book than I read.
g. This book, I read.
These constructions have in common that the dependency can cross finite clause bounda-
ries.
(10) a. Which book did your mother think you read?
b. I asked which book your mother thought you read.
c. It was this book that my mother thought I read.
d. What my mother thought I read was this book.
e. The book that my mother thought I read was long.
f. She read a longer book than my mother thought I read.
g. This book, my mother thought I read.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-38-320.jpg)
![2. Syntactic Constructions 21
In this respect they contrast with passive, as illustrated in (11a), and for example raising,
as illustrated in (11b) (compare [12]).
(11) a. *The house was thought (by his mother) the owner sold.
b. *Ian seemed stayed at the house.
(12) a. It was thought the owner sold the house.
b. It seemed Ian stayed at the house.
In fact, the constructions in (9)−(10) can cross indefinitely many finite clause boundaries,
and for this reason are known as unbounded dependencies.
(13) a. Which book did you think your mother thought you read?
b. I asked which book you thought your mother thought you read.
c. It was this book that I thought my mother thought I read.
d. What I thought my mother thought I read was this book.
e. The book that I thought my mother thought I read was long.
f. She read a longer book than I thought my mother thought I read.
g. This book, I thought my mother thought I read.
The availability and properties of unbounded dependency constructions can vary some-
what from one language to the next. For example, sometimes there is a resumptive
pronoun in the gap position, and in other cases there is no displacement on the surface,
with the filler element remaining in situ. Unlike the case with argument structure alterna-
tions, this kind of variation tends not to be dependent on lexical items, though it may
be connected to the properties of functional elements such as that and which.
In fact, it has been proposed that properties of constructions are largely determined
by the properties of their heads. This is explicit in the name of the theory Head-Driven
Phrase Structure Grammar (see Chapter 27 on HPSG), but is also a common assumption
in other theories (e.g. Borer 1984). For example, the properties of a relative clause could
conceivably be entirely determined by the cluster of features contained on its (possibly
abstract) head. The head would be a kind of C[omplementizer] taking a finite TP comple-
ment, attracting a suitable nominal element to its specifier, and projecting a category
which could be used as a nominal modifier.
2.3. Complex constructions
In this regard, constructions like the What’s X doing Y? construction mentioned above
are complex, as they appear to involve an interdependency of several heads. Mainstream
theory would probably treat a construction of this kind as a special kind of idiom.
Construction Grammar, in contrast, holds that there is no principled difference be-
tween a fully general construction and a highly idiomatic one, or even between a fully
general abstract construction and a lexical item; each is a pairing of form with function,
broadly construed.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-39-320.jpg)
![I. Introduction: Syntax in Linguistics
28
describes two separate events expressed by two different verbs, each in a different clause,
with the cause of the second event being given in the first clause. Each verb is a separate
lexical item that corresponds to a separate syntactic predicate.
(2) Charlie caused Lily to laugh.
Not all causative constructions are formed in this way though, and the variation that
exists brings to light issues which have a direct bearing on the status of lexical items
and their interaction with the syntactic component of the grammar. In a Romance lan-
guage such as Catalan, the causative construction consists of two verbs, but in contrast
to its English counterpart, tests show that it is not biclausal but monoclausal. For exam-
ple, while the logical subject of llegir ‘read’ is realized as a grammatical subject in (3a),
in (3b) it appears as an object expressed as a prepositional phrase which must follow
the other object noun phrase un poema ‘a poem’ (Alsina 1996: 190, 210). This is consist-
ent with the causee (the boy) being the third argument of a single syntactic predicate
(‘make-read’) in a simple clause.
[Catalan]
(3) a. El
the
nen
boy
llegeiz
read
un
a
poema.
poem
‘The boy is reading a poem.’
b. El
the
mestre
teacher
fa
makes
llegir
read
un
a
poema
poem
al
to.the
nen.
boy
‘The teacher is making the boy read a poem.’
(Alsina 1996: 190)
In the Catalan monoclausal causative construction therefore two separate verb forms
combine to give a single syntactic predicate, demonstrating that the overt form of a verb
and its function can be separated in this instance.
In Japanese, the productive causative formation process involves affixation of the
morpheme -(s)ase to the verb, as in (3b).
[Japanese]
(4) a. Akira-ga
Akira-NOM
it-ta.
go-PST
‘Akira went.’
b. Hiroshi-ga
Hiroshi-NOM
Akira-o
Akira-ACC
ik-ase-ta.
go-CAUS-PST
‘Hiroshi made Akira go.’
This process would be classed as occurring in the syntax according to Wasow (1977)
because it is productive. I refer to this construction in theory-neutral terms as the produc-
tive sase causative.
Kuroda (1965), amongst others, argues that the productive sase causative construction
exemplified in (4) consists of two clauses rather than one, even though the verb is a single
word. Evidence for this analysis includes the fact that the inclusion of a manner adjunct,
which can be interpreted as modifying either the grammatical subject or the logical subject](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-46-320.jpg)
![3. Syntax and its Interfaces: An Overview 29
in a biclausal sentence, also results in ambiguity in a productive sase causative construction,
as (5) shows. Hence, productive sase causative constructions are biclausal.
[Japanese]
(5) Jon-wa
John-TOP
muriyari
forcibly
sono
the
ko-ni
child-DAT
sono
the
kutsushita-o
socks-ACC
ooyorokobide
happily
hak-ase-ta.
put.on-CAUS-PST
‘John forcibly made the child put on his socks happily.’
[ambiguous as to who was happy]
(Matsumoto 1998: 9)
In a sense, the productive sase causative in (4b) is the mirror image of the Catalan
causative in (3b): in the Japanese construction, a single verb form corresponds to two
syntactic predicates; in Catalan, two verb forms correspond to a single syntactic predi-
cate. Thus, data show that there is not necessarily a one-to-one correspondence between
function and verb form.
In addition to the biclausal productive sase causative construction in Japanese, a
certain group of transitive verbs form a monoclausal structure when the causative -(s)ase
morpheme is added. For example, the causative form of nomu ‘to drink’ (nom-ase-ru)
in (6) means ‘to feed/give a drink to’.
[Japanese]
(6) Hahaoya-wa
mother-TOP
akachan-ni
baby-DAT
miruku-o
milk-ACC
nom-ase-ta.
drink-CAUS-PST
‘The mother fed the baby with milk (from a bottle).’
(Matsumoto 1998: 5)
This process would be classed as occurring in the lexicon according to Wasow (1977)
because it is not fully productive. I refer to this construction as the non-productive
sase causative.
In contrast to a biclausal productive sase causative construction like (5), the interpre-
tation of adverbials is unambiguous in the case of a non-productive sase causative con-
struction such as (7). This and other data indicate that non-productive sase causatives
are monoclausal.
[Japanese]
(7) Jon-wa
John-TOP
sono
the
netakirino
bedridden
roojin-ni
old.man-DAT
sono
the
kutsushita-o
socks-ACC
ooyorokobide
happily
hak-ase-ta.
put.on-CAUS-PST
‘John happily put the socks on the feet of the bedridden old man.’
[unambiguously John who was happy]
(Matsumoto 1998: 9)
However, if the productive sase causative is formed in the syntax, as per Wasow (1977),
and the lexicon provides the input for the syntax, Manning et al. (1999) note that data
such as (8) are problematic. In this sentence, a lexical process (nominalization via suffix-
ation of -kata ‘way’) has applied after the formation of a productive sase causative.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-47-320.jpg)
![I. Introduction: Syntax in Linguistics
30
[Japanese]
(8) (?kodomo-e-no)
child-DAT-GEN
hon-no
book-GEN
yom-ase-kata
read-CAUS-way
‘the way to cause (the child) to read a book’
(Manning et al. 1999: 44)
Any analysis of the productive sase causative must account for (8), as well as the differ-
ent syntactic structures (monoclausal vs. biclausal) of the two sase causative construc-
tions. Three different approaches could be adopted: (i) both causatives are formed by a
syntactic process, but differ in terms of their syntax (Section 2.3.1); (ii) both are formed
by a lexical process, but differ lexically (Section 2.3.2); or (iii) the two processes are
fundamentally different because one is lexical and the other is syntactic (Section 2.3.3).
Comparing these approaches, each with distinct theoretical consequences, provides an
insight into a range of issues concerning the relationship between lexical items and
syntax and how they have been addressed.
2.3.1. Syntactic approaches
Harley (2008) analyses causativization in purely syntactic terms within the framework
of Distributed Morphology (Halle and Marantz 1993; see also Section 3.2), a theory in
which there is no lexicon and (syntactic) principles apply to both morphology and syntax.
The main semantic feature shared by different types of causatives is defined syntacti-
cally: the head CAUSE of a verbal projection is common to all causative constructions.
The two Japanese sase causatives are analysed as differing in terms of the structural
complexity of their syntax. In the case of the non-productive sase causative, -(s)ase
appears in vo
, immediately adjacent to a verb root which cannot have a subject of its
own. By contrast, in the biclausal productive sase causative, -(s)ase is the head of a
verbal CAUSE projection which subcategorizes for a clause with its own subject.
The possibility that the same morpheme may attach to a “high” or “low” projection,
and that this may correlate in a systematic way with the differences between them, is
common to a number of analyses. For example, Travis (2000) distinguishes different
types of causatives in Malagasy and Tagalog on the basis of the syntactic position that
the causative morpheme occupies. Travis (2000) states that the Event Phrase marks the
boundary between l(exical)-syntax and s(yntactic)-syntax; causatives formed in the do-
main of l-syntax (causative morpheme occurs below E, the head of the Event Phrase)
are non-productive, while those formed in s-syntax (causative morpheme occurs above
E) are productive. This approach to the issue of causativization, as well as those outlined
by Harley (2008) and Ramchand (2008: 150−192), treats syntax as a generative compo-
nent with respect to word formation.
However, data from languages other than Japanese challenge a uniform syntactic
analysis of causativization cross-linguistically. For example, Horvath and Siloni (2010)
argue that a productive causative construction in Hungarian is not biclausal but mono-
clausal (see also Pylkkänen 2008: 107). The productive/non-productive distinction thus
does not appear to correlate with the proposed difference in syntactic structure. Another
issue for a syntactic analysis is that in order to account for the data in (8), -kata nominals
must be derived in the syntax, as Harley (2008: 48, fn. 8) notes.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-48-320.jpg)
![3. Syntax and its Interfaces: An Overview 33
a morpheme from its form. For example, an abstract L(exical)-derivation rule of plurali-
zation changes the features of an English singular count noun thus: [−Plural, +Singular]
→ [+Plural, −Singular] (Beard 1995: 160). The form of the plural is determined in a
separate morphological spelling (MS) component. MS-rules include suffixation, e.g.
dog−dogs, and umlaut, e.g. man−men, which accounts for the multiple exponence of
the morphosyntactic category ‘plural’ in English. This approach is often adopted in the
analysis of inflection, and has also been extended to derivation by some researchers (see
Beard 1998).
For those who do not subscribe to the view that syntax subsumes morphology (Sec-
tion 3.2) and instead adhere to a strong or weak version of the Lexicalist Hypothesis
(Section 3.3), the challenge is not only to identify those primitives and operations which
are uniquely morphological in their nature and define them, but also to characterize
the interface between syntax and morphology (Section 3.4). Key data are examples of
phenomena which exhibit a combination of properties that appear to be neither wholly
morphological nor wholly syntactic. These include cliticization (e.g. Anderson 2005;
Halpern 1995; Zwicky 1994), incorporation (e.g. Baker 1988; Mithun 1984) and inflec-
tion (e.g. Anderson 1992; Steele 1995; Stump 2001). Indeed, approaches to morphology
can also be classified according to the relations they posit between the morphology and
morphosyntactic properties of an inflected word (lexical vs. inferential and incremental
vs. realizational). See Stump (2001).
The proposed analyses of incorporation, undertaken in a variety of frameworks, pro-
vide a useful insight into some major issues.
3.2. Syntax subsumes morphology
Noun incorporation is a productive process found in a number of languages including
Classical Nahuatl, a North American language. The two sentences in (11) are very simi-
lar in meaning and include the same morphemes, but in (11b) the direct object naca
‘meat’ is part of the verb qua ‘eat’, the only word in the sentence.
[Classical Nahuatl]
(11) a. ni-c-qua
1SG.SBJ-3SG.OBJ-eat
in
DET
naca-tl.
meat-ABS
‘I eat (the) meat.’
b. ni-naca-qua.
1SG.SBJ-meat-eat
‘I eat meat.’
(Iturrioz Leza 2001: 715)
The process involved in forming (11b) is productive and may be the usual way to convey
a particular meaning. This is reminiscent of a syntactic process, though the result of this
noun incorporation is a single word. As Mithun (1984: 847) states, such noun incorpora-
tion is “perhaps the most nearly syntactic of all morphological processes”.
Baker’s (1988) influential approach to noun incorporation is an example of how word
formation can be analysed as a syntactic process: syntax is argued to manipulate mor-
phemes to form words in cases such as (11b). Specifically, complex words are formed](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-51-320.jpg)
![I. Introduction: Syntax in Linguistics
34
by head movement, a syntactic operation restricted by the Head Movement Constraint
(Travis 1984: 131), or possibly the Empty Category Principle (Chomsky 1981: 250) or
Relativized Minimality (Rizzi 1990), as shown in (12). This type of derivational rule is
known as Incorporation.
(12)
Noun incorporation, a way of forming a complex word, is thus analysed in terms of
established principles of derivational syntax. In this way, syntactic operations are pro-
posed to be responsible for generating structure below as well as above word level.
Other syntactic analyses of noun incorporation are presented in van Geenhoven (1998),
Koopman and Szabolcsi (2000) and Massam (2001). Baker (2009) compares his head-
movement analysis to these approaches in the context of the Minimalist Program (Chom-
sky 1995).
Baker (1988) also uses Incorporation to account for other morphosyntactic phenom-
ena, including passivization and applicativization, two processes in which affixation may
accompany a change in argument structure. Baker (1985: 375) further claims that gram-
mar is constrained by the Mirror Principle: “morphological derivations must directly
reflect syntactic derivations (and vice versa)”. The order of affixes relative to each other
and the base is thus predicted to reflect the order in which the relevant syntactic opera-
tions applied in the derivation of a complex word. If syntax subsumes morphology en-
tirely, the Mirror Principle should always constrain word-internal structure. However,
while it represents a strong tendency, the Mirror Principle does not appear to be inviola-
ble, as Hyman et al. (2009) show. For example, when an applicative is passivized, the
order of syntactic operations is applicative then passive, so the applicative morpheme
should be affixed closer to the base than the passive one according to the Mirror Princi-
ple. This is the case in the example from the Bantu language Ndebele in (13a). However,
when a passive is subject to applicativization in Ndebele, and hence the order in which
the syntactic operations apply is reversed, as in (13b), the same morphemes appear in
the same order as in (13a).
(13) a. Passivized Applicative [Ndebele]
abantwana
children
b-a-phek-el-w-a
they-PST-cook-APPL-PASS-FS
ukudla
food
‘The children were cooked food’](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-52-320.jpg)
![CHAPTER XII
LABOR
David Lamar—Labor's National Peace Council—The embargo
conference—The attempted longshoremen's strike—Dr. Dumba's
recall.
Labor produced munitions. The hands of labor could be frightened
away from work by explosions, their handiwork could be bombed on
the railways, the wharves, the lighters, and the ships, but a surer
method than either of those was the perversion of the hearts of
labor. So thought Count von Bernstorff and Dr. Albert, who dealt in
men. So thought Berlin—the General Staff sent this message to
America:
"January 26—For Military Attaché. You can obtain particulars as
to persons suitable for carrying on sabotage in the United States
and Canada from the following persons: (1) Joseph McGarrity,
Philadelphia; (2) John P. Keating, Michigan Avenue, Chicago; (3)
Jeremiah O'Leary, 16 Park Row, New York.
"One and two are absolutely reliable and discreet. Three is
reliable, but not always discreet. These persons were indicated
by Sir Roger Casement. In the United States sabotage can be
carried out on every kind of factory for supplying munitions of
war."
(Signed) "Representative of General Staff."[3]
So too thought von Rintelen, who hired men—usually the wrong
ones.](https://image.slidesharecdn.com/25170315-250522042115-4a72f131/85/Syntax-Theory-And-Analysis-Volume-1-Unknown-73-320.jpg)
Syntax Theory And Analysis Volume 1 Unknown
- 1. Syntax Theory And Analysis Volume 1 Unknown download https://ebookbell.com/product/syntax-theory-and-analysis- volume-1-unknown-50340630 Explore and download more ebooks at ebookbell.com
- 2. Here are some recommended products that we believe you will be interested in. You can click the link to download. Syntax Theory And Analysis Volume 3 Unknown https://ebookbell.com/product/syntax-theory-and-analysis- volume-3-unknown-50340470 Syntax Theory And Analysis Volume 2 Tibor Kiss Editor Artemis Alexiadou Editor https://ebookbell.com/product/syntax-theory-and-analysis- volume-2-tibor-kiss-editor-artemis-alexiadou-editor-50340518 Prosodic Syntax In Chinese Theory And Facts Feng Shengli https://ebookbell.com/product/prosodic-syntax-in-chinese-theory-and- facts-feng-shengli-10017024 Historical Syntax And Linguistic Theory Paola Crisma Giuseppe Longobardi https://ebookbell.com/product/historical-syntax-and-linguistic-theory- paola-crisma-giuseppe-longobardi-1407210
- 3. Key Terms In Syntax And Syntactic Theory Silvia Luraghi Claudia Parodi https://ebookbell.com/product/key-terms-in-syntax-and-syntactic- theory-silvia-luraghi-claudia-parodi-4415978 A Theory Of Syntax Minimal Operations And Universal Grammar Norbert Hornstein https://ebookbell.com/product/a-theory-of-syntax-minimal-operations- and-universal-grammar-norbert-hornstein-1677754 The Syntax And Semantics Of Noun Modifiers And The Theory Of Universal Grammar A Korean Perspective 1st Ed Minjoo Kim https://ebookbell.com/product/the-syntax-and-semantics-of-noun- modifiers-and-the-theory-of-universal-grammar-a-korean- perspective-1st-ed-minjoo-kim-9962614 Directival Theory Of Meaning From Syntax And Pragmatics To Narrow Linguistic Content 1st Ed Pawe Grabarczyk https://ebookbell.com/product/directival-theory-of-meaning-from- syntax-and-pragmatics-to-narrow-linguistic-content-1st-ed-pawe- grabarczyk-10487942 Essentials Of Grammatical Theory A Consensus View Of Syntax And Morphology D J Allerton https://ebookbell.com/product/essentials-of-grammatical-theory-a- consensus-view-of-syntax-and-morphology-d-j-allerton-6722192
- 5. Syntax − Theory and Analysis HSK 42.1
- 6. Handbücher zur Sprach- und Kommunikations- wissenschaft Handbooks of Linguistics and Communication Science Manuels de linguistique et des sciences de communication Mitbegründet von Gerold Ungeheuer Mitherausgegeben (1985−2001) von Hugo Steger Herausgegeben von / Edited by / Edités par Herbert Ernst Wiegand Band 42.1 De Gruyter Mouton
- 7. Syntax − Theory and Analysis An International Handbook Volume 1 Edited by Tibor Kiss Artemis Alexiadou De Gruyter Mouton
- 8. ISBN 978-3-11-020276-2 e-ISBN (PDF) 978-3-11-037740-8 e-ISBN (EPUB) 978-3-11-039423-8 ISSN 1861-5090 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. © 2015 Walter de Gruyter GmbH, Berlin/Munich/Boston Typesetting: Meta Systems Publishing & Printservices GmbH, Wustermark Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen Cover design: Martin Zech, Bremen 앪 앝 Printed on acid-free paper Printed in Germany www.degruyter.com
- 9. This handbook is dedicated to the memory of our dear friend Ursula Kleinhenz (1965−2010). The light that burns twice as bright burns half as long.
- 11. Preface When our friend Ursula Kleinhenz approached us during the annual meeting of the German Linguistic Society (DGfS) in Mainz in 2004, it was not to piece together the events of the previous evening. Being without the benefit of a screenwriter and a director, we usually had to figure out the events from the nebula of a hangover ourselves (but we never woke up next to a tiger). Ursula was just as hilarious as she was professional and could switch between these personas just as easily as she could be both simultaneously. If any consolation can be found in her premature death, it is only because her life was as intense as that of two others. Ursula’s proposal was that we should edit the syntax handbook in the HSK series. Like other handbooks in the series, the syntax handbook had a predecessor (published in 1993 and 1995), but Ursula (and de Gruyter) thought that it was about time to take a fresh look. Presumably, she also had a deeper insight in the process of conceiving, compiling, and editing such a handbook. Perhaps she could even foresee how much time would pass until the idea became a published book. With its present structure, the handbook aims to provide a valuable source not only for the professional syntactician but also for the linguist who wants to gain information about the current state of the art in syntax; in particular it should facilitate the advanced student’s way into syntax. Syntax can look back on a long tradition. The term itself is ambiguous. On the one hand, it is understood as a means of structural and descriptive analysis of individual languages using clearly defined instruments. Naturally, syntactic analyses can be compar- ative, spanning several languages. On the other hand, syntax is understood as syntactic theory, the aim of which is to decide which instruments can be sensibly applied to syntactic analysis. Syntactic theory thus defines the aims of syntactic research. Especially in the last 50 years, however, the interaction between syntactic theory and syntactic analysis has led to a rapid increase in analyses and theoretical suggestions. This increase has been accompanied with the impression that syntax is a fragmented discipline. This impression is not wrong in so far as syntactic theory cannot be traced back to one school, but rather to a great number of approaches in competition with each other. It should, however, not be forgotten that the competitive situation in most cases was triggered by empirical problems and that complex problems may be approached in more than one way. Precisely this situation has led to very decisive progress in syntactic analysis in the last 50 years. This result would probably have never been achieved if syntax had been limited to just one dispute about theory or method. This handbook – which is spread out over three volumes, containing 61 articles in nine sections – adopts these unifying perspectives and places at the fore the increase in knowledge that results from the fruitful argumentation between syntactic analysis and syntactic theory. To reflect this, the handbook contains articles on syntactic phenomena from two different angles: one perspective is mainly descriptive, allowing linguists to grasp what is at issue when a particular phenomenon is subjected to sometimes heated debate. Thus Section III (Syntactic Phenomena), which covers the bulk of Volume I of the handbook, contains descriptions of a set of phenomena that are called syntactic. The phenomena comprise the argument-adjunct distinction, negation, agreement, word order, ellipsis, and idioms, among others. We would like to note that the phenomena were not selected on the basis of the personal preference of the editors. Instead, we asked our
- 12. Preface viii colleagues around the world to take part in a survey in order to determine which phenom- ena should be dealt with in such a handbook. Four of the phenomena are taken up again in Section VI (Theoretical Approaches to Selected Syntactic Phenomena) and receive theoretical analyses. The reader of the hand- book may thus approach syntactic phenomena from the perspective of what is in need of an analysis. We assume that a description can be provided prior to (even if it is not entirely independent of) an analysis since the phenomena are natural – at least in parts. Alternatively, the reader may approach syntactic phenomena from the perspective of what has been, and also what has not been, covered by an analysis. We believe that those issues in particular that have been omitted from current analysis provide good starting points for future syntacticians to gain access to the syntactic community. Such gaps in analysis should thus not been seen as flaws but as open questions to be dealt with promptly. As for the structure of the handbook as a whole, the first volume (Sections I−III) is devoted to the position of syntax in linguistics (including its interfaces to other linguistic domains), to the syntactic tradition prior to the advent of structuralism and generativism, and − as already mentioned − to syntactic phenomena. The second volume (Sections IV−VI) begins with a survey on the dominant syntactic theories and frameworks (in the Introduction to the handbook we deal with the perennial question of why we have more than one theory (cf. 1)), and continues with a detailed description of the relationship between syntax and its major interfaces. Considering the relationship of syntax to other neighbouring disciplines, we note that, right up to the 1990s, an artificial comparison between formal models and models dealing with content was postulated, for example using the key phrase of “the autonomy of syntax”. However, it has become clear in the meantime that syntax based on formal options interacts with other components of linguistic knowledge, for example through correspondence rules or through defined interfaces. The chapters in Section V (together with the overview on Syntax and its interfaces in Section I) provide a picture of syntax as placed among the other linguistic domains. The second volume closes with theoretical analyses of several of the aforemen- tioned phenomena. The third volume (Sections VII−IX) provides syntactic sketches of various languages, language types, and language families (the Bantu language family, Bora, Creole lan- guages, Georgian, German, Hindi-Urdu, Japanese, Mandarin, Northern Straits Salish, Tagalog, and Warlpiri). It then deals with the cognitive environment of syntax, covering language acquisition, language disorders, and language processing, and finally deals with the broader role of syntax when dealing with corpora, lexicographic resources, stylistics, computational linguistics, the development of reference grammars and the documentation of (endangered) languages, and finally, what role syntax might play in the classroom. In conclusion, this handbook highlights syntax as a mature discipline, whose state of knowledge concerning the languages of the world has rapidly increased. This brief preface cannot and should not be concluded without an expression our gratitude to the various groups of people without whose participation this handbook would have been an even more impossible mission. We thank our authors, which are not named here, as their articles speak for them- selves.
- 13. Preface ix Each article has been subjected to an anonymous review process, and individual articles as well as the handbook as a whole profited enormously from the comments and the suggestions made by these reviewers. We would like to thank Peter Ackema, Susan Bejar, Eva Belke, Adriana Belletti, Miriam Butt, Katharina Colomo, Jeroen van Craenen- broeck, Berthold Crysmann, Holger Diessel, Laura Downing, Stanley Dubinsky, Susann Fischer, Bart Geurts, Kook-Hee Gill, Michael Hahn, Hubert Haider, Daniel Hole, Joachim Jacobs, Jaklin Kornflit, Beth Levin, Terje Lohndal, Anke Lüdeling, Stefan Mül- ler, Timothy Osborne, Eric Potsdam, Beatrice Primus, Frank Richter, Norvin Richards, Björn Rothstein, Vieri Samek-Lodovici, Florian Schäfer, Sabine Schulte im Walde, Frank Seifart, Wolfgang Sternefeld, Jesse Tseng, Marc de Vries, Helmut Weiß, Martina Wiltschko, and Niina Ning Zhang. We would also like to thank Barbara Karlson, Birgit Sievert, Anke Beck and Wolf- gang Konwitschny at Walter de Gruyter for their assistance and their gentle way of midwifery. We also thank Nils Hirsch and Patrick Lindert for editorial assistance. Many thanks to Miriam Butt and Gereon Müller for discussions early on in the process of structuring the content of this handbook. Finally, a very special thank you to Alicia Katharina Börner. Without her rigorous and thoroughgoing assistance in the preparation of the final manuscripts, we would per- haps still be awaiting the publication of the handbook.
- 15. Contents Volume 1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii I. Introduction: Syntax in Linguistics 1. Syntax − The State of a Controversial Art · Tibor Kiss and Artemis Alexiadou . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. Syntactic Constructions · Peter Svenonius . . . . . . . . . . . . . . . . . . . . 15 3. Syntax and its Interfaces: An Overview · Louise Mycock . . . . . . . . . . . 24 II. The Syntactic Tradition 4. The Indian Grammatical Tradition · Peter Raster . . . . . . . . . . . . . . . . 70 5. Arabic Syntactic Research · Jonathan Owens . . . . . . . . . . . . . . . . . . 99 6. Prestructuralist and Structuralist Approaches to Syntax · Pieter A. M. Seuren 134 III. Syntactic Phenomena 7. Syntactic Categories and Subcategories · Hans-Jürgen Sasse . . . . . . . . . 158 8. Grammatical Relations · Beatrice Primus . . . . . . . . . . . . . . . . . . . . . 218 9. Arguments and Adjuncts · Peter Ackema . . . . . . . . . . . . . . . . . . . . . 246 10. The Morpho-Syntactic Realisation of Negation · Hedde Zeijlstra . . . . . . . 274 11. The Syntactic Role of Agreement · Stephen Wechsler . . . . . . . . . . . . . 309 12. Verb Second · Anders Holmberg . . . . . . . . . . . . . . . . . . . . . . . . . . 342 13. Discourse Configurationality · Katalin É. Kiss . . . . . . . . . . . . . . . . . . 383 14. Control · Barbara Stiebels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 15. Pronominal Anaphora · Silke Fischer . . . . . . . . . . . . . . . . . . . . . . . 446 16. Coordination · Katharina Hartmann . . . . . . . . . . . . . . . . . . . . . . . . 478 17. Word Order · Werner Frey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514 18. Ellipsis · Lobke Aelbrecht . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562 19. Syntactic Effects of Cliticization · Anna Cardinaletti . . . . . . . . . . . . . . 595 20. Ergativity · Amy Rose Deal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 21. Relative Clauses and Correlatives · Rajesh Bhatt . . . . . . . . . . . . . . . . 708 22. Voice and Valence Change · Edit Doron . . . . . . . . . . . . . . . . . . . . . 749 23. Syntax and Grammar of Idioms and Collocations · Christiane Fellbaum . . . 777
- 16. Contents xii Volume 2 IV. Syntactic Models 24. Minimalism · Marc Richards . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803 25. Lexical-Functional Grammar · Miriam Butt and Tracy Holloway King . . . 839 26. Optimality-Theoretic Syntax · Gereon Müller . . . . . . . . . . . . . . . . . . 875 27. HPSG − A Synopsis · Stefan Müller . . . . . . . . . . . . . . . . . . . . . . . 937 28. Construction Grammar · Mirjam Fried . . . . . . . . . . . . . . . . . . . . . . 974 29. Foundations of Dependency and Valency Theory · Michael Klotz . . . . . . 1004 30. Dependency Grammar · Timothy Osborne . . . . . . . . . . . . . . . . . . . . 1027 31. Categorial Grammar · Jason Baldridge and Frederick Hoyt . . . . . . . . . . 1045 V. Interfaces 32. Syntax and the Lexicon · Artemis Alexiadou . . . . . . . . . . . . . . . . . . 1088 33. The Syntax-Morphology Interface · Heidi Harley . . . . . . . . . . . . . . . . 1128 34. Phonological Evidence in Syntax · Michael Wagner . . . . . . . . . . . . . . 1154 35. The Syntax-Semantics Interface · Winfried Lechner . . . . . . . . . . . . . . 1198 36. The Syntax-Pragmatics Interface · George Tsoulas . . . . . . . . . . . . . . . 1256 VI. Theoretical Approaches to Selected Syntactic Phenomena 37. Arguments and Adjuncts · Daniel Hole . . . . . . . . . . . . . . . . . . . . . . 1284 38. Models of Control · Tibor Kiss . . . . . . . . . . . . . . . . . . . . . . . . . . 1321 39. Theories of Binding · Silke Fischer . . . . . . . . . . . . . . . . . . . . . . . . 1357 40. Word Order · Klaus Abels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1400 Volume 3 VII. Syntactic Sketches 41. German: A Grammatical Sketch · Stefan Müller . . . . . . . . . . . . . . . . 1447 42. Hindi-Urdu: Central Issues in Syntax · Alice Davison . . . . . . . . . . . . . 1478 43. Mandarin · Lisa L.-S. Cheng and Rint Sybesma . . . . . . . . . . . . . . . . . 1518 44. Japanese · Takao Gunji . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1559 45. Georgian · Alice C. Harris and Nino Amiridze . . . . . . . . . . . . . . . . . 1588 46. The Bantu Languages · Leston Chandler Buell . . . . . . . . . . . . . . . . . 1622 47. Tagalog · Paul Schachter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1658 48. Warlpiri · Kenneth L. Hale, Mary Laughren and Jane Simpson . . . . . . . . 1677 49. Creole Languages · Derek Bickerton . . . . . . . . . . . . . . . . . . . . . . . 1710 50. Northern Straits Salish · Ewa Czaykowska-Higgins and Janet Leonard . . . 1726 51. Syntactic Sketch: Bora · Frank Seifart . . . . . . . . . . . . . . . . . . . . . . 1764
- 17. Contents xiii VIII. The Cognitive Perspective 52. Syntax and Language Acquisition · Sonja Eisenbeiss . . . . . . . . . . . . . . 1792 53. Syntax and Language Disorders · Martina Penke . . . . . . . . . . . . . . . . 1833 54. Syntax and Language Processing · Claudia Felser . . . . . . . . . . . . . . . . 1875 IX. Beyond Syntax 55. Syntax and Corpora · Heike Zinsmeister . . . . . . . . . . . . . . . . . . . . . 1912 56. Syntax and Stylistics · Monika Doherty . . . . . . . . . . . . . . . . . . . . . 1942 57. Syntax and Lexicography · Rainer Osswald . . . . . . . . . . . . . . . . . . . 1963 58. Computational Syntax · Emily M. Bender, Stephen Clark and Tracy Holloway King . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2001 59. Reference Grammars · Irina Nikolaeva . . . . . . . . . . . . . . . . . . . . . . 2036 60. Language Documentation · Eva Schultze-Berndt . . . . . . . . . . . . . . . . 2063 61. Grammar in the Classroom · Anke Holler and Markus Steinbach . . . . . . . 2095 Indexes Language index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2127 Subject index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2133
- 19. I. Introduction: Syntax in Linguistics 1. Syntax − The State of a Controversial Art 1. Introduction 2. Syntax relates sound and meaning 3. Sentence grammar differs from extra-sentential (or sub-sentential) modes of combination 4. There are (no) universals 5. What is the object of study? 6. Generative grammar 7. There is more than one way to do it (presumably) 8. References (selected) Abstract This article provides a broad introduction to Syntax, as it is conceived in this handbook. We will present a number of assumptions that have been considered to constitute common ground for all syntacticians. But they are as fundamental as they are controversial for syntactic theory and analysis. In fact, approaches to syntactic analysis as well as syntac- tic theories quite often take controversial positions, leading to the apparent conclusion that syntactic research is fractured. It is not our goal to offer a band-aid to patch up what better should be kept separate. Our conclusion is that many issues might indeed be controversial within syntactic research but that such controversy should be seen as the driving force behind progress in syntax. 1. Introduction Perhaps the simplest way to approach syntax is to define it as the linguistic component that relates sound and meaning, and thus fulfils this crucial task in mapping the two intelligible sides of the Saussurean sign. Syntax must thus be able to break up the con- tinuous flow of sound signals into elementary tokens, must be able to provide a structure for the tokens, and enable the semantic component to access the tokens − and the larger parts made of them in order to account for the interpretation of complex units. It goes without saying that the same story could be told if we started with the interpretational component, letting syntax map meaning to sound. A grammar consists of a complex formal unit that requires analysis, a rule system relating atomic and complex units to the complex formal unit under analysis, and a characterization of the atomic or elementary elements, and how they relate to the rule system. Typically, the sentence (S) is the formal unit that requires analysis, there are all kinds of rule systems resulting in a sentence, and the lexical units are taken to be the atomic elements, which are assigned appropriate categories so that the rule system can make use of them. This characterization seems to be uncontroversial, and it owns its
- 20. I. Introduction: Syntax in Linguistics 2 uncontroversial status to its relative abstractness, e.g. the lack of detail concerning the rule system, the use of categories, and more or less everything else required to actually carry out a syntactic analysis. When it comes to detail, however, syntax is mostly conceived as fractured. It is often forgotten that, within the broader realm of Saussurean linguistics, proper syntax started as a straggler, coming to maturity only approximately 60 to 70 years ago, and that it has achieved quite a lot in this short lifetime. And yet, a necessary precondition of syntax seems to be that the linguist becomes a member of one of the many different schools of thought. Schools that sometimes are engaged in fierce fights (giving rise to terms like “Linguistics Wars”, cf. Harris 1993), and sometimes are content with peaceful coexis- tence, or mutual neglect. In fact, one realizes soon that “schools” are partitioned in various dimensions, as for instance formal in opposition to functional or cognitive, for- mal in opposition to descriptivist, and so forth. Deep within syntax, different views pertain to the structures that are actually used to carve up the continuous flow of sound signals (be they tree structures, dependency arcs, grammatical relations, derivations etc.). This situation may leave those who want to study syntax in a state of confusion, about which school to choose, or whether to choose one at all, or more than one. (In addition to confusion, students may also feel pressured since syntax is interfacial by definition, and hence it will become necessary for the apprentice to be familiar with one of the major components with which syntax interfaces.) It should be noted, however, that disa- greement is not pertinent to syntax alone, but can be observed in other areas of linguistics as well (or linguistics as a whole), cf. the discussion following the influential article of Evans and Levinson (2009). The goal of the present article is twofold: On the one hand, we would like to address several issues that seem to be controversial within syntax as a research area, and shed some light on real or apparent controversies. On the other hand, we would like to make clear that controversy in itself does not mean that a field is unripe or uninteresting, but to the contrary that disagreement is an expression about the relative maturity of syntax. In doing so, we will take recourse to Henk van Riemsdijk’s presentation of a set of background assumptions “which (…) virtually every linguist (…) shares” (van Riems- dijk 1984: 1). Writing in 1984, van Riemsdijk believed that the assumptions presented were widely shared. We have chosen a selection from van Riemsdijk’s assumptions, the ones that we believe to have the strongest bearing on current research. Relating current positions to the positions held back then when van Riemsdijk presented them also allows us to evaluate the progress made in syntax and syntactic theory. In fact, we take controversial issues in syntax to be an indicator of scientific progress. So, it is good that there is contro- versy. 2. Syntax relates sound and meaning While the very idea of syntax relating sound and meaning has to be considered uncontro- versial, with syntax breaking up the stream of sounds into units which accordingly re- ceive an interpretation, it is at least unclear whether sound and meaning are relevant to syntactic operations or not. Mainstream generative grammar has assumed from Chomsky
- 21. 1. Syntax − The State of a Controversial Art 3 (1965) onwards that the syntactic component of the grammar is “generative” (we will come to this term below), while the other components are “interpretative”, i.e. work on structures provided by syntax. This position has been challenged by a variety of frame- works. HPSG (Pollard and Sag 1994; cf. S. Müller, this volume) assumes that syntactic and semantic properties receive the same formal representation, and interact in agreement and anaphoric binding (the reader is referred to Wagner’s contribution to this handbook to discuss the relation between syntax and phonology). Pustejovsky’s generative lexicon (Pustejovsky 1995) raises the question of how syntactic categorization is influenced by lexical semantics to account for coercion processes. These are just two proposals that challenge a model of syntax, in which syntax is taken to differ fundamentally even from the linguistic components with which it shares an interface (cf. also Mycock, this volume). 3. Sentence grammar differs from extra-sentential (or sub-sentential) modes of combination Riemsdijk (1984: 2) states that “[the structure of the sentence] is determined by rules and principles sui generis”. As often when contemplating questions about syntax, the statement seems to be clear and intuitive at the outset, but invites further considerations after second thoughts. What we can assume is that there are sentence internal operations (rules and principles), whose application to units that consist of sentences would be nonsensical, i.e. would not lead to plausible results. Case marking might be a case in point, as well as reference to syntactic arguments or grammatical relations as discourse relations. But the composition of sentences may also rely on rules, which do not exclu- sively apply within a sentence. Pronominal agreement can serve as an illustration. Con- sider the examples in (1). (1) a. John told Paulai that shei would have to apply for the dean position on Mon- day the latest. b. John spoke with Paulai about the dean position. Shei understood that she would have to apply on Monday the latest. To begin with, we are talking about one sentence in (1a), but two sentences in (1b). And yet, the relationship between the object of tell/speak to the subject of the embedded sentence and the subject of the second sentence, respectively, seems to be similar. In particular, we see agreement in gender, person, and number. Going further, we may assume that the agreement is a consequence of Principle B of Binding Theory (Chomsky 1981; Pollard and Sag 1994) in (1a). Since identification of the indices of antecedent and pronoun implies agreement of antecedent and pronoun in number, person, and gender, the observed agreement can be deduced from a condition on syntactic elements within a clause. In (1b), however, the agreement may come about because the discourse referent of the antecedent is identified with the discourse referent of the pronoun, as this is usually done in Discourse Representation Theory (Kamp and Reyle 1993). Since they are identical, and since discourse referents either cannot be formally distinguished from
- 22. I. Introduction: Syntax in Linguistics 4 indices (as in the analysis of Pollard and Sag 1994) or bear the very same features, once again agreement emerges because two indices are identified, and hence the features that make up the indices are identified as well. At this point, we may ask whether index identification is a syntactic rule sui generis. It would be possible to argue that index identification started as rule of sentence grammar and spread into the rule system for discourses. If this interpretation is assumed, we may consider van Riemsdijk’s view as uncontroversial. The statement would have to be translated to the effect that rules of sentence grammar might form the basis for rules outside of sentence grammar, but that the reverse is not likely to happen. The reader may decide whether this characterization still allows the maintenance of the position that the structure of the sentence is formed by rules sui generis. We would like to note that there is a second interpretation of Riemsdijk’s statement. Even if we take it that the rules within a sentence are different from the rules outside of it, we may still ask whether the rules are restricted in their application to atomic units or whether the sentence is the smallest unit to which specific rules apply. This question translates into the relationship between syntax and morphology, or more generally into the relationship between syntax and the lexicon (for the latter, cf. Alexiadou, this vol- ume). One position has been initiated by lexical integrity (Bresnan and Mchombo 1995). According to the lexical integrity hypothesis, morphological processes and syntactic rules form two separate domains of linguistic rule systems. Interestingly, the lexical integrity hypothesis has not only been attacked from the position of syntactic access to sub-syntactic units, but also by lexical rules that seem to affect syntax and morphology (for a recent vindication of lexical rules against constructional analyses, cf. Müller and Wechsler 2014). Another position gives up the distinction between morphology and syn- tax (cf. Harley, this volume). Can morphology for instance be taken to be word syntax? This question requires careful investigation as well, and sometimes, we find surprising evidence. Consider the relationship between syntactic combinations and morphological composition. If morpho- logical composition crucially differs from syntactic modes of combination, the former should not be able to satisfy requirements imposed by the latter. But this does not seen to be correct. It is well known that morphological operations − such as composition and derivation − affect the valency of lexemes, but one sometimes even finds that modifica- tion is mirrored in morphology. As an illustration consider mass terms in German that may only appear together with an indefinite determiner if they are internally modified, as illustrated with Stahl (‘steel’) in (2). [German] (2) a. Dies this ist is ein a Stahl steel *(von of besonderer particular Güte). quality ‘This is a high-quality steel.’ b. Dies this ist is ein a Gütestahl. quality.steel ‘This is a high-quality steel.’ The combination with an indefinite determiner is ungrammatical without internal modifi- cation in (2a), and similarly in (2b). But in (2b), the modifier is realized as part of a compound. If (morphological) composition and (syntactic) modification are aspects of
- 23. 1. Syntax − The State of a Controversial Art 5 modification within sentence grammar, the observation in (2) would not be surprising. If morphological operations are separated from syntactic operations, the observation is in need of an explanation. The analysis of polysynthetic languages forms another challenge for the distinction of syntactic and sub-syntactic rule systems. With regard to polysynthetic languages, some researchers argue to give up a distinction between morphology and syntax (cf. Evans and Levinson 2009: 432). Eventually, the analysis of polysynthetic languages will form a touchstone for the relation between syntax and morphology, and will thus allow to answer the question whether syntactic rules are rules sui generis when compared to super-syntactic operations, or perhaps also when compared to sub-syntactic structure building. With these provisos given, one could still say that the vast majority of syntacticians agree that the sentence is the unit within which syntactic relations must be established. By the same line of reasoning, syntacticians assume that syntactic relations are not em- ployed to describe relations between sentences. The clear distinction between rule sys- tems affecting the sentence and rule systems affecting super-sentential units gives the sentence a psychological reality, as well as a formal definition. 4. There are (no) universals The concept of universal grammar, and with it, the concept of a syntactic universal has always been controversial. Van Riemsdijk was well aware of this. He thus simply states that “universals are there to be found”. He goes on and says: “we all think we must find them”. Nowadays, this view is contested from various points of view. Evans and Levin- son (2009) argue against it from the perspective of linguistic typology. Construction Grammar (cf. Fried, this volume) rejects the idea of a universal from the perspective of syntax being derived in a bottom-up fashion from individual instances. Let us restrict ourselves to the term universal (we will come back to the term univer- sal grammar below), and let us further restrict ourselves by explicitly excluding the Greenbergian concept of a universal from further discussion. One reason for excluding this concept is that it is not the entity that most syntacticians have in mind when using the term universal. Another reason is that the Greenbergian concept is controversial in itself. As an illustration, consider Dryer’s (1988) refutation of the conditional universal relating verb object and adjective noun orders. Even under this restriction, it is amazing to observe the sheer magnitude of misconceptions that the term universal invites. Chom- sky (1965: 27−30) initiated the debate. It is thus instructive to look at Chomsky’s intro- duction of the terms of substantive and formal universals. The former are characterized as follows: “A theory of substantive universals claims that items of a particular kind in any language must be drawn from a fixed class of items.” (Chomsky 1965: 28) Formal universals receive the following introduction: “The property of having a grammar meet- ing a certain abstract condition might be called a formal linguistic universal, if shown to be a general property of natural languages. Recent attempts to specify the abstract conditions that a generative grammar must meet have produced a variety of proposals concerning formal universals (…) For example, consider the proposal that the syntactic component of a grammar must contain transformational rules (…) mapping semantically
- 24. I. Introduction: Syntax in Linguistics 6 interpreted deep structures into phonetically interpreted surface structures, or the pro- posal that the phonological component of a grammar consists of a sequence of rules, a subset of which may apply cyclically to successively more dominant constituents of the surface structure (a transformational cycle …)” (Chomsky 1965: 29). Criticism has focussed much more on substantive than on formal universals. Within the generative syntactic tradition, criticism has also been encouraged by undertaking research into the identification of universals only sparsely. The most advanced attempt to provide such universals, within the Theory of Principles and Parameters (Chomsky and Lasnik 1993), has failed by general consensus within and outside of generative linguistics. With regard to formal universals, there has been an extended debate between 1979 and 1994 (approximately) on whether transformational rules are formal universals or not for the reason that their very existence has been called into question (cf. Gazdar 1981; Gazdar et al. 1985). But this debate has not broadened our understanding towards the concept of formal universals. Evans and Levinson (2009: 430−431) have a clear idea what substantive universals should be, but are much less clear in their characterization of formal universals. We would doubt that subjacency would actually be considered a formal universal (cf. Evans and Levinson 2009: 436−437). But discussions of the term substantive universal suffer from problems as well. Evans and Levinson (2009: 432) make the following claim with regard to the existence of apparently substantive universals despite their superficial absence in individual lan- guages: “The differences (…) are somehow superficial (…) This (…) is wrong, in the straightforward sense that the experts (…) do not agree that it is true.” This conclusion seems to suggest that an assumption is proven to be wrong because not every expert agrees on its truth. There might be something that escapes us here, but major scientific breakthroughs have quite often been accompanied by denials of experts. So it seems that the existence of universal categories that are not attested in every language is controver- sial, and not that their existence has been falsified. But controversy can lead to research results, even if these results falsify the initial assumptions of the researcher. Consider the discussion around the claim that well-established syntactic categories (such as noun, verb, adjective, or preposition) form part of the inventory of every language. Pinker and Bloom (1990) claim − perhaps somewhat prematurely − that all languages employ the aforementioned categories. Evans and Levinson (2009: 434) adduce arguments to show that adjectives and adverbs may not exist in all languages and that certain languages may even relinquish the distinction between noun and verb. (Northern) Straits Salish is usually cited as a language that does not make use of a noun/verb distinction. The current discussion, however, as also presented in Czaykowska-Higgins and Leonard (this volume), suggests that the distinction between verbs and nouns is present in Northern Straits Salish. Here, Evans and Levinson (2009) apparently underestimate the insights that syntacticians can gain from pursuing such assumptions. We argue that it might be premature to assume the universality of lexical categories that we have taken over from Latin and Greek grammarians. But the true nature of lexical categories can only be revealed once we have at least shown negatively that we cannot work with these pur- ported categories. In this respect, a claim about the universality of a lexical category, while proven wrong in further research, opens a venue to refine our understanding of the concept of a lexical category in itself. But such claims are always couched in linguis- tic (syntactic) theories. A modest syntactic theory may at least offer a means to adduce
- 25. 1. Syntax − The State of a Controversial Art 7 evidence that an entity considered to be a prerequisite does not exist, with the usual repercussions for the theory. More concrete entities usually abound in the controversies about syntactic universals. More abstract categories or rules are not typically the subject of such debates, and with it the existence of more abstract (possibly formal) universals is not often discussed. We would like to illustrate this point with three examples: the exhaustive constant partial order (ECPO) property of ID/LP grammars in GPSG as an example of a universal that is directly derived from the formal properties of the grammar, the presumed univer- sality of endocentricity (a.k.a. Merge in Minimalism), and the terminological non-univer- sality of relative clauses. The ECPO property of ID/LP grammars states that constraints on ordering must be context-insensitive insofar as an ordering must neither depend on the mother under which the ordering takes place, nor on the head of a phrase. This property has been introduced as a general property of grammars that separate dominance and linear precedence in GPSG (Gazdar et al. 1985). It should be noted here that other grammar models that assume this distinction do not take over the ECPO, as e.g. HPSG (Pollard and Sag 1994). The ECPO predicts that languages that require contextual ordering as mentioned above do not exist. The ECPO is thus a strong candidate for a formal universal. Gazdar et al. (1985: 49) carefully note that imposing the ECPO restricts the possible languages ana- lysed by a GPSG severely, and hence that they would even be surprised “if ECPO turned out to be a linguistic universal”. Nevertheless, Gazdar et al. (1985: 49) are “committed to the rather strong claim that it turns out to be universal”. In discussing the ECPO from hindsight, it is not really important that it became clear rather soon that it could not be a universal. With regard to this point (but not actually addressing the ECPO), Haider (1993) noted that unmarked orders between German com- plements depend on the verbal head of which they are complements. Hence, it had to be admitted that German violates the ECPO (with regard to the ordering constraints in Ger- man, (cf. Frey, this volume)). What is important is that the ECPO is even more abstract than a formal universal in that it provides a property that is directly derived from the formal apparatus of GPSG. ID/LP grammars of the form presented in GPSG must obey the ECPO. It seems to us that uncovering potential formal universals like the ECPO is more rewarding than collecting indicators that a particular syntactic category does not occur in a language − even if the potential formal universals are refuted eventually. Let us now turn to the treatment of complementation, which perhaps is sometimes considered an operation so ubiquitous that its character is not even mentioned. We will illustrate the problem with the treatment of complementation in HPSG, but similar con- siderations apply to the operation Merge in syntactic Minimalism. HPSG (Pollard and Sag 1994; cf. the introduction by S. Müller, this volume) assumes that Immediate Domi- nance Schemata form part of Universal Grammar. One such schema is the head-comple- ment-schema, which combines a head with one or more complements, and which is a subtype of endocentric phrases. They illustrate their analysis mostly with English (and thus invite criticism by linguistic typologists), but the very point in question can already be made clear by adding another related language (German) to the picture. For English, Pollard and Sag (1994) introduce Schemata 1 and 2 as instances of the head-comple- ment-schema. They are instances of the head-complement-schema for English (the first combining sole arguments with phrasal projections, the second combining all arguments save one with lexical entities), and as such could not be used to analyse German, where
- 26. I. Introduction: Syntax in Linguistics 8 it seems much more plausible to assume that complementation is instantiated by a single schema, which yields uniformly binary branching structures. Schema 1 and Schema 2 as well as the purported schema for German are instances of complementation, but complementation is a universal mode of combination (cf. below for a qualification), and as such requires a formal representation within the model. It turns out that a schema more general than Schema 1 and Schema 2 in English, as well as more general than the schema employed for German cannot be easily defined. Still, the combination of a head with its argument(s) seems to be a plausible candidate for a universal rule, which may even apply to free word order languages (Evans and Levinson 2009: 441), i.e. languages that are not apparently confined by phrasal bound- aries. Latin is a familiar instance of a free word order language, as opposed to languages employing free order of phrases. Evans and Levinson (2009: 441) illustrate for Latin that tree-like structures may not be appropriate to describe the relations between the verb and its subject, or between an adnominal modifier and the head noun of the subject, as illustrated in (3). [Classical Latin] (3) ultima last.NOM Cumaei Cumae.GEN venit come.3SG.PST iam now carminis song.GEN aetas age.NOM ‘The last age of the Cumaen song has now arrived.’ Evans and Levinson (2009: 441) We can observe that the adjective ultima is severed from the noun aetas, and the relation- ship between this subject (last age) and the verb venit is interrupted by an adverbial and the postnominal genitive carminis Cumaei. An analysis in terms of non-branching trees is presumably not easily provided for (3). What we can say, however, is that endocentric- ity plays a crucial role in this example, just as it does in structures that can be described more easily in tree structures: aetas is the head of the subject, and ultima is its adjectival modifier, and the featural make-up (nominative) must be related to the head, and not to the postnominal complement of the head. One can say that the discontinuity of ultima and aetas becomes a discontinuity only because they are conceived as a unit at some level of representation, that they together fulfil the subject argument slot of the verb, and as such that they together are subject to complementation. While there may not be evidence for continuous phrases, there is clearly evidence that features target the ‘right’ elements, which must be achieved somehow, and is usually achieved by invoking endocentricity. So while Pollard and Sag’s Immediate Dominance Schemata translate into more usual phrase structures for German and English, this does not have to be the case, and struc- tures akin to Dependency Grammar (cf. Osborne, this volume) may be the result. Still, we would like to maintain for Latin that the head-complement-schema is at work, and more generally that either the schema or the concept of endocentricity count as a univer- sal. The universality of complementation is not entirely undisputed. Koenig and Michel- son (2012) discuss Oneida and suggests that the head-complement-schema is not em- ployed in this language. Even if this conclusion turns out to be correct, we would still require a formal concept of complementation to begin with. Van Riemsdijk (1984: 2) also takes structure dependence to be an uncontroversial tenet of syntactic theory, but ignores the fact that structure dependency invites the ques- tion which kind of structure is required. It might well be that not all languages of the
- 27. 1. Syntax − The State of a Controversial Art 9 world can be analysed by employing tree-like structures − in fact, it strikes us that tree- like structures are more often used as a lingua franca of dialects of syntactic theories than as basic building blocks of the theory, consider Categorial Grammar (Baldridge and Hoyt, this volume) and Minimalism (Richards, this volume) in this respect − but it still seems plausible to assume that syntax adduces structure, be it for the simple reason that semantic interpretation works more nicely, if structure is adduced. The question whether tree-like structures or dependency structures (or functional structures, as employed in LFG, cf. Butt and King, this volume) or something entirely different will be appropriate can only be determined by making strong claims that stand the test against the diversity of natural languages. Linguists criticising syntactic universals often adhere to diversity as a means in itself, but it strikes us that only the interplay between claims embedded in syntactic theories and the diversity of natural languages may eventually allow linguis- tics to become a mature science. With regard to the complementation schema discussed above, we would like to assume that language specific complementation schemata are formally represented as instances of a more abstract, and possibly universal complemen- tation schema, which forms an instance of endocentricity (with a modification schema as another instance of endocentricity, and further instances as well). Couched within the respective theory, the relation between the universal and the language-specific schemata should be represented as simply as possible (if it turns out that the appropriate definition may only refer to dependencies, tree-like structures employed in German or English may be viewed as epiphenomena). Ideally, language-specific schemata should be represented as more specific instances of universal schemata; the specific schemata are still abstract (for instance, they do not make reference to particular categories), and yet they are much more concrete than their universal super-types. Imposing a concept like complementation requires its formal representation within syntactic theory. Hence, it becomes a prerequisite for determining the adequate model- ling in the theory, and a touchstone for the theory itself. The quest for formal universals has thus an immediate bearing on determining the competition of syntactic theories, and clearly cannot be established outside the realm of a theory (which may lead to the confusion that is e.g. expressed in Evans and Levinson (2009: 436−437)). With this background, let us enter unto a discussion of what we see as misunderstood universal categorizations. We assume that most (if not all) linguists would agree that a term like relative clause is much less a universal category than a pure means of descrip- tion. It is less clear whether a relative clause is an instance of a more general operation, where a phrasal modifier is related to a modified element through a marker (the marker itself may be present or not, and may be more or less abstract). Gil (2001: 107−108) tries to show that clausal association to nominals in Hokkien Chinese, as illustrated in (4), should not be confused with the concept of a relative clause in English or German. [Hokkien Chinese] (4) a44(>44) -beŋ24 Ah Beng bue53 buy e24>22 ASSOC ph eŋ24>22 -ko53 apple ‘apples that Ah Beng bought’ (Gil 2001: 107) Because the attributive marker e24 (the superscript indicates a tone; x > y indicates tonal sandhi) is not only used in Hokkien to establish a relationship between a clause and a nominal, but is also required to establish relations between adjectives and nouns, PPs
- 28. I. Introduction: Syntax in Linguistics 10 and nouns, and further adnominal elements and nouns, Gil (2001: 108) concludes that the concept relative clause should not be used in Hokkien grammar. While this may be correct, it also seems trivial. However, it seems plausible to assume that the attributive construction in Hokkien and relative clauses in other languages share a common core in that a nominal element is formally linked to another phrase, and establishes a restrictive interpretation through the combination. A relative marker (not to be confused with the relative pronoun, and hence possibly without phonetic signature) in English or German establishes what is established by e24 in Hokkien. If we abstract over this particular property and indicate that we are not calling a relative clause a universal, we might be able to identify a possibly universal mode of combination (cf. for instance, Kayne 1994 and subsequent work). With regard to the question raised in this section, and the current state of affairs regarding its answer, it seems to us that syntactic research should focus on identifying or rejecting more abstract universals as part of syntactic theories (as e.g. complementa- tion). In addition, it seems that at least some rather basic universals (such as employing noun and verb as categories) have survived recent iconoclastic attempts to demolish them (once again, the reader is referred to the language sketch of Northern Straits Salish in Czaykowska-Higgins and Leonard, this volume). So, the nature of linguistic universals is controversial, but it is a fruitful controversy. 5. What is the object of study? After Saussure’s introduction of the distinction between langue, parole and faculté de langage, every syntactic theory must position itself with regard to its actual object of study. Saussure made it clear that the object of study is mesmerizingly difficult: “Quel est l’objet à la fois intégral et concret de la linguistique? La question est particulièrement difficile” (Engler 1968: 123). Van Riemsdijk (1984: 1) states that “we regard the object which we study as a real object, a mental organ as Chomsky has put it. From this it also follows that what we study is the grammar, not the language.” Van Riemsdijk concludes that language acquisition is the central problem of syntactic theory. This position has been attacked from various perspectives − not the least one being that there seems to a conflation of a variety of concepts. For instance, there can be real objects without being related to a mental organ. In their introduction to one of the strictest attempts to provide a formal analysis of natural language, Gazdar et al. (1985: 5) state that “[o]ur linguistic theory is not a theory of how a child abstracts from the surrounding hubbub of linguistic and non-linguistic noises enough evidence to gain a mental grasp of a natural language.” Other syntacticians attack the idea that grammar comes first and that language is derived from grammar. At least some versions of Construction Grammar (cf. Fried, this volume) assume that the grammar is induced from language data. It seems uncontroversial, how- ever, that language in general is a bio-cultural hybrid, partly the product of biological evolution, but also partly the product of our own making, but this characterization applies to human beings in general, and may thus not be a specific property of language, or a property only in the slightly circular sense that it is human language we are dealing with. The objects of syntactic study have been delimited by means of various distinctions that prove to be problematic in themselves. There is the celebrated distinction between
- 29. 1. Syntax − The State of a Controversial Art 11 competence and performance, as introduced in Chomsky (1965), and later supplanted by a distinction between internal and external language in Chomsky (1986). Here, van Riemsdisk (1984: 3) remarks that “everyone lives by it”. We are not so sure that this is actually correct. The latter distinction has been utilized to separate theories into those that address internal and those that address external language. But even if we followed this direction and assumed that Gazdar et al. (1985) investigated external language, while Chomsky (1986) was concerned with internal language, we cannot be sure that the separation actually yields different objects, as it is purely noumenal. Drawing such an arbitrary distinction will lead to confusion at best, and will weaken the field as a whole at worst. In any case, there are many syntactic models that are agnostic to the distinction, as well as to the distinction between competence and performance. Similarly, the object under syntactic study was delimited by the distinction between core and periphery within the Theory of Principles and Parameters. If one goes through the discussions of the past 20 years, however, one easily detects that today’s periphery might become tomorrow’s core and vice versa. Müller (2013) discusses Exceptional Case Marking (ECM, more familiar in traditional grammar under the rubric ‘accusative- cum-infinitive construction’) as an illustration for this tendency: Chomsky (1981: 70) assigned ECM to the periphery, but 15 years later, ECM became a building block of case assignment in specifier-head-configurations. The morale to be drawn here is threefold. Firstly, it seems to us that the distinctions (competence/performance, core/periphery, etc.) have been useful in the past since it al- lowed syntactic research to start somewhere without being constantly endangered by requests to also account for other, presumably related phenomena. In this sense, early syntactic theories had to be reductionist, and introduced distinctions to this end. Syntax in its current guise should give up reductionism. Secondly, it seems that these distinctions have failed to provide an instrument to gauge whether syntactic models make headway or not. If we take a fixed concept of a core that has to be described or explained by syntactic theory, then we can assume that a model should be abandoned (or at least very critically scrutinized) that does not provide analyses for phenomena that unequivocally belong to a “core”. Progress could be charac- terized by accounting for an agreed-upon set of phenomena, and could be gauged by consecutively extending the core that is covered in this way, until it becomes clear that further extensions would lead to analyses of phenomena that form part of the periphery because they emerge as cultural artefacts that have no biological basis. It would also allow syntactic analysis to establish a canon of analysed phenomena (where a canon is understood in the sense of engineering and natural science, i.e. as foundational part that has already been successfully covered), and to proceed from this canon to areas of syntactic research that still await detailed scrutiny. Yet, syntactic theories often have not developed in this way. As a striking illustration consider the treatment of negative place- ment in inverted English sentences. (5) a. Can John not leave? b. Will John not leave? Almost 25 years after their celebrated analysis in Syntactic Structures (Chomsky 1957), Lasnik (2000: 151) admits that “[t]here are about a dozen really fancy ways in the literature to keep [the sentences in (5)] from violating the [Head Movement Constraint].
- 30. I. Introduction: Syntax in Linguistics 12 To the best of my knowledge, not one of them really works.” One does not have to understand the workings of the Head Movement Constraint to see that any syntactic analysis blocking the derivation of the examples in (5) may be in need of refurbishment. While such renovations took place, they were not initiated by indications of prior failure and did not lead to an overall improvement. What are the conclusions here? It sounds awkward to assume that negative placement in English belongs to the periphery, and even absurd to delegate it to issues of perfor- mance. But it might very well be that syntactic theories should not generalize from rather singular language-specific properties to propose universal constraints. Perhaps the general ignorance towards this construction has to do with the conclusion that the con- struction is in fact peculiar − but a distinction between competence and performance, or core and periphery adds little to this insight. Thirdly, and finally, it should be noted that the distinction between competence and performance also provides the ground for the distinction between grammaticality and acceptability (the former only affected by competence, the latter affected by factors of performance as well). This is important insofar as psycholinguistic experimentation, which seems to be the most plausible methodology to arrive at conclusions about the biological side of syntax, can only access acceptability, i.e. is always “blemished” by performance (cf. Schütze 1996 and Felser, this volume). In light of the problems mentioned, it seems wise to consider the question of the object under study with a bit of agnosticism. Currently, it is a majority view that the objects of study are real objects (but one also finds the position that language is princi- pally a cultural creation), but it does not follow that all syntacticians agree in investigat- ing a mental organ, and moreover that it remains unclear how such an organ could be assessed in the first place. 6. Generative grammar With the term generative grammar, we seem to have arrived at a completely uncontrover- sial spot, given the definition provided by van Riemsdijk (1984: 2): “anyone is a genera- tive grammarian (…) who holds that the grammar must be a finite rule system which explicitly characterizes an infinite set of sentences.” Even linguists who would not char- acterize themselves as generative grammarians agree to versions of this statement, as e.g. Martin Haspelmath (2010: 666): “To describe a language, one needs categories, because it is not possible to list all the acceptable sentences of a language.” 7. There is more than one way to do it (presumably) The final quote from Henk van Riemsdijk’s list serves to illustrate that syntactic research has indeed progressed considerably in the past 30 years: “Modern linguistic research is characterized by a comparative freedom of strict methodological imperatives. In this sense, it is clearly different from corpus linguistics, (…) and the statistical straight-jacket of the social sciences. (…) most grammarians will tend to have most faith in serious introspection.” (van Riemsdijk 1984: 3) This statement is interesting insofar as linguistics
- 31. 1. Syntax − The State of a Controversial Art 13 (syntax in particular) strives to become acknowledged not as part of the humanities, but as part of the natural sciences. Whatever one might say about this, it should be clear that the methodological constraints within the natural sciences might be even more rigid than the ones of the social sciences (with the usual proviso that we should not forget the lesson learned by Paul Feyerabend’s Against Method). But the quote reflects the time when syntax had to work with the resources that were available. Without proper comput- ing machinery, the development of corpora was problematic, syntactic annotation was practically unknown, and hence data analysis was much less a question of methodologi- cal orthodoxy than of impossibility. All this has changed. In addition, we have a much better understanding of the limits of introspection, and may also use controlled experi- ments to arrive at conclusions about data. So, from hindsight, it appears that 30 years ago, syntactic research was free from one particular methodological imperative for the simple reason that the methodology was still in its infancy. With the advent of advanced computing devices and a deeper understanding of controlled experimentation, the current syntactic researcher has not only much more freedom to decide whether data should be derived from introspection, from corpus search, or from experimentation. We also under- stand now that data sometimes cannot be assessed through introspection, just as scarcely occurring data cannot easily be assessed through corpus-based methodologies (for a case study, cf. Meurers and Müller 2009). It is thus handy that more than one method is at our disposal. While methods flourish, other − more practical, as well as socio-economical − con- straints affect syntactic research (as well as linguistics in general). Linguistics remains a rather small field, the subject under study, however, provides an abundance of data that in many cases resist automated analysis and thus requires human analysis (in this respect, we fully agree with Evans and Levinson 2009). It would be ideal if all languages could receive the same degree of attention, but syntactic research is not often recognized as a goal in itself. Finally, there is not only more than one way to do it method-wise, but also with regard to the different frameworks that the syntactician may employ. More often, syntac- ticians do not choose from these frameworks because they belong to schools, but because they see that for the questions at hand, different frameworks may allow different entry points, which eventually may lead to similar conclusions, or even similar ways of analys- ing syntactic data. As an illustration, consider the treatment of long distance dependen- cies. When Gazdar (1981) proposed that certain long distance dependencies could be handled without transformations but by employing complex (SLASH) categories, Chom- sky retorted that GPSG (under which name the new transformationless framework be- came known) was just a notational variant of the current version of transformational grammar. While it seems illogical to call something a “notational variant” that employs less machinery than the theory of which it is considered a variant, the lesson learned was that “notational variants” are to be dodged, if only because they are superfluous. If one compares the current-day treatment of long distance dependencies in certain versions of Minimalism (cf. Richards, this volume) with the treatment of long distance dependen- cies in HPSG (S. Müller, this volume, HPSG in this sense is an offspring of GPSG), one is tempted to assume that the treatment must be very different for the very reason that the first framework calls itself derivational, while the second is termed representational. Closer scrutiny reveals, however, that the strictly local representational treatment (in terms of local trees) is mirrored in strictly local versions of derivational Minimalism. It
- 32. I. Introduction: Syntax in Linguistics 14 does not matter whether an analysis is based on representations or derivations, if both are strictly local. Just as a syntactician cannot look “into” a local representation, he or she cannot look into the history of a local derivation. In this light, the so-called uniform- path analyses in derivational Minimalism cannot be distinguished from SLASH-based analyses in representational HPSG (cf. Alexiadou, Kiss, and Müller 2012: 30). What is more, a replacement of uniform-path analyses in favour of punctuated-path analyses, which would not be “notational variants” of SLASH projection, cannot be justified by observable consequences (i.e. empirical differences) (again, cf. Alexiadou, Kiss, and Müller 2012: 30−31). The new lesson to be learned is that it may perhaps be not only be useful to attack problems from different vantage points, but also to observe what the other camp is doing. If the results gained are similar, they are so not because they are “notational variants”, but perhaps because they represent the limit of current expertise in the worst case, and the true nature of the object under investigation in the best. 1. Acknowledgements This article has benefited from comments by Stefan Müller, for which we are grateful. 8. References (selected) Alexiadou, Artemis, Tibor Kiss, and Gereon Müller 2012 Local modelling of non-local dependencies in syntax: An introduction. In: Artemis Alex- iadou, Tibor Kiss and Gereon Müller (eds.), Local Modelling of Non-local Dependencies in Syntax, 1−48. Berlin/Boston: Walter de Gruyter. Chomsky, Noam 1965 Aspects of the Theory of Syntax. Cambridge/London: The MIT Press. Chomsky, Noam 1986 Knowledge of Language. Its Nature, Origin, and Use. New York: Praeger. Engler, Rudolf (ed.) 1968 Ferdinand de Saussure, Cours de Linguistique Generale. (Édition critique par Rudolf Engler, tome 1.) Wiesbaden: Harrasowitz. Evans, Nicolas, and Stephen C. Levinson 2009 The myth of language universals: Language diversity and its importance for cognitive science. Behavioral and Brain Sciences 32: 429−492. Gazdar, Gerald, Ewan Klein, Geoffrey K. Pullum, and Ivan A. Sag 1985 Generalized Phrase Structure Grammar. London: Blackwell. Gil, David 2001 Escaping Eurocentrism: fieldwork as a process of unlearning. In: Paul Newman and Martha Ratliff (eds.), Linguistic Fieldwork. Cambridge: Cambridge University Press, 102−132. Harris, Randy Allan 1993 The Linguistics Wars. Oxford: Oxford University Press. Haspelmath, Martin 2010 Comparative concepts and descriptive categories in crosslinguistic studies. Language 86: 663−687.
- 33. 2. Syntactic Constructions 15 Koenig, Jean-Pierre, and Karin Michelson 2012 The (non)universality of syntactic selection and functional application. In: Christopher Piñon (ed.), Empirical Issues in Syntax and Semantics 9: 185−205. [http://www.cssp. cnrs.fr/eiss9/] Müller, Stefan, and Stephen Wechsler 2014 Lexical approaches to argument structure. To appear in Theoretical Linguistics 40. Pollard, Carl, and Ivan A. Sag 1994 Head-driven Phrase Structure Grammar. Chicago: University of Chicago Press. Riemsdijk, Henk van 1984 Introductory remarks. In: Wim de Geest and Yvan Putseys (eds.), Sentential Complemen- tation, 1−9. Dordrecht: Foris Publications. Schütze, Carson 1996 The Empirical Base of Linguistics: Grammaticality Judgments and Linguistic Methodol- ogy. University of Chicago Press. Tibor Kiss, Bochum (Germany) Artemis Alexiadou, Stuttgart (Germany) 2. Syntactic Constructions 1. Introduction 2. Examples of constructions 3. Constructions and syntactic theory 4. Conclusion 5. References (selected) Abstract The term construction is widely used descriptively in discussing grammar, and is still used informally in most theoretical work for characteristic formal patterns of syntactic categories or features, usually associated with a meaning and/or function. Modern lin- guistic theories employ a range of formal devices to produce or characterize surface constructions; they may be rules, or schemata, or constraints. It is usually assumed that competence in a language consists largely of these formal devices together with a lexi- con; the constructions themselves are epiphenomenal. As such, constructions are an abstraction over the data which linguistic theory must analyze; insight in syntax is achieved through discovering generalizations across constructions. 1. Introduction The term construction is ubiquitous in contemporary syntactic literature, being used informally to refer to linguistic expressions in a variety of ways. The term also has a technical sense in the theory of Construction Grammar, as detailed in Chapter 28.
- 34. I. Introduction: Syntax in Linguistics 16 The term construction is widely used to characterize certain kinds of form−meaning pairings, as when we refer to “possessive constructions” or “the verb−particle construc- tion” to refer to examples like those below. (1) Three examples of possessive constructions a. Seymour’s new friend b. a new friend of Seymour’s c. Seymour has a new friend (2) Three examples of the verb-particle construction a. We picked up a lamp at the flea market. b. We picked a lamp up at the flea market. c. What did you pick up at the flea market? In general, linguists would not refer to the three examples in (1) as comprising a single possessive construction, because they are too different in their syntax; in (1a) the posses- sor precedes the possessed noun, in (1b) the possessor follows the possessed noun, and in (1c) the possessor is expressed as a distinct argument outside the possessed noun phrase. These differences represent three different ways of expressing the concept of possession, in English. In (2), on the other hand, many linguists would be inclined to refer to all three sentences as manifesting a single verb-particle construction, on the basis of the perceived similarity of the syntax of the three cases. There is a very large class of verb−particle pairings which allow the ordering alternation shown in (2a)−(2b), where the order re- flects no apparent difference in meaning (such as drop off, smash up, fix up, turn on, leave out). In such cases, the object can systematically be the focus of a question, as in (2c); so the general consensus would be that these three sentences illustrate the verb- particle construction. On this view, (1) illustrates three different form−meaning pairings, even though one component of the meaning is shared across all three, while (2) illustrates one form− meaning pairing, even though independent factors distort the shared form (and the cor- rect characterization of the meaning component may be elusive). At the same time, (2c) illustrates a wh-question construction, in addition to the verb- particle construction. Since the properties of the wh-question construction (e.g. wh-ex- pression in clause-initial position, auxiliary in second position) are independent of the verb-particle construction (e.g. the predicate includes a particle like up, down, out, etc.), there is no motivation for formulating a distinct “verb-particle wh-question construction.” 1.1. Toward a definition The term construction is not a technical term (outside of Construction Grammar), and consequently it is difficult to define. As an approximation, it can be defined as follows. (3) A construction is a characteristic formal pattern of syntactic categories or features, usually associated with some meaning and/or discourse function.
- 35. 2. Syntactic Constructions 17 The use of the word pattern here is an attempt to be as theory-neutral as possible; a pattern might be a structure, or a template, or the output of a rule. The notion formal is meant to include aspects of form which are of significance to grammar. In some theories word order is a primitive of grammar, while in other theories word order is derived from structure, such that structure, but not linear order, would count as a formal property (see Chapters 17 and 40 on Word Order). The notion syntactic categories is intended to include major parts of speech but also minor or functional categories such as the class of English verb-particles, or the class of determiners. The notion syntactic features in the definition is meant to include morpho- syntactic features such as the past participle or dative case but also semantico-syntactic features such as negation or “wh” (borne by interrogative expression like what and who). Together syntactic categories or features includes function words such as infinitival to and bound formatives such as possessive ’s, on any analysis. The definition in (3) is meant to exclude phonology and surface exponence, which do not characterize constructions as the term is ordinarily used in mainstream syntax. For example, we would not expect to find a construction which necessarily involved words beginning with the phoneme /w/, even if we speak loosely of various kinds of wh-constructions. Similarly, if there is more than one formal category in English which is spelled out by the suffix -ing, then we expect a construction to be identified by the underlying features which are being spelled out (e.g. progressive, or gerund), not by the phonological form of the exponent doing the spelling out. Though we might descriptively call something an Acc-ing construction, for example, in a more careful statement of its characteristics, we would distinguish the feature or category that -ing manifests. Thus, the definition offered above is intended to stress syntactic form, not phonologi- cal form. This is in accord with the usual use of the word construction in syntax. An idiom like kick the bucket meaning ‘die’ requires the lexical items kick and bucket, and hence makes direct reference to exponents with phonological content (see Chapter 23 on Idioms). As such, ordinary usage would not make reference to a kick the bucket construction (Construction Grammar, however, is different here: idioms are considered to be constructions). In this way, more or less functional elements like the interrogative pronoun what and the light verb do are treated together with syntax as opposed to lexical items like kick and bucket. Thus is it not controversial to speak of a construction of the form What’s X doing Y? meaning roughly ‘Why is X Y?,’ where X is a subject and Y is a predicate. For example: What’s the newspaper doing in the bushes? or What are you doing leaving without your shoes? This construction requires what and do as well as the progressive with an appropriate form of be. The definition offered in (3) also suggests that a construction is usually associated with some meaning and/or discourse function. The importance of meaning is somewhat loosely applied in practice. Thus, it is not usually considered necessary to have a rigorous statement of the meaning of a possessive construction like the one in (1a) in order to call it a construction, if it has a clearly defined syntactic form. But if there are two disjoint meanings involved, then it is common to think of them as involving two distinct constructions. For example, in English, the auxiliary inverts with the subject when a wh- item is fronted, but also when a negative element is fronted, as in (4) (cf. Chapter 12, on V2).
- 36. I. Introduction: Syntax in Linguistics 18 (4) a. Which of them would he recommend? b. None of them would he recommend. Even if the syntax of the inversion is identical in the two cases, it would be most natural here to speak of subject-auxiliary inversion constructions in the plural, rather than a single subject-auxiliary inversion construction which was indifferent to whether the ini- tial element was a negative or an interrogative phrase−though practice varies somewhat here (and in Construction Grammar, there is no limit to how abstract a construction can be). The notion of discourse function in (3) is intended broadly, to include various prag- matic inferences and affect. For example, the What’s X doing Y? construction is only used when there is some sense that it is incongruous or inappropriate for X to be Y (as discussed by Kay and Fillmore 1999). Thus the question Why are men rebelling? can be asked in a range of contexts, but What are men doing rebelling? can only be asked if there is some salient sense (perhaps the speaker’s opinion, but not necessarily) that it is inappropriate, incongruous, or outrageous for them to be doing so. 1.2. Applying the definition Returning to the examples in (1)−(2), we can apply the definition offered in (3) to show that it is harmonious with the common intuition that (1) illustrates three different posses- sive constructions while (2) represents three different manifestations of a single construc- tion. According to (3), a class of phrases or sentences must share a characteristic formal pattern in order to belong to a single construction. The formal differences among the three examples in (1) are fairly clear; (1a) lacks of and the possessor precedes the posses- sum, while (1a) contains the function word of, and the possessor follows the possessum. Furthermore, in (1a), the possessum is understood as definite, while in (1b), the posses- sum is indefinite. The example in (1c) is predicative, and contains the verb have. So the fact that the three expressions describe the same semantic relation is not normally taken to imply the existence of a single possessive construction. Thus it seems that the defini- tion appropriately picks out each of the three as a construction. Turning to (2), we can first address the question of whether (2c) represents a different construction from the other two. Of course it does, as it involves wh-movement, but this is irrelevant to the verb−particle construction. The interrogative construction simply ap- plies to a clause that has a verbal particle in the predicate, just as it does in an ordinary transitive clause. The second question is whether there is motivation to treat examples like (2a) and (2b) as distinct constructions. This cannot be conclusively determined without formal analysis. By and large, the two are distinguished only by word order, not by meaning, nor by functional categories or features. There are some differences, for example (2b) allows the object to be pronominal (We picked it up), but (2a) does not (*We picked up it). If such differences can be independently explained, then an analysis can be motivated in which there is a single construction with some flexibility of order, that is, a single ‘characteristic formal pattern of syntactic categories or features’ in which whatever deter-
- 37. 2. Syntactic Constructions 19 mines the placement of the particle before or after the internal argument is not character- istic, or is not a syntactic feature. This is the usual consensus (see Ramchand and Sven- onius 2002 for one such account), though alternative analyses exist in which the two represent distinct constructions (see Farrell 2005). 2. Examples of constructions To further illustrate the notion of construction, several examples of constructions are listed in this section. No detailed characterization or analysis is attempted. The examples serve simply to give an impression of a small part of the range of syntactic constructions. 2.1. Argument structure, grammatical functions A number of constructions involve valency, argument structure, and grammatical func- tions (see Chapter 8 on Grammatical Relations). For example, a passive construction involves the demotion of the external argument, compared to the active use of the same verb (see Chapter 22 on Voice). In English, the demoted external argument can be ex- pressed in a by-phrase or left implicit, and the verb appears in a past participle form, with a form of the auxiliary be. The implicit argument can control a purpose clause, as illustrated in (5b), just as with the active construction in (5a). English also has a middle construction, in which the external argument cannot appear in a by-phrase, and is not syntactically active as diagosed by a purpose clause, as illustrated in (5c). (5) a. The owner sold the house (to pay off debts). b. The house was sold (by the owner) (to pay off debts). c. Houses sell easily (*by the owner) (*to pay off debts). In a conative construction, illustrated in (6b), the internal argument is oblique and is not as fully affected as in a regular transitive construction, compare (6a). In a benefactive construction, illustrated in (6c), an indirect object derives some benefit from the action, or comes into possession of the direct object (see Chapters 9 and 37 on Arguments and Adjuncts). (6) a. The baker cut the bread. b. The baker cut at the bread. c. The baker cut me a piece of bread. Resultative and depictive constructions involve secondary predicates, as illustrated in (7a) and (7b) respectively. (7) a. They drank the bar dry. b. They ate the bread dry. Control, raising, and so-called Exceptional Case Marking (ECM, or accusative-with- infinitive, or subject-to-object raising) constructions usually involve infinitive comple-
- 38. I. Introduction: Syntax in Linguistics 20 ments in which arguments are, descriptively speaking, shared across the two clauses (see Chapters 14 and 38 on Control). (8) a. Ian wants to stay at the house. (control) b. Ian seemed to stay at the house. (raising) c. We believed Ian to stay at the house. (ECM) English provides many more examples of constructions involving various configurations of arguments, and other languages provide yet more. In some cases, the availability of a construction may be tied to the availability of a lexical item; for example, it may be that if a language has an ECM construction if and only if it has a verb with the selectional properties exhibited by English believe in (8c). However, in other cases, the availability of a construction in a given language does not seem to be connected to lexical items. An example is the resultative construction illustrated in (7a), which many languages lack, despite having verbs and adjectives which are otherwise like the ones used in the English resultative construction (Snyder 2001). 2.2. Unbounded dependencies There is a range of constructions which have been analyzed as involving displaced con- stituents, or filler-gap dependencies. These include (in the order in which they appear in [9]): wh-questions, embedded wh-questions, clefts, pseudoclefts, relative clauses, com- parative constructions, and (contrastive) topicalization (for some discussion of some of these and related constructions, see Chapter 12 on V2, Chapter 13 on Discourse Configu- rationality, and especially Chapter 21 on Relative Clauses and Correlatives). (9) a. Which book did you read? b. I asked which book you read. c. It was this book that I read. d. What I read was this book. e. The book that I read was long. f. She read a longer book than I read. g. This book, I read. These constructions have in common that the dependency can cross finite clause bounda- ries. (10) a. Which book did your mother think you read? b. I asked which book your mother thought you read. c. It was this book that my mother thought I read. d. What my mother thought I read was this book. e. The book that my mother thought I read was long. f. She read a longer book than my mother thought I read. g. This book, my mother thought I read.
- 39. 2. Syntactic Constructions 21 In this respect they contrast with passive, as illustrated in (11a), and for example raising, as illustrated in (11b) (compare [12]). (11) a. *The house was thought (by his mother) the owner sold. b. *Ian seemed stayed at the house. (12) a. It was thought the owner sold the house. b. It seemed Ian stayed at the house. In fact, the constructions in (9)−(10) can cross indefinitely many finite clause boundaries, and for this reason are known as unbounded dependencies. (13) a. Which book did you think your mother thought you read? b. I asked which book you thought your mother thought you read. c. It was this book that I thought my mother thought I read. d. What I thought my mother thought I read was this book. e. The book that I thought my mother thought I read was long. f. She read a longer book than I thought my mother thought I read. g. This book, I thought my mother thought I read. The availability and properties of unbounded dependency constructions can vary some- what from one language to the next. For example, sometimes there is a resumptive pronoun in the gap position, and in other cases there is no displacement on the surface, with the filler element remaining in situ. Unlike the case with argument structure alterna- tions, this kind of variation tends not to be dependent on lexical items, though it may be connected to the properties of functional elements such as that and which. In fact, it has been proposed that properties of constructions are largely determined by the properties of their heads. This is explicit in the name of the theory Head-Driven Phrase Structure Grammar (see Chapter 27 on HPSG), but is also a common assumption in other theories (e.g. Borer 1984). For example, the properties of a relative clause could conceivably be entirely determined by the cluster of features contained on its (possibly abstract) head. The head would be a kind of C[omplementizer] taking a finite TP comple- ment, attracting a suitable nominal element to its specifier, and projecting a category which could be used as a nominal modifier. 2.3. Complex constructions In this regard, constructions like the What’s X doing Y? construction mentioned above are complex, as they appear to involve an interdependency of several heads. Mainstream theory would probably treat a construction of this kind as a special kind of idiom. Construction Grammar, in contrast, holds that there is no principled difference be- tween a fully general construction and a highly idiomatic one, or even between a fully general abstract construction and a lexical item; each is a pairing of form with function, broadly construed.
- 40. I. Introduction: Syntax in Linguistics 22 3. Constructions and syntactic theory Traditional descriptive grammars may characterize and exemplify a list of constructions in a given language. They may organize the constructions according to perceived similar- ities, and may attempt to state generalizations which transcend the individual construc- tions. This much is extremely useful in a reference grammar (see Chapter 59 on Refer- ence Grammars). Modern syntactic theory necessarily goes further, and is based on the assumption that higher-level generalizations are necessary in order to achieve what Chomsky (1964) called explanatory adequacy, a model of language which accounts for how individual languages are learnable by children. Traditional grammar posited rules to characterize or generate constructions, such as passive and relative clauses. Generative grammar took this as a starting point and went on to abstract properties from classes of rules, such as elementary transformations and different kinds of formal conditions constraining them. Generative grammar in the 1970’s explored the ways in which the properties interacted in rule systems, and sought to discover constraints on them. For example, Emonds (1976) observed that transformational rules did not produce structures unlike those which had to be base-generated (“structure preservation”), sug- gesting that the full power of transformations was unneeded. Chomsky (1977) showed that the class of unbounded dependencies displayed highly uniform properties, suggest- ing that they were not produced by distinct rules. Subsequent work has increasingly focused on higher-level generalizations over rule types, shedding much light on the nature of grammar. Thus, generalizations about constructions involving valence and argument structure led to those being analyzed in Lexical-Functional Grammar as the output of lexical rules (see Chapter 25 on LFG), accounting, for example, for their structure-preserving proper- ties and their relative sensitivity to lexically listed traits of individual verbs. Similarly, generalizations about the class of unbounded dependencies as a whole led to the develop- ment of Move-α in Government-Binding Theory (Chomsky 1981), and the SLASH fea- ture as used in Head-Driven Phrase Structure Grammar (see Chapter 27 on HPSG). The development of Principles and Parameters theory (Chomsky 1981) involved re- thinking the nature of rules entirely; once deconstructed into a system of invariant univer- sal principles interacting with parametric points of variation, there are no rules per se. This is expressed in the following quote from Lectures on Government and Binding: “The notions “passive,” “relativization,” etc., can be reconstructed as processes of a more general nature, with a functional role in grammar, but they are not “rules of gram- mar”” (Chomsky 1981: 7). Since constructions were the output of rules in the traditional conception of grammar, the elimination of rules from the theory means that in a Principles and Parameters frame- work, constructions are epiphenomenal, as reflected in the following quote, also from Chomsky but a decade later: “A language is not, then, a system of rules, but a set of specifications for parameters in an invariant system of principles of Universal Grammar (UG); and traditional grammatical constructions are perhaps best regarded as taxonomic epiphenomena, collections of structures with properties resulting from the interaction of fixed principles with parameters set one or another way” (Chomsky 1991: 417).
- 41. 2. Syntactic Constructions 23 Since that time, although the notion of parameter has been substantially rethought, mainstream syntactic theories continue to regard the notion of construction, like the notion of rule or transformation, as a descriptive stepping stone on the path to greater understanding rather than as an analytic result in its own right. Work in Construction Grammar, too, recognizes that insight does not come from simply listing the individual surface constructions in a language, and therefore, like other theories, seeks generalizations over constructions; the difference between Construction Grammar and other theories mentioned here is that in Construction Grammar, the gener- alizations are themselves modelled as abstract constructions. Nor is this just a termino- logical distinction: the claim in Construction Grammar is that the generalizations have the same kinds of properties as the constructions themselves, at a suitable level of ab- straction. 4. Conclusion I have characterized a construction as a characteristic formal pattern of syntactic features, usually associated with some meaning or discourse function. The way constructions are characterized in descriptive work tends to combine morphological, syntactic, and seman- tic facts. Purely phonological facts, however, tend not to be significant for the characteri- zation of constructions. The precision of the syntactic characterization is usually taken to be more important than the semantic description of the meaning of a construction (e.g. even if two possessive expressions had exactly the same meaning, substantially different syntaxes would usually lead to them being classified as distinct constructions). Some such notion is descriptively indispensible in syntactic work. Descriptive gram- mars must contain detailed characterizations of surface properties of constructions, and all careful empirical work makes reference to numerous construction types, as will be seen in the other chapters in this work. Linguistic theory advances through the careful examination of the properties of con- structions. Ultimately, principles are sought which transcend the individual constructions themselves, and the construction itself can be taken to be a taxonomic artifact. 5. References (selected) Borer, Hagit 1984 Parametric Syntax. Dordrecht: Foris. Chomsky, Noam 1964 Current Issues in Linguistic Theory. The Hague: Mouton. Chomsky, Noam 1977 On wh-movement. In: Peter Culicover, Thomas Wasow, and Adrian Akmajian, (eds.), Formal Syntax, 71−132. New York: Academic Press. Chomsky, Noam 1981 Lectures on Government and Binding. Dordrecht: Foris. Chomsky, Noam 1991 Some notes on economy of derivation and representation. In: Robert Freidin (ed.), Prin- ciples and Parameters in Comparative Grammar, 417−454. Cambridge, Massachusetts: MIT Press.
- 42. I. Introduction: Syntax in Linguistics 24 Emonds, Joseph E. 1976 A Transformational Approach to English Syntax: Root, Structure-Preserving, and Local Transformations. New York: Academic Press. Farrell, Patrick 2005 English verb-preposition constructions: Constituency and order. Language 81 1: 96−137. Kay, Paul, and Charles J. Fillmore 1999 Grammatical constructions and linguistic generalizations: The What’s X doing Y? con- struction. Language 75 1: 1−33. Ramchand, Gillian, and Peter Svenonius 2002 The lexical syntax and lexical semantics of the verb-particle construction. In: Line Mik- kelsen and Christopher Potts (eds.), Proceedings of WCCFL 21, 387−400. Somerville, Massachusetts: Cascadilla Press. Snyder, William 2001 On the nature of syntactic variation: Evidence from complex predicates and complex word-formation. Language 77: 324−342. Peter Svenonius, Tromsø (Norway) 3. Syntax and its Interfaces: An Overview 1. Introduction 2. Syntax and the lexicon 3. Syntax and morphology 4. Syntax and phonology 5. Syntax and semantics 6. Syntax and pragmatics 7. Conclusion 8. Abbreviations 9. References (selected) Abstract The notion that knowledge of language comprises separate modules or levels is central to linguistic research, though the precise subcomponents of grammar, their rules and primitives, the extent to which they overlap, and the nature of the architecture of which they are a part are still the subject of intense debate. Investigation into interface phenom- ena necessarily straddles the traditional boundaries of linguistic enquiry and thus pro- vides an important perspective not only on specific empirical issues, but also on broader theoretical matters including the organization of the grammar itself. The focus of this chapter is the interaction of syntax with other modules of grammar which have been posited: the lexicon, morphology, phonology, semantics, and pragmatics. Each section explores a different syntactic interface through discussion of contemporary empirical and theoretical research undertaken in a variety of frameworks.
- 43. 3. Syntax and its Interfaces: An Overview 25 1. Introduction The aim of work on syntax and its interfaces is to identify the basic properties which other modules share with syntax and thus the precise domain of syntax within the gram- mar. In this respect, there are three central questions to be answered: How is syntax distinct from other components of the grammar? To what extent do syntactic processes operate within the grammar? How is syntax related to other levels of linguistic represen- tation? Addressing these issues has implications for fundamental aspects of the grammat- ical architecture. Any theoretical framework will ultimately be judged on the extent to which it is able to capture the relevant generalizations about syntax and its interfaces with regard to the autonomy of posited components, the input−output relations which exist between them, and language structure versus usage. The issue of autonomy naturally precedes any discussion of the interfaces which link modules of the grammar. With respect to syntax, the question is to what extent its primi- tives and operations influence aspects of linguistic structure which might otherwise be treated as belonging to another component of the grammar, and vice versa. For instance, debate continues regarding whether morphology exists as an autonomous component of the grammar with its own combinatory principles. It has been argued that positing a separate part of the grammar to account for apparently morphological phenomena is unmotivated and unnecessary because the features of complex word formation are con- sistent with independent syntactic principles. Data which indicate that the relationship between morphology and syntax is not always isomorphic undermine the claim that morphology can be reduced entirely to syntax. Some approaches imply that syntax itself is not an independent component, raising the question of whether syntax represents an abstraction of the linguist or a natural object. For example, Head-Driven Phrase Structure Grammar (Pollard and Sag 1994) includes a single level of structure SYNTAX-SEMANTICS. Any approach which posits autonomous components of the grammar must also define the input−output relations which connect them. Precisely how syntax interacts with each component in terms of whether the relationship is serial or parallel, unidirectional or bidirectional is still the subject of much debate. For instance, a strictly serialist concep- tion of the interface between syntax and phonology, according to which the former serves as input to the latter, predicts the influence of syntax on phonology and excludes the possibility of interaction in the opposite direction. Data which appear to show that pro- sodic weight may influence syntax represent a challenge to such a unidirectional charac- terization of the syntax−phonology interface. The nature of the links between different components of the grammar continues to be the focus of research. Even the standard assumption that the lexicon serves as input to syntax has been challenged (Borer 2005). Finally, there is the matter of how the architecture of the grammar reflects a structural as opposed to a usage-based approach to language, an issue perhaps most clearly illus- trated by the interaction between syntax and pragmatics. For instance, Mandarin Chinese is claimed to be a language whose syntax is organized in terms of the distinction between topic (that which is being discussed) and comment (what is being said about the topic) rather than grammatical relations such as subject and object (Li and Thompson 1976). Accounts of such interaction in terms of the organization of the grammar will contribute to the wider debate on the relative merits of a structural versus a usage-based model of language. Research into syntax and its interfaces thus has implications for the architecture of the grammar which are relevant across theoretical frameworks and sub-disciplines.
- 44. I. Introduction: Syntax in Linguistics 26 2. Syntax and the lexicon 2.1. Introduction While it is often posited that the lexicon exists as a separate part of the grammar, this has not always been the case, nor is it true of all current theoretical frameworks. In those approaches which distinguish the lexicon from the syntactic component, the issue of where the boundary between the two lies has consequences for the interface between them, as well as the organization of the grammar as a whole. In early generative work such as Chomsky (1957), lexical items and morphemes were those units which could be inserted at the terminal nodes of Deep Structure trees by phrase structure rules. Despite their unique status, these elements were not initially treated as members of a component distinct from syntax with its own rules and types of structure. The lexicon was identified as a part of the syntactic Base Component in Stand- ard Theory (Chomsky 1965) and defined as an unordered list of entries containing infor- mation only about a lexical item’s idiosyncratic features (cf. Bloomfield 1933). However, subsequent research has indicated that at least some word-formation processes may differ in key respects from those which form syntactic structure, prompting a reassessment of the contents of the lexicon and its nature. Today, the Lexicalist Hypothesis (Chomsky 1970) still lies at the heart of the debate on the relationship between lexical items and syntax (Section 2.2), as the different analyses of processes such as causative formation (Section 2.3), undertaken within a variety of theoretical frameworks, illustrate. 2.2. The Lexicalist Hypothesis Chomsky (1970) identified regularities in the relation between a nominal and the verb from which it had been derived (1a), but noted that this nominalization process had idiosyncratic features such as restricted productivity (1b), which made it incompatible with syntactic rules characterized as applying without exception. (1) a. John criticized the book. V John’s criticism of the book N b. John amused (interested) the children with his stories. V *John’s amusement (interest) of the children with his stories N (Chomsky 1970: 187−189) To account for such data, Chomsky argued for the inclusion of some operations in the lexicon. Under this lexicalist view, the lexicon served not only as a repository for lexical items, but also as a generative component in its own right. Wasow (1977: 331) contrasted lexical rules with syntactic ones in terms of five criteria; the former and not the latter affected structure, could relate items of different categories, were local, could have ex- ceptions, and applied before any syntactic rules. These distinctions, and the last criterion in particular, assume an architecture of grammar in which the lexicon interacts in a limited way with syntax, providing the minimal units manipulated by syntactic opera- tions. These assumptions underlie the Principles and Parameters model (Chomsky 1981),
- 45. 3. Syntax and its Interfaces: An Overview 27 in which the lexicon is no longer treated as a part of a syntactic component, but rather has equal status with syntax and the other independent levels of linguistic structure posited (Logical Form and Phonetic Form). Central to all formulations of the Lexicalist Hypothesis is the notion that syntactic operations, which follow those responsible for word-formation, are blind to the internal structure of lexical items. This principle distinguishes lexicalist from non-lexicalist ap- proaches to grammar and thus is an important one with far-reaching consequences. Sys- tematic regularities in word-internal structure are accounted for in terms of lexicon- internal processes that are fundamentally different to syntactic ones. The Lexicalist Hy- pothesis has often been revised, and both strong and weak forms of it have been advo- cated, the difference being whether inflection is classed as a syntactic operation along with derivation and compounding or not. Within transformational theory, the Lexicalist Hypothesis has been challenged by Baker (1988), who proposed that the incorporation of a noun into a verb is a word- formation process that is not lexical but syntactic (Section 3.2). This syntactic analysis has been used to account for other word-formation processes such as causative formation (Section 2.3.3) as well. Some researchers go further and reject the Lexicalist Hypothesis entirely, claiming that primitives and processes relevant to lexical items and word-forma- tion are wholly syntactic (e.g. Ramchand 2008). Other work broadly compatible with the Minimalist Program, by contrast, adopts some version of the Lexicalist Hypothesis (e.g. Ackema and Neeleman 2004). In contrast to approaches such as the Minimalist Program, which seek to relate syntac- tic structures via transformational operations and hence are derivational, many lexicalist theories are non-derivational. The latter provide a framework within which to explore grammaticality in terms of how constraints on the correspondences between different levels of linguistic structure, which are simultaneously present, can be satisfied. Such correspondence theories include Head-Driven Phrase Structure Grammar (HPSG; Pollard and Sag 1994), Lexical-Functional Grammar (LFG; Bresnan 2001), Role and Reference Grammar (Van Valin and LaPolla 1997) and Culicover and Jackendoff’s (2005) Simpler Syntax. Construction Grammar (e.g. Fried and Östman 2004) takes a different overall approach, treating the relationship between the lexicon and syntax as a continuum on the basis that both lexical and syntactic constructions represent pairs of form and mean- ing which simply differ in terms of internal complexity. The relationship between syntax and the lexicon has been explored in each framework through analysis of a variety of interface phenomena. In addition to the issue of the relationship between morphology and syntax (Section 3), research has centred on accounting for the realization of argu- ments and argument structure alternations, for example the dative alternation (Goldberg 1995), unaccusativity (e.g. Alexiadou et al. 2004; Levin and Rappaport Hovav 1995), aspect (e.g. Erteschik-Shir and Rapoport 2005) and incorporation (Baker 1988). The different analyses of causativization which have been proposed serve to illustrate the major differences between these approaches with regards to a particular valency-chang- ing operation. 2.3. Causative constructions: lexical or syntactic? The English sentence (2) is an example of a periphrastic causative in which Charlie is the causer, Lily is the causee, and the caused event is Lily laughing. This sentence
- 46. I. Introduction: Syntax in Linguistics 28 describes two separate events expressed by two different verbs, each in a different clause, with the cause of the second event being given in the first clause. Each verb is a separate lexical item that corresponds to a separate syntactic predicate. (2) Charlie caused Lily to laugh. Not all causative constructions are formed in this way though, and the variation that exists brings to light issues which have a direct bearing on the status of lexical items and their interaction with the syntactic component of the grammar. In a Romance lan- guage such as Catalan, the causative construction consists of two verbs, but in contrast to its English counterpart, tests show that it is not biclausal but monoclausal. For exam- ple, while the logical subject of llegir ‘read’ is realized as a grammatical subject in (3a), in (3b) it appears as an object expressed as a prepositional phrase which must follow the other object noun phrase un poema ‘a poem’ (Alsina 1996: 190, 210). This is consist- ent with the causee (the boy) being the third argument of a single syntactic predicate (‘make-read’) in a simple clause. [Catalan] (3) a. El the nen boy llegeiz read un a poema. poem ‘The boy is reading a poem.’ b. El the mestre teacher fa makes llegir read un a poema poem al to.the nen. boy ‘The teacher is making the boy read a poem.’ (Alsina 1996: 190) In the Catalan monoclausal causative construction therefore two separate verb forms combine to give a single syntactic predicate, demonstrating that the overt form of a verb and its function can be separated in this instance. In Japanese, the productive causative formation process involves affixation of the morpheme -(s)ase to the verb, as in (3b). [Japanese] (4) a. Akira-ga Akira-NOM it-ta. go-PST ‘Akira went.’ b. Hiroshi-ga Hiroshi-NOM Akira-o Akira-ACC ik-ase-ta. go-CAUS-PST ‘Hiroshi made Akira go.’ This process would be classed as occurring in the syntax according to Wasow (1977) because it is productive. I refer to this construction in theory-neutral terms as the produc- tive sase causative. Kuroda (1965), amongst others, argues that the productive sase causative construction exemplified in (4) consists of two clauses rather than one, even though the verb is a single word. Evidence for this analysis includes the fact that the inclusion of a manner adjunct, which can be interpreted as modifying either the grammatical subject or the logical subject
- 47. 3. Syntax and its Interfaces: An Overview 29 in a biclausal sentence, also results in ambiguity in a productive sase causative construction, as (5) shows. Hence, productive sase causative constructions are biclausal. [Japanese] (5) Jon-wa John-TOP muriyari forcibly sono the ko-ni child-DAT sono the kutsushita-o socks-ACC ooyorokobide happily hak-ase-ta. put.on-CAUS-PST ‘John forcibly made the child put on his socks happily.’ [ambiguous as to who was happy] (Matsumoto 1998: 9) In a sense, the productive sase causative in (4b) is the mirror image of the Catalan causative in (3b): in the Japanese construction, a single verb form corresponds to two syntactic predicates; in Catalan, two verb forms correspond to a single syntactic predi- cate. Thus, data show that there is not necessarily a one-to-one correspondence between function and verb form. In addition to the biclausal productive sase causative construction in Japanese, a certain group of transitive verbs form a monoclausal structure when the causative -(s)ase morpheme is added. For example, the causative form of nomu ‘to drink’ (nom-ase-ru) in (6) means ‘to feed/give a drink to’. [Japanese] (6) Hahaoya-wa mother-TOP akachan-ni baby-DAT miruku-o milk-ACC nom-ase-ta. drink-CAUS-PST ‘The mother fed the baby with milk (from a bottle).’ (Matsumoto 1998: 5) This process would be classed as occurring in the lexicon according to Wasow (1977) because it is not fully productive. I refer to this construction as the non-productive sase causative. In contrast to a biclausal productive sase causative construction like (5), the interpre- tation of adverbials is unambiguous in the case of a non-productive sase causative con- struction such as (7). This and other data indicate that non-productive sase causatives are monoclausal. [Japanese] (7) Jon-wa John-TOP sono the netakirino bedridden roojin-ni old.man-DAT sono the kutsushita-o socks-ACC ooyorokobide happily hak-ase-ta. put.on-CAUS-PST ‘John happily put the socks on the feet of the bedridden old man.’ [unambiguously John who was happy] (Matsumoto 1998: 9) However, if the productive sase causative is formed in the syntax, as per Wasow (1977), and the lexicon provides the input for the syntax, Manning et al. (1999) note that data such as (8) are problematic. In this sentence, a lexical process (nominalization via suffix- ation of -kata ‘way’) has applied after the formation of a productive sase causative.
- 48. I. Introduction: Syntax in Linguistics 30 [Japanese] (8) (?kodomo-e-no) child-DAT-GEN hon-no book-GEN yom-ase-kata read-CAUS-way ‘the way to cause (the child) to read a book’ (Manning et al. 1999: 44) Any analysis of the productive sase causative must account for (8), as well as the differ- ent syntactic structures (monoclausal vs. biclausal) of the two sase causative construc- tions. Three different approaches could be adopted: (i) both causatives are formed by a syntactic process, but differ in terms of their syntax (Section 2.3.1); (ii) both are formed by a lexical process, but differ lexically (Section 2.3.2); or (iii) the two processes are fundamentally different because one is lexical and the other is syntactic (Section 2.3.3). Comparing these approaches, each with distinct theoretical consequences, provides an insight into a range of issues concerning the relationship between lexical items and syntax and how they have been addressed. 2.3.1. Syntactic approaches Harley (2008) analyses causativization in purely syntactic terms within the framework of Distributed Morphology (Halle and Marantz 1993; see also Section 3.2), a theory in which there is no lexicon and (syntactic) principles apply to both morphology and syntax. The main semantic feature shared by different types of causatives is defined syntacti- cally: the head CAUSE of a verbal projection is common to all causative constructions. The two Japanese sase causatives are analysed as differing in terms of the structural complexity of their syntax. In the case of the non-productive sase causative, -(s)ase appears in vo , immediately adjacent to a verb root which cannot have a subject of its own. By contrast, in the biclausal productive sase causative, -(s)ase is the head of a verbal CAUSE projection which subcategorizes for a clause with its own subject. The possibility that the same morpheme may attach to a “high” or “low” projection, and that this may correlate in a systematic way with the differences between them, is common to a number of analyses. For example, Travis (2000) distinguishes different types of causatives in Malagasy and Tagalog on the basis of the syntactic position that the causative morpheme occupies. Travis (2000) states that the Event Phrase marks the boundary between l(exical)-syntax and s(yntactic)-syntax; causatives formed in the do- main of l-syntax (causative morpheme occurs below E, the head of the Event Phrase) are non-productive, while those formed in s-syntax (causative morpheme occurs above E) are productive. This approach to the issue of causativization, as well as those outlined by Harley (2008) and Ramchand (2008: 150−192), treats syntax as a generative compo- nent with respect to word formation. However, data from languages other than Japanese challenge a uniform syntactic analysis of causativization cross-linguistically. For example, Horvath and Siloni (2010) argue that a productive causative construction in Hungarian is not biclausal but mono- clausal (see also Pylkkänen 2008: 107). The productive/non-productive distinction thus does not appear to correlate with the proposed difference in syntactic structure. Another issue for a syntactic analysis is that in order to account for the data in (8), -kata nominals must be derived in the syntax, as Harley (2008: 48, fn. 8) notes.
- 49. 3. Syntax and its Interfaces: An Overview 31 2.3.2. Lexicalist approaches Analyses such as Manning et al. (1999), working in the framework of HPSG, and Matsu- moto (1996, 1998), working in LFG, seek to account for the two Japanese sase causa- tives in terms of a lexical distinction. In both cases, the causative verb is derived from the base verb in the lexicon. The monoclausal/biclausal distinction is generally dealt with as a difference in structural complexity relating to the mappings between different types of linguistic structure. For example, Matsumoto (1998: 5−6) captures the fact that productive and non-productive sase causatives in Japanese share some features by proposing that they are both monoclausal at f(unctional)-structure, a level of syntactic structure at which grammatical relations such as subject and object are represented. Non- productive sase causatives involve a single predicate − the verb stem and causative morpheme together have a single argument structure − and are “purely monoclausal” (Matsumoto 1998: 5), i.e. at the level of a(rgument)-structure as well as at f-structure (9). (9) Non-productive sase causative, e.g. nom-ase-ta ‘fed/gave a drink to’ in (6) a-structure: feed < Agent, Recipient, Patient > f-structure: ‘feed < SBJ OBJ OBJpatient > (Matsumoto 1998: 6) By contrast, productive sase causatives are biclausal at the level of a(rgument)-structure. The causative morpheme introduces a separate predicate CAUSE with its own argument structure. The patient of CAUSE and the agent of the embedded predicate are fused to- gether in a-structure, as indicated by the dashed line in (10), the result being that two thematic roles are shared by a single “fused” argument (Alsina 1992). (10) Productive sase causative, e.g. nom-ase-ta ‘caused to drink’ n o i s u f a-structure: CAUSE < Agent, Patient, Subevent > drink< Agent, Patient > f-structure: ‘cause-to-drink’ < SBJ OBJ OBJpatient >’ (Matsumoto 1998: 5) Thus two different a-structures may map to the same f-structure, cf. (9) and (10), and the observed monoclausal/biclausal contrast is analysed as a reflex of argument structure. The fact that the productive sase causative can be nominalized by suffixation of -kata, as exemplified in (7), supports this lexicalist analysis because both are proposed to be formed in the lexicon. Accounting for the possibility of disjunction of sub-word constituents under the scope of causation (Kuroda 2003: 455−458) and the causativization of raising verbs (Horvath and Siloni 2010: 168) have been identified as challenges to the lexicalist analysis of the productive sase causative.
- 50. I. Introduction: Syntax in Linguistics 32 2.3.3. Mixed approaches A third possible approach to the two sase causatives would be to assume that they are inherently different, one being formed in the lexicon and the other being formed in the syntax. Baker (1988) represents a framework broadly consistent with the Weak Lexicalist Hypothesis within which to explore this possibility. According to this approach, the underlying structure of both causative constructions is the same because according to the Uniformity of Theta Assignment Hypothesis (UTAH; Baker 1988: 46), there is a one-to-one correspondence between syntactic positions and thematic roles as assigned by a specific type of predicate. The productive sase causative is formed in the syntax: the base verb is head of a lower VP that moves to SpecC' before its head is adjoined, via Incorporation (Section 3.2), to -(s)ase which is the head of a higher VP. The non- productive sase causative, on the other hand, is formed in the lexicon and represents a single verb in the input to the syntax. Harley (2008: 33) criticizes this type of Incorpora- tion analysis for failing to account for the systematic relationship which exists between the productive and non-productive sase causatives. For discussion of other problems with an Incorporation analysis of causatives, see Alsina (1992). Horvath and Siloni (2010) propose an analysis which does not seek to define causa- tive formation in purely syntactic or lexical terms either, defining the difference in terms of the setting of the Lexicon-Syntax Parameter (Reinhart and Siloni 2005), which states that valency-changing operations including causativization apply either in the lexicon or in the syntax in a language. In Japanese, causativization is claimed to apply in the syntax; non-productive sase causatives are hypothesised to be underived lexical items or relics and therefore irrelevant to the language’s parameter setting. It remains to be clarified how data such as (8), which shows that a syntactic causative can be the input to nominali- zation, is dealt with under such an analysis. 2.4. Conclusion The causative construction has been the focus of much research and its study continues to raise issues of importance. This and other data concerning the realization of arguments are key to understanding and defining the relationship between lexical items and syntax. A closely related issue is the interface between syntax and morphology. 3. Syntax and morphology 3.1. Introduction Different theoretical approaches account for the part which syntax plays in the formation of complex words in distinct ways. Defining the nature of morphology is an important consideration in itself. A key proposal that cuts across lexical and syntactic approaches to the syntax−morphology interface is the Separationist Hypothesis (see Beard 1995, amongst others), which advocates divorcing the meaning or morphological function of
- 51. 3. Syntax and its Interfaces: An Overview 33 a morpheme from its form. For example, an abstract L(exical)-derivation rule of plurali- zation changes the features of an English singular count noun thus: [−Plural, +Singular] → [+Plural, −Singular] (Beard 1995: 160). The form of the plural is determined in a separate morphological spelling (MS) component. MS-rules include suffixation, e.g. dog−dogs, and umlaut, e.g. man−men, which accounts for the multiple exponence of the morphosyntactic category ‘plural’ in English. This approach is often adopted in the analysis of inflection, and has also been extended to derivation by some researchers (see Beard 1998). For those who do not subscribe to the view that syntax subsumes morphology (Sec- tion 3.2) and instead adhere to a strong or weak version of the Lexicalist Hypothesis (Section 3.3), the challenge is not only to identify those primitives and operations which are uniquely morphological in their nature and define them, but also to characterize the interface between syntax and morphology (Section 3.4). Key data are examples of phenomena which exhibit a combination of properties that appear to be neither wholly morphological nor wholly syntactic. These include cliticization (e.g. Anderson 2005; Halpern 1995; Zwicky 1994), incorporation (e.g. Baker 1988; Mithun 1984) and inflec- tion (e.g. Anderson 1992; Steele 1995; Stump 2001). Indeed, approaches to morphology can also be classified according to the relations they posit between the morphology and morphosyntactic properties of an inflected word (lexical vs. inferential and incremental vs. realizational). See Stump (2001). The proposed analyses of incorporation, undertaken in a variety of frameworks, pro- vide a useful insight into some major issues. 3.2. Syntax subsumes morphology Noun incorporation is a productive process found in a number of languages including Classical Nahuatl, a North American language. The two sentences in (11) are very simi- lar in meaning and include the same morphemes, but in (11b) the direct object naca ‘meat’ is part of the verb qua ‘eat’, the only word in the sentence. [Classical Nahuatl] (11) a. ni-c-qua 1SG.SBJ-3SG.OBJ-eat in DET naca-tl. meat-ABS ‘I eat (the) meat.’ b. ni-naca-qua. 1SG.SBJ-meat-eat ‘I eat meat.’ (Iturrioz Leza 2001: 715) The process involved in forming (11b) is productive and may be the usual way to convey a particular meaning. This is reminiscent of a syntactic process, though the result of this noun incorporation is a single word. As Mithun (1984: 847) states, such noun incorpora- tion is “perhaps the most nearly syntactic of all morphological processes”. Baker’s (1988) influential approach to noun incorporation is an example of how word formation can be analysed as a syntactic process: syntax is argued to manipulate mor- phemes to form words in cases such as (11b). Specifically, complex words are formed
- 52. I. Introduction: Syntax in Linguistics 34 by head movement, a syntactic operation restricted by the Head Movement Constraint (Travis 1984: 131), or possibly the Empty Category Principle (Chomsky 1981: 250) or Relativized Minimality (Rizzi 1990), as shown in (12). This type of derivational rule is known as Incorporation. (12) Noun incorporation, a way of forming a complex word, is thus analysed in terms of established principles of derivational syntax. In this way, syntactic operations are pro- posed to be responsible for generating structure below as well as above word level. Other syntactic analyses of noun incorporation are presented in van Geenhoven (1998), Koopman and Szabolcsi (2000) and Massam (2001). Baker (2009) compares his head- movement analysis to these approaches in the context of the Minimalist Program (Chom- sky 1995). Baker (1988) also uses Incorporation to account for other morphosyntactic phenom- ena, including passivization and applicativization, two processes in which affixation may accompany a change in argument structure. Baker (1985: 375) further claims that gram- mar is constrained by the Mirror Principle: “morphological derivations must directly reflect syntactic derivations (and vice versa)”. The order of affixes relative to each other and the base is thus predicted to reflect the order in which the relevant syntactic opera- tions applied in the derivation of a complex word. If syntax subsumes morphology en- tirely, the Mirror Principle should always constrain word-internal structure. However, while it represents a strong tendency, the Mirror Principle does not appear to be inviola- ble, as Hyman et al. (2009) show. For example, when an applicative is passivized, the order of syntactic operations is applicative then passive, so the applicative morpheme should be affixed closer to the base than the passive one according to the Mirror Princi- ple. This is the case in the example from the Bantu language Ndebele in (13a). However, when a passive is subject to applicativization in Ndebele, and hence the order in which the syntactic operations apply is reversed, as in (13b), the same morphemes appear in the same order as in (13a). (13) a. Passivized Applicative [Ndebele] abantwana children b-a-phek-el-w-a they-PST-cook-APPL-PASS-FS ukudla food ‘The children were cooked food’
- 53. 3. Syntax and its Interfaces: An Overview 35 b. Applicativized Passive ukudla food kw-a-phek-el-w-a it-PST-cook-APPL-PASS-FS abantwana children ‘The food was cooked (for) the children’ (Hyman et al. 2009: 298) Thus, morphology does not reflect syntactic derivation in (13b). This undermines one of the strongest types of argument for a purely syntactic analysis of such productive com- plex word formation because in these cases the processes involved in syntactic and morphological structure building do not appear to be exactly the same. Baker himself does not advocate a syntactic analysis of all complex word formation (Baker 1993: 588) and Incorporation could not be straightforwardly extended to cover morphological processes which are non-concatenative. Others go further though, assum- ing that there is no separate part of the grammar which generates morphological struc- ture. Lack of isomorphism is an issue which syntactic approaches to morphology must address. Lieber (1992) seeks to capture complex word formation in terms of independent principles of syntax and phonology. Under this approach, syntactic rules based on revised principles of X-bar theory may apply to form a word from morphemes contained within the lexicon. In this linear model, the sole interaction between morphology and syntax is insertion of lexical items into syntax, at which point structure above and below word level is built. Lieber’s (1992) analysis rests on definitions of head and non-heads in- tended to apply both above and below word level, and the proposal that direction of headedness is a parameter set once for a language and thus that the order of syntactic units and morphemes will be identical, all things being equal. In support of this claim, Lieber (1992) notes that, for example, English is largely a head-final language with right-headed morphology (i.e. suffixation, Lieber 1992: 49−64, the head of a word being the part that determines its category) while Tagalog is largely a head-initial language with left-headed morphology (i.e. prefixation; Lieber 1992: 40−49). There are languages such as Warlpiri (Simpson 1991) though which exhibit freedom of word order and a rigid morpheme order that would be difficult to account for in terms of a headedness pa- rameter. A purely syntactic analysis must account for complex word formation in terms of independently motivated syntactic principles. However, Lieber (1992) finds it necessary to adapt X-bar principles to account for word-internal structure, indicating that it does differ in some respects from syntactic structure (see Borer 1998). Distributed Morphology (Halle and Marantz 1993; Harley and Noyer 1999; Marantz 1997) is another syntactic approach to morphology, according to which “all derivation of complex objects is syntactic” (Embick and Noyer 2007: 290), and therefore syntactic and morphological structure are fundamentally the same. Distributed Morphology is an attempt to incorporate Separationism in a model which in effect eliminates morphology altogether. Instances in which there is a lack of isomorphism between the two are ac- counted for by differences in the complexity of the derivation involved or the application of readjustment rules post-syntactically (Embick and Noyer 2007: 305−310). Distributed Morphology, in common with other syntactic treatments of morphological phenomena, is argued to be preferable on the grounds of parsimony: by having the syntactic component subsume some or all complex word-formation, it is possible to
- 54. I. Introduction: Syntax in Linguistics 36 reduce or eliminate the need for morphology as a separate part of the grammar. However, if it is necessary to posit new or revised operations which are not motivated on independ- ent syntactic grounds and may never apply in the syntax, or to add powerful post- syntactic stipulations in order to account for morphological processes, the parsimony argument in favour of a syntactic analysis is undermined. 3.3. Lexicalist approaches Contrasting with syntactic approaches to morphology are those such as Aronoff (1994) which maintain that the formation and structure of complex words is distinct from that of phrases and hence morphology is a separate component of the grammar. The Strong Lexicalist approach (e.g. Halle 1973), in which specific rules of word formation apply in one component of the grammar and feed into a separate syntactic component, has generally been rejected on the basis that inflection exhibits features consistent with syntactic principles. This is the weak lexicalist position found in the work of Aronoff (1976), whose word-formation rules (WFRs) account only for derivational morphology. According to the Weak Lexicalist Hypothesis, derivational processes apply in the lexicon, whereas “inflectional morphology is what is relevant to the syntax” (An- derson 1982: 587). Thus in English, nominalization by addition of -ion (derivational morphology) is analysed as being inherently different from affixation of -s to a verb stem indicating third person singular subject agreement (inflectional morphology) as the latter is integrated with the syntax. Inflectional processes such as agreement and case marking are characterized as a set of adjustment rules which serve to harmonize the output of the syntactic and morphological components in a Weak Lexicalist framework. Kiparsky’s (1982) Strong Lexicalist analysis, by contrast, relates the distinction between derivation and inflection to the cyclic ordering of rules of affixation relative to each other within the lexicon. Morphology is thus kept separate from syntax in this Lexical Phonology/Morphology model, but the price is to situate morphological and phonologi- cal rules in the lexicon. The process of incorporation is open to a lexicalist analysis, as work by Di Sciullo and Williams (1987), Mithun (1984) and Rosen (1989) shows. Under such an analysis, noun incorporation is the result of a verb combining with one of its arguments, similar to an English verbal compound such as horse riding. This is a lexical rather than a syntactic process. (Not all noun incorporation appears to be the same in this respect. Rosen 1989 identifies a type of incorporation, Classifier NI, which involves incorpora- tion of a noun that is not selected as an argument by the verb.) Anderson (2000: 135), concurring with Baker, notes that “the syntactic and lexical theories of Noun Incorporation are ‘tied’ in that each can be said to account for roughly the same range of phenomena the other can”, but goes on to discuss data which are problematic for a purely syntactic analysis. Specifically, Anderson (2000) discusses in- stances of morphosyntactic agreement in noun incorporation constructions. Baker (1996: 307−329) claims that when the syntactic operation of Incorporation has applied, a trace is left which occupies the position in which the incorporated noun was base generated, as in (12). There should not be marking on the verb of agreement with this trace because it does not possess any of the features of the moved element. Anderson
- 55. Random documents with unrelated content Scribd suggests to you:
- 56. mobilize; he was known to them as "Fred Hansen." If he wanted a naval reservist he called on Boy-Ed; if an army reservist was required von Papen sent him to "Hansen." Boy-Ed gave him data on ship sailings, von Papen on munitions plants, Koenig on secret service. His first task was to buy supplies and ship them to Germany. He boasted that there was no such thing as a British blockade. Using his pseudonyms of Gibbon and Hansen he made large purchases and with the aid of Captain Gustave Steinberg, a naval reservist, he chartered ships and dispatched them under false manifests to Italy and Norway, where their cargoes could be readily smuggled into Germany. Through Steinberg he importuned a chemist, Dr. Walter T. Scheele, to soak fertilizer in lubricating oil for shipment to the Fatherland, where the valuable oil could be easily extracted. Through the same intermediary von Rintelen gave Dr. Scheele $20,000 to ship a cargo of munitions under a false manifest as "farm implements"; Dr. Scheele kept the $20,000 and actually shipped a cargo of farm machinery. Rintelen's next venture attracted some unpleasant attention. The United States Government had condemned some 350,000 Krag- Joergensen rifles, which it refused to sell to any of the belligerents. Rintelen cast a fond eye in their direction. President Wilson had told a banker: "You will get those rifles only over my dead body." Rintelen heard, however, that by bribing certain officials he could obtain the guns, so he sent out agents to learn what they would cost, and found a man who said he could buy them for $17,826,000, part of which was to be used for effective bribery. "So close am I to the President," said the intermediary, "that two days after I deposit the money in the bank you can dandle his grandchild on your knee!" But just when the negotiations were growing bright, Rintelen was told that the man who proposed to sell him the rifles was a secret agent from another government. A certain "Dr. Alfred Meyer" was known to have been groping for those rifles, and the newspapers and government officials became suddenly interested in his real
- 57. identity. A dowdy woman's implication reached a reporter's ears; presently the newspapers burst out in the "discovery" that "Dr. Alfred Meyer" was none other than Dr. Meyer-Gerhardt, a German Red Cross envoy then in the United States. Like the popping of a machine gun, "correct versions of the facts" were published: "Dr. Meyer-Gerhardt denied vigorously that he was 'Dr. Alfred Meyer,'" then "'Dr. Alfred Meyer' was known to have left the United States on the same ship with Dr. Meyer-Gerhardt," then "an American citizen came forward anonymously and said that he had posed as 'Dr. Alfred Meyer' in order to test the good faith of the Government." This last announcement may have been true. It was made to a New York Sun reporter by a German, Karl Schimmel, who professed his allegiance to the United States, and by the "American citizen" who said he had posed as "Dr. Alfred Meyer." It may have been made to shield Rintelen himself, for the "American citizen" was an employe of a German newspaper in New York, a friend of Rintelen's, a friend of Schimmel's and Schimmel himself was in von Rintelen's pay. Let a pack of reporters loose on a half dozen tangents and they will probably scratch the truth. A Tribune man heard a whisper of the facts and set out on a hunt for "two Germans, Meyer and Hansen, who have been acting funny." He frightened the personnel right out of the office of the E. V. Gibbon Company. Captain Steinberg fled to Germany with a trunkful of reports on the necessity of concerted action to stop the shipment of munitions to the Allies, and Rintelen migrated to an office in the Woolworth Building. Some one heard of his activities there and he was evicted, taking final refuge in the Liberty Tower, in the office of Andrew M. Meloy, who had been in Germany to interest the German government in a scheme similar to Rintelen's own. In Meloy's office Rintelen posed as "E. V. Gates"— preserving the shadow of his identity as "Emil V. Gasche." So effective was his disappearance from the public view, that he was reported to have gone abroad as a secretary, and he sat in the tower and chuckled, and sent messages by wireless to Berlin through Sayville, and cablegrams to Berlin through England and Holland, and
- 58. enjoyed all the sensations of a man attending a triple funeral in his honor. "Meyer," "Hansen" and "Gasche" were all dead, and yet, here was Rintelen! Although his sojourn in New York covered a period which was the peak of the curve of German atrocities in the United States, Rintelen was a fifth wheel. No man came to America to accomplish more, and no man accomplished less. No German agent had his boldness of project, and no German executive met a more ignominious fate. Whatever he touched with his golden wand turned to dross. He was hoodwinked here and there by his own agents, and frustrated by the vigilance of the Allied and the United States governments. He has been introduced here because of his connection with subsequent events, and yet this picturesque figure played the major part in not one successful venture. Four months he passed in America, until it became too small for him. In August the capture of Dr. Albert's portfolio and the publication of certain of its contents frightened Rintelen, and he applied for a passport as "Edward V. Gates, an American citizen of Millersville, Pa.," but he did not dare claim it. Though he had bought tickets under the alias, and had had drafts made payable in that name, he did not occupy the "Gates" cabin on the Noordam, but at the last minute engaged passage under the renascent name of "Emil V. Gasche," the harmless Swiss. He eluded the Federal agents, and sailed safely to Falmouth, England, where, after a search of the ship, and an excellent attempt to bluff it through, he finally surrendered to the British authorities as a prisoner-of-war. Meloy and his secretary were captured with him. Rintelen was returned to the United States in 1916. He was convicted in 1917 and 1918 on successive charges of conspiracy to violate the Sherman Anti-Trust law, to obtain a fraudulent passport, and to destroy merchant ships—which combined to sentence him to a year in the Tombs and nine years in a Federal prison.
- 60. CHAPTER XI SHIP BOMBS Mobilizing destroying agents—The plotters in Hoboken—Von Kleist's arrest and confession—The Kirk Oswald trial—Further explosions—The Arabic—Robert Fay—His arrest—The ship plots decrease. The reader will recall a circular quoted in Chapter VIII, and issued November 18, 1914, from German Naval Headquarters, mobilizing all destroying agents in harbors overseas. On January 3, 1915, there was an explosion on board the munitions ship Orton, lying in Erie Basin, a part of New York harbor. On February 6 a bomb was found in the cargo of the Hannington. On February 27 the Carlton caught fire at sea. On April 20 two bombs were found in the cargo of the Lord Erne. One week later the same discovery was made in the hold of the Devon City. All of which accounts for the following charge: "George D. Barnitz, being duly sworn, deposes and says ... on information and belief that on the first day of January, 1915, and on every day thereafter down to and including the 13th day of April, 1916, the defendants Walter T. Scheele, Charles von Kleist, Otto Wolpert, Ernst Becker, (Charles) Karbade, the first name Charles being fictitious, the true first name of defendant being unknown, (Frederick) Praedel ... (Wilhelm) Paradis ... Eno Bode and Carl Schmidt ... did unlawfully, feloniously and corruptly conspire ... to manufacture bombs filled with chemicals and explosives and to place said bombs ... upon vessels belonging to others and laden with moneys, goods and merchandise...."
- 61. Ninety-one German ships were confined to American harbors by the activities of the British fleet, ranging from the Neptun, of 197 tons, in San Francisco Bay, to the Vaterland, of 54,000 tons, the largest vessel on the seven seas, tied up to accrue barnacles at her Hoboken pier, and later, as the Leviathan, to transport American troops to France. Every one of the ninety-one ships was a nest of German agents. Only a moderate watch was kept on their crews, and there were many restless men among them. Every man aboard was liable to command from Captain Boy-Ed, for the German merchant marine was part of the formal naval organization. The interned sailors found shortly that they could be of distinct service to their country without stirring from their ships. Not far from the North German Lloyd piers in Hoboken lived Captain Charles von Kleist, 67 years old, a chemist and former German army officer. One day there came to him one who spoke the German tongue and who said he came from Wolf von Igel, in von Papen's office. Those were good credentials, especially since the gentleman was inquiring on von Igel's behalf whether Kleist needed any money in the work he was doing. The polite caller returned a few days later with another man, who spoke no German. Von Kleist asked whether he was also from the Fatherland, and was told no, but "we have to use all kinds of people in our business—that's how we fool these Yankees!" Von Kleist laughed heartily, and wagged his head, and went out in the garden and dug up a bomb-case and showed the visitors how it had been made. The visitors were Detectives Barth and Barnitz. They assured Kleist that von Igel wanted to know precisely what he and his associates were doing, so no money might be paid to the wrong parties. The aged captain wrote out a memorandum of his activities, which he signed, and the detectives proposed a trip to Coney Island as an evidence of good faith, so the three had a pleasant afternoon at the Hotel Shelburne, and the officers then suggested: "Let's go up and see the chief." "Chief" to von Kleist
- 62. meant von Igel; he agreed, and was taken gently into the arms of the chief of detectives. He implicated, as he sat there answering questions, Captain Eno Bode, pier superintendent of the Hamburg-American Line, Captain Otto Wolpert, pier superintendent of the Atlas Line, and Ernst Becker, an electrician on the North German Lloyd liner Friedrich der Grosse, tied up at Hoboken. The other conspirators were induced to come to New York, and were arrested at once. Bode and Wolpert, powerful bullies of Paul Koenig's own stamp, proved defiant in the extreme. Becker, knowing no word of English, was pathetically courteous and ready to answer. But it remained for von Kleist to supply the narrative. Becker, working on the sunny deck of the Friedrich der Grosse, had made numerous bomb cases, rolling sheet lead into a cylinder, and inserting in the tube a cup-shaped aluminum partition. These containers he turned over to Dr. Walter Scheele at his "New Jersey Agricultural Company," where he filled one compartment with nitroglycerine, the other with sulphuric acid. Scheele supplied the mechanics with sheet lead for the purpose. The bombs were then sealed and packed in sand for distribution to various German gathering places, such as, for example, the Turn Verein in the Brooklyn Labor Lyceum. Wolpert appeared there at a meeting one night and berated the Germans present for talking too much and acting too little; he wanted results, he said. Eugene Reister, the proprietor of the place, said that shortly afterward Walter Uhde and one Klein (who died before the police reached him) had taken away a bundle of bombs from the Turn Verein and had placed them on the Lusitania, just before her last voyage, and added that Klein, when he heard of the destruction of the ship, expressed regret that he had done it. Karl Schimmel—the same who had negotiated for the Krag rifles—said later to Reister: "I really put bombs on that boat, but I don't believe that fellow Klein ever did." Following Kleist's information, agents of the Department of Justice and New York police inspected the Friedrich der Grosse, and found
- 63. quantities of chlorate of potash and other chemicals. They brought back with them also Garbode (mentioned in the charge as "Karbade"), Paradis and Praedel, fourth engineers on the ship, who had assisted in making the bombs, and Carl Schmidt, the chief engineer. All of the group were implicated in the plot to the complete satisfaction of a jury which concluded their cases in May, 1917, by convicting them of "conspiracy to destroy ships through the use of fire bombs placed thereon." Kleist and Schmidt received sentences of two years each in Atlanta Penitentiary and were each fined $5,000; Becker, Karbade, Praedel and Paradis were fined $500 apiece and sentenced to six months in prison. Dr. Scheele fled from justice, and was arrested in March, 1918, in Havana. A liberal supply of vicious chemicals and explosives discovered in his "New Jersey Agricultural Company" implicated him thoroughly, if the evidence given by his fellows had not already done so. When he was finally captured he faced two federal indictments: one with Steinberg and von Igel for smuggling lubricating oil out of the country as fertilizer, under false customs manifests; the other the somewhat more criminal charge of bombing. On April 29, 1915, the Cressington caught fire at sea. Three days later, in the hold of the Kirk Oswald, a sailor found a bomb tucked away in a hiding place where its later explosion would have started a serious fire. So it came about that when the four lesser conspirators of the fire-bomb plot had served their six months' sentences, they were at once rearrested on the specific charge of having actually planted that bomb in the Kirk Oswald. The burly dock captains, Bode and Wolpert, who had blustered their innocence in the previous trial, and had succeeded in securing heavy bail from the Hamburg- American Line pending separate trials for themselves, were nipped this time with evidence which let none slip through. Rintelen was haled from his cell to answer to his part in the Kirk Oswald affair, and the jury, in January, 1918, declared the nine plotters "guilty as charged" and Judge Howe sentenced them to long terms in prison. Rintelen, alone of the group, as they sat in court, had an air of anything but wretched fanatic querulousness. He followed the
- 64. proceedings closely, and once took the trial into his own hands in a flash of temper when the State kept referring to the loss of the Lusitania. It went hard with the nobleman to be herded into a common American court with a riff-raff of hireling crooks and treated with impartial justice. In Germany it never could have happened! If those trials had occurred in May, 1915, the history of the transport of arms and shells would not have been marred by such entries as these: May 8—Bankdale; two bombs found in cargo. May 13—Samland; afire at sea. May 21—Anglo-Saxon; bomb found aboard. June 2—Strathway; afire at sea. July 4—Minnehaha; bomb exploded at sea. (The magnetos.) July 13—Touraine; afire at sea. July 14—Lord Downshire; afire. July 20—Knutford; afire in hold. July 24—Craigside; five fires in hold. July 27—Arabic; two bombs found aboard. Aug. 9—Asuncion de Larriñaga; afire at sea. Aug. 13—Williston; bombs in cargo. Aug. 27—Lighter Dixie; fire while loading. On August 31 the White Star liner Arabic, nineteen hours out of Liverpool was torpedoed by a German submarine and sank in eleven minutes, taking 39 lives, of which two were American. Germany, on September 9, declared that the U-boat commander attacked the Arabic without warning, contrary to his instructions, but only after he
- 65. was convinced that the liner was trying to ram him; the Imperial Government expressed regret for the loss of American lives, but disclaimed any liability for indemnity, and suggested arbitration. On October 5, however, the government in Berlin had changed its tune to the extent of issuing a note expressing regret for having sunk the ship, disavowing the act of the submarine commander, and assuring the United States that new orders to submarines were so strict that a recurrence of any such action was "considered out of the question." If the cargoes could be fired at sea, no submarine issue need be raised. And so fires and bombs continued to be discovered on ships just as consistently as before. The log, resumed, runs thus: Sept. 1—Rotterdam; fire at sea. Sept. 7—Santa Anna; fire at sea. Sept. 29—San Guglielmo; dynamite found on pier. Now von Rintelen's handiwork was revealed in the adventures of Robert Fay, or "Fae," as he was known in the Fatherland. In spite of the imaginative quality of the enterprise, and the additional guilt which it heaped upon the executives of the spy system, it was not successful. There were vibrant moments, though, when only the mobilization of police from two states and special agents from the Secret Service and Department of Justice averted what would have developed into a profitable method of destroying ships. Lieutenant Robert Fay was born in Cologne, where he lived until 1902. In that year he migrated to Canada, where he worked on a farm, and later to Chicago, where he was employed as a bookkeeper until 1905. He then returned to Germany for his military service, and went to work again in Cologne, in the office of Thomas Cook & Sons. After a period in a Mannheim machine shop he went home and devoted himself to certain mechanical inventions, and was at work upon them when he was called out for war service on August 1, 1914.
- 66. His regiment went into the trenches, and the lieutenant had some success in dynamiting a French position. Conniving with a superior officer, he deserted his command, and was sent to America by a German reputed to be the head of the secret service, one Jonnersen. Jonnersen gave Fay 20,000 marks for expenses in carrying out a plan to stop shipments of munitions from America, and Fay arrived in New York April 23, 1915, on the Rotterdam. Dr. Herbert Kienzle, a clock-maker, of 309 West 86th Street, had written to his father in Germany bitterly assailing the United States for shipping munitions, and enclosed in his letters information of certain American firms, such as Browne & Sharp, of Providence, and the Chalmers Motor Car Company, of Detroit, who were reputed to be manufacturing them. These letters had been turned over to Jonnersen, who showed them to Fay as suggestions. Upon his arrival in New York, then, Fay called on Kienzle, who, though he was friendly enough, was reluctant to know of the details Fay had planned. Dr. Kienzle introduced Fay to von Papen, and later to Max Breitung, from whom he purchased a quantity of potassium chlorate. The deserter found his brother-in-law, Walter Scholz, working as a gardener on an estate near Waterford, Connecticut, and brought him to New York on a salary of $25 a week. The two crossed the Hudson to Weehawken, N. J., and set to work to make bombs. Fay had a theory that a bomb might be attached to the rudder of a ship, and so set as to explode when the rudder, swinging to port, wound a ratchet inside the device which would release a hammer upon a percussion cap. Their plan was to have the parts manufactured at machine shops, assemble and fill them themselves, and then steal up the waterfront in the small hours and attach the infernal machines to outward bound vessels. Fay even counted on disarming the police boats before setting out. It took the two some three months to get the parts made and properly adjusted. Meanwhile they employed their spare hours in cruising about the harbor in a motor-boat. A machinist in West 42nd Street, New York, made the zinc tank which they used as a model,
- 67. and the two conspirators shortly opened a garage in Weehawken where they could duplicate the bomb cases unmolested. There came a time when the devices were satisfactory, and Fay actually attached one to the rudder of a ship to make sure that his adjustments were correct. The next move was to obtain explosives. Fay's prejudice against bombs placed in a ship's hold was that they rarely succeeded in sinking the craft; seventy or eighty pounds of high explosive detonated at the stern of a vessel, however, would blow the rudder away and not only cripple the ship but would probably burst a hole in the stern, mangle the screw, and split the shaft. Captain Tunney, of the Bomb Squad, heard in October that two Germans were trying to buy picric acid from a man who stopped at the Hotel Breslin, and who called himself Paul Seib and Karl F. Oppegaarde, as the occasion demanded. Tunney's men located the two Germans, and some days later learned that they had placed an order for fifty-two pounds of TNT, to be delivered at the Weehawken garage. The delivery was intercepted, a similar but harmless substance substituted for the explosive, and two detective-truckmen took the package away on their truck to deliver it to Fay and Scholz. While they were in New Jersey, Detectives Coy, Sterrett and Walsh found Fay at the Breslin, and followed him back to Weehawken. As he left the garage in the evening in his automobile, the automobile of Police Commissioner Woods followed at a discreet distance. Up the Palisades the two cars paraded, until in a grove near Grantwood, Fay and Scholz got out of their car and disappeared into the woods with a lantern. After a time they reappeared, and returned to the garage, the police following. Next morning Chief Flynn was called into the hunt—the morning of Saturday, October 23—and he assigned two special agents to the case. The police department directed two detectives to watch the woods at Grantwood where the conspirators had gone the night before. Detectives Murphy and Fennelly, each equipped with linemen's climbers, arrived at the wood-road about noon, and spent
- 68. the next eleven hours in the branches of a great oak tree which commanded the road. The perch was high and the night wind chilly, but the watchers were rewarded at last by the twin searchlights of an approaching car. Out of it stepped Fay and Scholz. The men in the branches saw by the light of the lantern which Scholz carried that Fay placed a package underneath a distant tree, walked to a safe distance, exploded a percussion cap, watched the tree topple over and went away, apparently satisfied with the power of his explosives.
- 69. Robert Fay, who made bombs with which he hoped to cripple the shipment of munitions to Europe Meanwhile other detectives were watching the rooming house at Union Hill where Fay and Scholz lived, and they saw the two come in about 4 o'clock in the morning. Scholz had very little sleep, for there was a ship leaving next day for Liverpool. He left the house at 7 A. M. and went to the garage. Thereupon three detectives returned to the
- 70. great oak tree at Grantwood. About noon Fay and his brother-in-law drove up, and unlocking the door of a rude hut in the wood, took out a bag, from which they poured a few grains of powder on the surface of a rock. Fay struck the rock with a hammer; a loud report followed, and the hammer broke in his hand. A moment later he heard a twig snap behind him. He turned, and saw a small army of detectives with drawn revolvers closing in on him. Fay protested and pleaded, and offered to bribe the detectives for his freedom, but he was locked up with Scholz. The two had stored in a warehouse several cases containing their completed bomb mechanisms; the police confiscated from their various caches five new bombs, 25 pounds of TNT, 25 sticks of dynamite, 150 pounds of chlorate of potash, two hundred bomb cylinders, 400 percussion caps, one motor-boat, one chart of New York harbor showing all its fortifications and piers, one foreign automobile, two German automatic pistols and a long knife—a considerable arsenal. Their confessions caused the arrest of Paul Daeche, who had furnished them with explosives, Dr. Kienzle, Breitung, and Engelbert Bronkhorst. Fay received a sentence of eight years in the penitentiary, but after America went to war, Atlanta became too confining for his adventurous spirit, and he escaped the prison, and is believed to have crossed the Mexican border to safety. Scholz was sentenced to four years, and Daeche to three. Kienzle, Breitung and Bronkhorst were not tried, their apparent ignorance of Fay's designs outweighing in the jury's mind their obvious German sympathies. Kienzle, upon the declaration of war of April 6, 1917, became an enemy alien, and was interned. So Lieutenant Fay never qualified in active service as a destroying agent. Yet he was profligate in his intentions. He offered two men $500,000 if they could intrigue among the shippers in order that a ship laden with copper for England might wander from the path of convoy into German hands, and he even entertained the fantastic hope, with his chart and his motor-boat and his bombs, of stealing out of the harbor to the cordon of British cruisers who hung outside
- 71. the three-mile limit and attaching his bombs to their rudders, that the German merchantmen might escape into the open sea. On October 26 the Rio Lages caught fire at sea; fire broke out in the hold of the Euterpe on November 3; three days later there was fire aboard the Rochambeau at sea; the next day an explosion occurred aboard the Ancona. And so the list runs on: Dec. 4—Tynningham, two fires on ship. Dec. 24—Alston, dynamite found in cargo. Dec. 26—Inchmoor, fire in hold. 1916 Jan. 19—Sygna, fire at sea. Jan. 19—Ryndam, bomb explosion at sea. Jan. 22—Rosebank, two bombs in cargo. Feb. 16—Dalton, fire at sea. Feb. 21—Tennyson, bomb explosion at sea. Feb. 26—Livingston Court, fire in Gravesend Bay. April saw the round-up of the group who had been working under the Hamburg-American captains, and although numerous fires occurred during May, 1916, in almost every case they were traced to natural accidents. The number mounted more slowly as the year advanced. With the entrance of America into the war, and the tightening of the police cordon along the waterfront, the chance of planting bombs was still further reduced, but waterfront fires kept recurring, and until the day of ultimate judgment in Berlin, when each of Germany's arsonists in America comes to claim his reward, none will know the total of loss at their hands. It was enormous in the damage it inflicted upon cargo, but it is improbable that it had
- 72. any perceptible effect upon the whole export of shells for Flanders and France.
- 73. CHAPTER XII LABOR David Lamar—Labor's National Peace Council—The embargo conference—The attempted longshoremen's strike—Dr. Dumba's recall. Labor produced munitions. The hands of labor could be frightened away from work by explosions, their handiwork could be bombed on the railways, the wharves, the lighters, and the ships, but a surer method than either of those was the perversion of the hearts of labor. So thought Count von Bernstorff and Dr. Albert, who dealt in men. So thought Berlin—the General Staff sent this message to America: "January 26—For Military Attaché. You can obtain particulars as to persons suitable for carrying on sabotage in the United States and Canada from the following persons: (1) Joseph McGarrity, Philadelphia; (2) John P. Keating, Michigan Avenue, Chicago; (3) Jeremiah O'Leary, 16 Park Row, New York. "One and two are absolutely reliable and discreet. Three is reliable, but not always discreet. These persons were indicated by Sir Roger Casement. In the United States sabotage can be carried out on every kind of factory for supplying munitions of war." (Signed) "Representative of General Staff."[3] So too thought von Rintelen, who hired men—usually the wrong ones.
- 74. Full of his project, he cast about for an intermediary. No sly chemist or muscular wharf-rat would do for this delicate task of anesthetizing men with the gas of German propaganda while it tied their hands and amputated their centres of right and wrong; the candidate must be a man of affairs, intimate with the chiefs of labor, skillful in execution, and the abler the better. Von Rintelen would pay handsomely for the right man. Whereupon David Lamar, the "Wolf of Wall Street," appeared on the scene and applied for the job—an entrance auspicious for the United States, for the newcomer's philosophy (if one could judge from his previous career) was "Me First." In an attempt to defraud J. P. Morgan & Co., and the United States Steel Corporation Lamar had once impersonated Representative A. Mitchell Palmer in certain telephone interviews. (Palmer became custodian of alien property after the United States entered the war.) He was convicted and sentenced to two years' imprisonment in Atlanta Penitentiary. He appealed the case, and while he was out on bail pending the appeal, he fell in with Rintelen. In April, 1915, a New Yorker who dealt in publicity was introduced to Rintelen, or "Hansen," by Dr. Schimmel. Rintelen offered the publicity man $25,000 to conduct a campaign of propaganda for more friendly relations with Germany, to offset the commercial power Great Britain bade fair to have at the end of the war, and assured him that he would go to any extreme to prevent shipments of munitions to the Allies. The war, he said, would be decided not in Europe but in America. There must be strikes in the munitions factories. When the publicity man heard also that Rintelen was trying to stir up trouble with Mexico, he wrote on May 13 to Joseph Tumulty, President Wilson's secretary, informing him of the German's intentions. He was referred to the Department of Justice, and at their dictation continued in contact with Rintelen. Shortly thereafter David Lamar and his friend Henry Martin took a trip to Minneapolis, where they met Congressman Frank Buchanan and Ex-Congressman
- 75. Robert Fowler, both of Illinois. Out of that conference grew a plan for forming a labor organization the object of which was ostensibly peace, and actually an embargo upon the shipment of munitions abroad, but whether Buchanan and Fowler knew of von Rintelen's connection with the scheme remains to be proved. It can be readily seen that such a labor organization, if it had actually represented organized labor, could have forced such a stoppage, either by its collective potential voting power and influence, or by fostering a nation-wide strike of munitions workers. The nucleus formed in Chicago, about one William F. Kramer. "Buchanan and Fowler came to me in June here in Chicago," said Kramer, "and told me about their plan to form a council. We opened headquarters, and we engaged two organizers, James Short and J. J. Cundiff, who got $50 a week apiece, a secretary, L. P. Straube, who got $50 a week, and a stenographer. I was a vice-president, but I didn't get anything. We were known then as Labor's Peace Council of Chicago, and we were supposed to be in it because of our convictions against the shipment of munitions. And I'll say that organized labor was made the goat." Buchanan had no idea of restricting the council to one city. He called upon Samuel Gompers, head of the American Federation of Labor, at Atlantic City on June 9 and tried to induce him to back a movement in Washington for an embargo. Gompers refused flatly and completely to have anything to do with the plan, especially when Buchanan made known his associates. Those associates were busy meanwhile lobbying in Congress, representing themselves as friends of organized labor, and pressing the embargo question. About a week later Congressman Buchanan inflated the Chicago organization into Labor's National Peace Council, with headquarters at Washington, to recommend the convocation of a special session of Congress at once to "promote universal peace," which meant simply "to promote the introduction and enactment of an embargo." Its members met frequently, and annoyed the President and other
- 76. important men,—even Andrew Carnegie,—with their importunings for attention, and got exactly what they wanted—wide publicity. About July 10 Andrew D. Meloy, whose office in New York Rintelen was sharing at the time, noticed that his German associate began to keep a clipping-file of news of the Council. Meloy learned of the project, and assured Rintelen that he was foolhardy to attempt, by bribery of labor officials, to divert common labor from earning high wages. To which Rintelen replied brusquely: "Thanks. You come into this business about 11:45 o'clock." Rintelen sent a telegram to Lamar in Chicago on July 16, the text of which follows: "E. Ruskay, Room 700 B, Sherman Hotel, Chicago. "Party who receives $12,500 monthly from competitors is now interfering with business in hand. Do you know of any way and means to check him? Wire. "F. Brown." "Ruskay" was Lamar. Later in the day the German sent this message: "Twelve thousand five hundred now at capitol. Conference here today plans to guarantee outsiders and settlement possible within few days. New issue urgently needed. Notify B." The "party" mentioned in the first despatch was the code designation for Gompers, and he was indicated in the second message as "Twelve thousand five hundred." "B" was Buchanan, upon whose connection with labor Rintelen told Meloy the success of the plan rested. Lamar hurried to New York, arriving July 19, and met Rintelen in a limousine at the 100th Street entrance to Central Park; on the ride which followed the "Wolf" told Rintelen that a strike then going on among the munitions workers at Bridgeport was "only a beginning of his efforts," and that within thirty days the
- 77. industry would be paralyzed throughout the country. Meloy advanced the information that Gompers had just gone to Bridgeport to stop the strike, to which Lamar replied: "Buchanan will settle Gompers within twenty-four hours!" The clippings kept coming in as testimony to the vigorous work being done by the organization's press bureau: the Council attacked the Federal Reserve Banks as "munitions trusts," it cited on July 8 nine ships lying in port awaiting munitions cargoes, and attacked Dudley Field Malone, then Collector of the Port of New York, for permitting such ships to clear; it claimed to represent a million labor votes, and four million and a half farmers; it listened eagerly to an address by Hannis Taylor, a disciple of the late warmhearted Secretary of State, Mr. Bryan, in which Taylor criticized President Wilson and was roundly cheered by the German-American element in the audience. Semi-occasionally during the midsummer heat Charles Oberwager, attorney for the Council (whose firm had received handsome fees from von Papen), rose to deny any German connection with the organization. The Council assailed Secretary Lansing as a man "whose radicalism was liable to plunge this nation into war." The Council assailed, in fact, any project which furthered the interests of the Allies. Rintelen began to have his doubts of the effectiveness of Lamar's work. The bank account in the Trans- Atlantic Trust Company had dwindled from $800,000 to $40,000, and Rintelen admitted that his transactions with Lamar cost him several hundred thousand dollars. Labor's National Peace Conference died quietly, Lamar flitted away to a country estate at Pittsfield, Mass., and Rintelen started across the Atlantic Ocean. August wore on. The Council was getting ready for a second gaseous session, when Milton Snelling, a representative of the Washington Central Labor Union, who had been elected a first vice- president of the Council, withdrew from its membership, because he "discovered persons participating in the meetings who have been hanging on the fringe of the labor movement for their own personal aggrandizement, men who have been discarded ... others never
- 78. having been members of any organization of labor," and because Jacob C. Taylor, the cigar-making delegate from East Orange, N. J., said, in answer to a query as to the Council's purpose: "We want to stop the export of munitions to the Allies. You see Germany can make all the munitions she wants." Then—and it may be coincidence —about one week later the New York World began its publication of certain of the papers found in the brief case which Dr. Heinrich Albert, of the German Embassy, allowed to escape him on a New York elevated train; on August 19 Buchanan resigned the Council, and Taylor was elected to succeed him. Indictments were returned against Rintelen, as well as against Lamar, Martin, Buchanan and their associates, on December 28, 1915. Buchanan at once exploded with a retaliatory demand for the impeachment of United States District Attorney Marshall, upon which Congress dared not take action. Marshall gracefully retired from the trial in May, 1916, lest he prejudice the Government's case, and Lamar, Martin and Rintelen were convicted of infraction of the Sherman Anti-Trust Law and sentenced to one year each in a New Jersey prison. Thus ended Labor's National Peace Council, thanks to David Lamar. The project for an embargo looked attractive to the Embassy, however—so attractive that while the Council was at the height of its activity, Baron Kurt von Reiswitz wrote on July 22, 1915, from Chicago to Dr. Albert: "Everything else concerning the proposed embargo conference you will find in the enclosed copy of the report to the Ambassador. A change has, however, come up, as the mass meeting will have to be postponed on account of there being insufficient time for the necessary preparations. It will probably be held there in about two weeks. "Among others the following have agreed to coöperate: Senator Hitchcock, Congressman Buchanan, William Bayard Hale of New York
- 79. and the well known pulpit orator, Dr. Aked (born an Englishman), from San Francisco. "Hitchcock seemed to be very strong for the plan. He told our representative at a conference in Omaha: 'If this matter is organized in the right way you will sweep the United States.' "For your confidential information I would further inform you that the leadership of the movement thus far lies in the hands of two gentlemen (one in Detroit and one in Chicago) who are firmly resolved to work toward the end that the German community, which, of course, will be with us without further urging, shall above all things remain in the background, and that the movement, to all outward appearances, shall have a purely American character. I have known both the gentlemen very well for a long time and know that personal interest does not count with them; the results will bring their own reward. "For the purposes of the inner organization, to which we attribute particular importance, we have assured ourselves of the coöperation of the local Democratic boss, Roger C. Sullivan, as also Messrs. Sparman, Lewis and McDonald, the latter of the Chicago American. Sullivan was formerly leader of the Wilson campaign and is a deadly enemy of Wilson, as the latter did not keep his word to make him a Senator; therefore, principally, the sympathy of our cause." One is inclined to wonder where Rintelen's vast credits went, during his short visits in 1915. Lamar took a goodly sum, as we have seen; the negotiations for the purchase of the Krag rifles cost him no small amount; his ship bomb activities required a considerable payroll. But as further evidence of the high cost of causing trouble, we must consider briefly the profligate methods he employed in other attempts to inflame and seduce labor. A walkout by the longshoremen of the Atlantic coast would cripple the supply of munitions to Europe, and might be successful enough to cause a shell famine in France of which the Central Powers could
- 80. readily take advantage. There were 23,000 dock-workers in American ports; they must be guaranteed a certain wage for five weeks of strike; the cost in wages alone would therefore amount to about $1,635,000, besides service fees to intermediaries. He had the money, and the first step was taken in the otherwise placid city of Boston. On May 7, 1915, the day the Lusitania sank, William P. Dempsey, the secretary-treasurer of the Atlantic Coast International Longshoremen's Union, met Dennis Driscoll, a Boston labor leader and former city office-holder, at the old Quincy House in Hanover Street. Driscoll said that Matthew Cummings, a wealthy Boston grocer, had outlined to him the plan for the strike, and said he was acting for parties who were willing to pay a million dollars. Dempsey maintained his poise when the startling information was recited, but he was frightened, and at the conclusion of the interview he telegraphed at once to T. V. O'Connor, the president of the union, requesting an interview. The two union men met in Albany and discussed the affair pro and con, arriving at the conclusion that they had best reveal the plot to the Government. O'Connor accordingly told of the negotiations to Secretary Wilson of the Department of Labor, and then in connivance with the Secret Service, went on dealing with the grocer, constantly pressing him for the identity of the principals who, he said, were prepared to supply all the necessary money. He implicated George Sylvestor Viereck, the editor of a subsidized German propaganda-weekly called The Fatherland, and said that he had been introduced to him by Edmund von Mach. Neither of those men figured except as intermediaries, and Cummings suggested that Dr. Bernhard Dernburg, a loyal propagandist then in the United States, was the director of the enterprises. Owing to the high pitch of public feeling over the Lusitania, Cummings could not receive permission from his superiors to go ahead with O'Connor, but he did his best to keep O'Connor interested. The latter, fearing that German agents were at work on the Pacific coast, took a trip to the far West, and during his absence Cummings telegraphed him twice. There the affair ended, for
- 81. O'Connor ignored the message, and on July 14 returned to New York to find that a German attempt to force a walkout on the New York waterfront had failed, and that Cummings had stopped playing with fire and had gone back to his grocery in Boston. When the Government turned the story over to a newspaper to publish on September 13, the time was not ripe to fix the responsibility for the attempt. Dr. Dernburg was a popular scapegoat at the time, and the implication of his authority in the attempt was allowed to stand. Rintelen was in Donington Hall, a prison camp in England, and it was months thereafter before the United States and British Secret Services had fully compared notes on him. By that time there were other charges lying against him which promised better cases than an abortive attempt to promote a strike 'longshore. We have witnessed the cumulative influence of newspaper reports in surrounding Labor's National Peace Council with an almost genuine atmosphere of national interest; we have been able to picture the hostility which the publication of the longshoremen's strike story aroused in legitimately organized labor; and although as a typical instance of newspaper influence we should postpone the following incident, it is a temptation too great to resist. It is the story of The Story That Cost an Ambassador, and if any further plea for its introduction be needed, let it be that it is another subtle attempt upon labor in the summer of 1915.
- 82. Dr. Constantin Dumba, Austrian ambassador to the United States, recalled after the disclosures of the correspon- dence captured on the war correspondent, Archibald James F. J. Archibald, an American correspondent who had seen most of the wars of recent years, and who wanted to see more, set sail from New York on August 21, 1915, for Amsterdam, with his wife, his campaign clothes, and a portfolio. At Falmouth, England, the usual search party came aboard, and inspected the papers in the portfolio. Archibald proved to be an unofficial despatch-bearer, upon
- 83. whom his German and Austrian acquaintances in the United States placed great reliance—such men as Papen, Bernstorff, and Dr. Constantine Dumba sent reports to their governments in his care. On September 5 the New York World burst forth with the text of one of the letters—one from Dr. Dumba, the Austro-Hungarian ambassador at Washington, to his chief in the foreign office at Vienna, Baron Burian. It is worth reproducing here intact: "New York, August 20." "Your Excellency: "Yesterday evening Consul-General von Nuber received the enclosed aide memoire from the chief editor of the local influential paper Szabadsag, after a previous conversation with me in pursuance of his verbal proposals to arrange for strikes at Bethlehem in Schwab's steel and munitions factory and also in the middle West. "Archibald, who is well known to your Excellency, leaves today at 12 o'clock on board the Rotterdam for Berlin and Vienna. I take this rare and safe opportunity of warmly recommending these proposals to your Excellency's favorable consideration. It is my impression that we can disorganize and hold up for months, if not entirely prevent, the manufacture of munitions in Bethlehem and the middle West, which, in the opinion of the German military attaché, is of great importance and amply outweighs the comparatively small expenditure of money involved. "But even if strikes do not occur it is probable that we should extort under pressure more favorable conditions of labor for our poor downtrodden fellow countrymen in Bethlehem. These white slaves are now working twelve hours a day, seven days a week. All weak persons succumb and become consumptive. So far as German workmen are found among the skilled hands means of leaving will be provided immediately for them.
- 84. "Besides this, a private German registry office has been established which provides employment for persons who voluntarily have given up their places. It already is working well. We shall also join in and the widest support is assured us. "I beg your Excellency to be so good as to inform me with reference to this letter by wireless. Reply whether you agree. I remain, with great haste and respect, "Dumba." The aide memoire, written by the editor of a Hungarian weekly, proposed to create unrest by a campaign in foreign language newspapers circulated free to labor, muck-raking labor conditions in Bethlehem, Youngstown, Cleveland, Pittsburg, and Bridgeport, where there were great numbers of foreign workmen, Hungarians, Austrians, and Germans. This was to be supplemented by a "horror novel" similar to the bloody effort of Upton Sinclair to describe the Chicago stockyards. Special agents of unrest, roll-turners, steel workers, soapbox orators, picnic organizers, were all to be insinuated into the plants to stir up the workmen. This editor had stirred them up a few weeks before at Bridgeport—the strike which Lamar claimed as his own accomplishment—and he presented to Baron Burian a really comprehensive plan for creating unrest through his well-subsidized foreign-language press. And in passing it on, Dr. Dumba stood sponsor for it. The British government saw in the discovery of the letter and the cool impudence of it, a rare chance for propaganda in America. So, as has been said, the World published the story, and at once the wrath of the truly American people justified President Wilson in doing what he and Secretary Lansing had already determined to do —to send Dr. Dumba home. Perhaps Dumba's reference to the "self- willed temperament of the President" in another note found on Archibald had something to do with the haste with which the Ambassador's recall was demanded; it followed on the heels of the publication of the letter:
- 85. Welcome to our website – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com