Download the full version of the ebook at ebookfinal.

com

Mediating Between Concepts and Grammar Holden

Hartl

https://ebookfinal.com/download/mediating-between-concepts-
and-grammar-holden-hartl/

OR CLICK BUTTON

DOWNLOAD EBOOK

Download more ebook instantly today at https://ebookfinal.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Semantics and morphosyntactic variation qualities and the

grammar of property concepts 1st Edition Francez

https://ebookfinal.com/download/semantics-and-morphosyntactic-
variation-qualities-and-the-grammar-of-property-concepts-1st-edition-
francez/
ebookfinal.com

Chaos Arun V. Holden (Editor)

https://ebookfinal.com/download/chaos-arun-v-holden-editor/

ebookfinal.com

Principles of Population Genetics 4th Edition Daniel L.

Hartl

https://ebookfinal.com/download/principles-of-population-genetics-4th-
edition-daniel-l-hartl/

ebookfinal.com

Ruby On Rails 3 Tutorial 1st Edition Michael Hartl

https://ebookfinal.com/download/ruby-on-rails-3-tutorial-1st-edition-
michael-hartl/

ebookfinal.com
The HDI Handbook 1st Edition Happy Holden

https://ebookfinal.com/download/the-hdi-handbook-1st-edition-happy-
holden/

ebookfinal.com

An Introduction to Physical Geography and the Environment

2nd Edition Joseph Holden

https://ebookfinal.com/download/an-introduction-to-physical-geography-
and-the-environment-2nd-edition-joseph-holden/

ebookfinal.com

Basic Irish A Grammar and Workbook Grammar Workbooks Nancy

Stenson

https://ebookfinal.com/download/basic-irish-a-grammar-and-workbook-
grammar-workbooks-nancy-stenson/

ebookfinal.com

The Global Social Policy Reader First Edition Nicola

Yeates & Chris Holden

https://ebookfinal.com/download/the-global-social-policy-reader-first-
edition-nicola-yeates-chris-holden/

ebookfinal.com

Mediating International Crises Routledge Advances in

International Relations and Global Politics 1st ed Edition
Jon Wilkenfeld
https://ebookfinal.com/download/mediating-international-crises-
routledge-advances-in-international-relations-and-global-politics-1st-
ed-edition-jon-wilkenfeld/
ebookfinal.com
Mediating between Concepts and Grammar

WDE

G
Trends in Linguistics
Studies and Monographs 152

Editors
Walter Bisang
(main editor for this volume)
Hans Henrich Hock
Werner Winter

Mouton de Gruyter
Berlin · New York
Mediating between
Concepts and Grammar

Edited by
Holden Härtl
Heike Tappe

Mouton de Gruyter
Berlin · New York
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter G m b H & Co. KG, Berlin.

© Printed on acid-free paper which falls within the guidelines

of the ANSI to ensure permanence and durability.

ISBN 3-11-017902-4

Bibliographic information published by Die Deutsche Bibliothek

Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data is available in the Internet at <http://dnb.ddb.de>.

© Copyright 2003 by Walter de Gruyter G m b H & Co. KG, D-10785 Berlin

All rights reserved, including those of translation into foreign languages. No part of this
book may be reproduced or transmitted in any form or by any means, electronic or mechan-
ical, including photocopy, recording or any information storage and retrieval system, with-
out permission in writing from the publisher.
Cover design: Christopher Schneider, Berlin.
Printed in Germany.
Contents

Mediating between concepts and language -

Processing structures 1
Heike Tappe and Holden Härtl

Mediating between
non-linguistic and linguistic structures

Coordination of eye gaze and speech in sentence

production 39
Femke F. van der Meulen

Time patterns in visual reception and written phrase

production 65
Philip Cummins, Boris Gutbrod
and Rüdiger Weingarten

Animacy effects in language production: From mental

model to formulator 101
Kathy Y. van Nice and Rainer Dietrich

Incremental preverbal messages 119

Markus Guhe

Word order scrambling as a consequence of incremental

sentence production 141
Gerard Kempen and Karin Harbusch

The linearization of argument DPs and its semantic

reflection 165
Andreas Späth
vi Contents

Semantics as a gateway to language 197

Heike Wiese

Mediating between
event conceptualization and verbalization 223

Temporal relations between event concepts 225

Elke van der Meer, Reinhard Beyer, Herbert Hagendorf,
Dirk Strauch and Matthias Kolbe

Segmenting event sequences for speaking 255

Ralf Nüse

Events: Processing and neurological properties 277

Maria Mercedes Pinango

Aspectual (re-)interpretation: Structural representation

and processing 303
Johannes Dölling

Type coercion from a natural language generation point

of view 323
Markus Egg and Kristina Striegnitz

The mediating function of the lexicon 349

The thematic interpretation of plural nominalizations 351

Veronika Ehrich

Competing principles in the lexicon 379

Andrea Schalley
Concepts of motion and their linguistic encoding
LadinaB. Tschander

Too abstract for agents? The syntax and semantics of

agentivity in abstracts of English research articles
Heidrun Dorgeloh and Anja Wanner

Index of names

Index of subjects
Mediating between concepts and language -
Processing structures*

Heike Tappe and Holden Härtl

1. Modules and interfaces

One of the main functions of language is to abstract over complex

non-verbal message structures. The language system generates
highly compact linguistic material which, however, must still enable
the recipient of the corresponding linear grammatical sequence to
fully infer the intended message. To guarantee this a device is re-
quired which links concepts and grammar in a systematic fashion by
negotiating the requirements of both the generalized linguistic struc-
tures and the underlying conceptual complexes. Typically, this me-
diating function is instantiated by an interface. Any interface device
has to satisfy procedural requirements because linguistic structure
building must accommodate the fact that different types of informa-
tion are available at different points in time.
Regarding aspects of design, an interface is a virtual or an actual
surface forming a common boundary between independent func-
tional units. It can be defined as a point of information transition and
communication. In a technical sense, an interface definition encom-
passes rules for information transfer and calls for a characterization
of the kind of data that can be handed over from one unit to the
other. This also entails the specification of structure-sensitive opera-
tions over those representations that are the output structures of one
functional component and serve at the same time as input structures
for the subsequent component. The diction independent functional
unit is akin to the term module, in that both notions imply a more or
less autonomous and specialized computational system to solve a
very restricted class of problems and uses information - which are its
proprietary - to solve them (cf. Fodor 1998).
2 Heike Tappe and Holden Härtl

In cognitive science it is widely held, that at least some human

cognitive mechanisms are organized in modules.1 Fodor (1983) de-
fines them as cognitive systems characterized by nine criteria, some
of which concern module-internal information processing with im-
plications for how the interface between such modules is to be de-
fined. The most prominent of these criteria are informational
encapsulation and domain specificity, meaning that; first, the inner
workings of a module cannot be directly influenced from the outside.
Second, that each module computes information of one distinct type,
which, however, has to be of tremendous significance to the species.
Further characteristic features are the following: Unconsciousness,
i.e. module-internal processing is opaque to introspection. Speed and
shallow output, which characterize modules as extremely fast cogni-
tive sub-systems producing a particular output, albeit without provid-
ing information about the mediating stages preceding it.
Additionally, modules are processing pre-determined inputs, which
in turn result in pre-determined outputs devoid any contextual influ-
ence (obligatory firing)} Since it was advanced Fodor's notion of
modularity has stimulated a vivid controversy and an enormous body
of research. In particular the idea of information encapsulation has
become fundamental to computer science. Many standard technolo-
gies of programming are based on this feature. Modularity also plays
a key role in artificial intelligence and computational linguistics: To-
day even systems within sub-symbolic intelligence such as neural
network systems depart from their traditional homogenous architec-
tures and use somewhat modular approaches especially so to natural
language processing (cf. McShane and Zacharski 2001). While it is
thus largely agreed upon that the human mind/brain is organized into
domain specific components (except in rigorous connectionist ap-
proaches), it can be witnessed that the current interpretation of the
term module varies immensely depending on the underlying general
framework (cognitivist, neuro-psychological, evolutionary connec-
tionist, etc.). Generally, it seems that Fodor's modularity assump-
tions are only partly shared in existing models of the human
mind/brain, i.e. the proposed modules are not usually held to possess
all nine Fodorian criteria.
 Mediating between concepts and language - Processing structures 3

The related question whether the human language system is

carved up into functional units and - more strongly - whether these
or some of these are full-fledged modules in Fodor's sense has been
a hotly debated question in linguistics, philosophy and psychology
over the last two decades. Because of space limitation we cannot
reiterate this intricate discussion (but cf. e.g. Karmiloff-Smith 1992,
Marshall 1984, Frazier 1987a, Smolensky 1988 and Müller 1996 for
varying viewpoints on modularity). Generally, this thematic complex
is closely connected with a persistent delimitation effort in linguis-
tics. It is broadly held indispensable for both the definition of the
discipline and for scientific distinctness to accomplish an analysis of
language as a formal system. This endeavor dating back to de Saus-
sure (1916) has had its reflex in syntactic and semantic theory alike.
Consequently, the predominant position subsumes under the term
syntax language specific competencies of how symbols of some lan-
guage may be combined independent of meaning, of other cognitive
computations, and of socio-cultural requirements (cf. e.g. Chomsky
1986). Likewise, formal semantics strives to explicitly identify those
aspects of meaning that are genuinely linguistic, i.e. abstract-able
from general world knowledge, and at the same time persistent in all
syntactic alternation contexts (cf. Cresswell 1978, Montague 1970,
for an overview Bäuerle 1985). In the consequence formal ap-
proaches in linguistics have to date been primarily engaged in con-
sistently explicating language internal structures.
Starting in the 70ies, research in cognitive science, anthropology,
and psychology inspired approaches that deny the autonomy of syn-
tax - and of linguistic subsystems in general - in relation to concep-
tual structure. They interpret grammatical phenomena in terms of
more general cognitive principles with applications outside lan-
guage. These have been subsumed under the terms of cognitive
grammar and functional grammar (e.g. Bates and MacWhinney
1984, Deane 1992, Lakoff 1991, Langacker 1987, 1990, 1991,
Gärdenfors 2000, and Tomasello 2000a, 2000b). Without the as-
sumption of functional units that are engaged in some kind of divi-
sion of labor, the notion of a restrictive mapping device becomes
superfluous as the different parts of the language faculty are concep-
4 Heike Tappe and Holden Härtl

tualized as being highly interactive and having access to basically

the same information and knowledge sources.
The epistemological question, whether a formalist or a functional-
ist conception is preferable, gains in relevance when we take into
consideration language processing. The overarching endeavor to de-
velop models for language production and comprehension systems
calls for a specification the relevant sub-components and carries in
itself the need to describe and to explain the interaction of informa-
tional sources. This objective is characteristic for approaches that
attempt to preserve some of theoretically sound and the empirically
founded assumptions of theoretical linguistics and to incorporate
them into a language processing framework (cf. e.g. Levelt 1989,
Bierwisch and Schreuder 1992). Language production and language
comprehension processes are based on representations, on which
they operate. The computation of the linguistic meaning and thus the
communication of information are impossible without an accessibil-
ity of both general and linguistic knowledge. From this follows the
prime question: Which kinds of information interact in what fashion
and at what points in time during language processing? What we are
addressing here is the processing criterion, i.e. are the representa-
tions a given linguistic theory proposes computable by a language
processing system (Marcus 1982, Fodor 1983, Frazier 1987b, Fra-
zier, Clifton and Randall 1983, Berwick and Weinberg 1983). This
means that if we assume that grammars are theories of abstract lin-
guistic competence (e.g., Chomsky 1986), we have to ask whether
they may or may not provide an appropriate framework for under-
standing the mental processing of language (Stillings et al. 1998:
435).
Unfortunately, the discussion between different schools in lan-
guage research remains - as Newmeyer (1998) points out - to date
largely unsatisfactory. They tend to avoid direct confrontation and
thus they generally are unaware of the compatibility of their results.
For the most part this observation also characterizes the interdisci-
plinary communication on matters of modularity and in the conse-
quence on the structures and processes, which play a role at the
interfaces in question. While the understanding of how the linguistic
 Mediating between concepts and language - Processing structures 5

and the non-linguistic system interact, constitutes one of the most

interesting and central questions in language research, both an intra-
disciplinary and interdisciplinary convergence seems to be a long
way off. The respective definitions of the interface between grammar
and concepts - as well as its allocated character and scope - vary
substantially subject to the vigorousness of the underlying modular-
ity assumption. In the well-established Levelt model (1989) - that
has provided the reference architecture for the majority of research
in language production - the most intensively discussed interface
representation is the so-called preverbal message. In the rigorous
interpretation it links non-linguistic and linguistic structures. How-
ever, the question whether the preverbal message itself is to be inter-
preted as purely non-linguistic is to date still hotly debated. And, in
the consequence there exist profound controversial assumptions
about its general character and content. From this follows that the
impact of features in the preverbal message on the subsequent repre-
sentations remains under discussion especially regarding the realiza-
tion of this information by the sub-components of the linguistic
system. This concerns e.g., the question whether the linguistic reali-
zation of a preverbal message such as the word order of the utterance
is determined by the order in which concepts are selected, or, is the
outcome of purely grammatical operations.
In order to enhance both intra- and interdisciplinary exchange
about these issues, the current volume brings together researchers
both from theoretical linguistics and from language processing as
well as researchers from adjacent disciplines such as computer-
science and psychology. While all contributors acknowledge some
division of labor between lexical(-semantic), morphological, syntac-
tic, and phonological structuring, it is not surprising that they do not
define the respective sub-components and their substance in the
same way. Especially the term semantics receives different interpre-
tations as notions relating to meaning have long and often controver-
sial histories within the disciplines that contribute to this volume,
which are related to foundational and methodological differences. As
a consequence, the current volume comprises contributions that a
traditional perspective on the interface function in question would
6 Heike Tappe and Holden Härtl

not integrate. Although findings from language comprehension stud-

ies are also discussed, the main body of contributions center around
aspects of language production. In this field available definitions of
the concrete interaction between the conceptual/semantic and the
grammatical level are to date still of a tentative nature. The dispute
in the book will shed light on this issue by exploring the several
stages of processing ranging from the conceptual knowledge, its re-
cruitment, and preverbal preparation for linguistic computation, to
finally its grammatical realization.
In the following paragraphs we give an overview of prominent -
and in the interest of space selected - interface conceptions from the
perspectives of both the grammatical and the conceptual systems and
relate those to questions of language processing. Subsequently, we
introduce the contributions to this volume, which demonstrate vari-
ous parallels and common attitudes in spite of differences in focus,
research background, and modeling.

1.1. Linking to syntax

The assumption that a linguistic capacity of the human mind/brain

enables speakers to competently master their native language is
tightly intertwined with the influential work of Fodor (1983) and
Chomsky (1986). Both assert the existence of a specialized language
faculty, which is conceived as a mental organ3 and as being internal-
ly organized into several functional subsystems. Especially Chom-
sky's arguments in favor for a linguistic module are based on pheno-
mena which are hard to explain on other but syntax-internal
grounds.4 Further compelling evidence for genuine linguistic syntac-
tic principles are found in language acquisition (e.g. Meisel 1990,
Stromswold 1992, Tappe 1999) and Creole language data (e.g. Bick-
erton 1990).5 The division of the cognitive system into functional
sub-components implies the existence of specific principles organiz-
ing the representations within each component. More importantly in
the present context, it follows from this conception that mapping
mechanisms between the components be specified.
 Mediating between concepts and language - Processing structures 7

It is generally acknowledged that for a successful coupling be-

tween (lexical) semantics and syntax predicates have to provide such
lexical information as the number of arguments and the syntactic
structure into which these arguments are to be integrated. In spite of
this broad consensus, the proposals about how such an interrelation
between syntactic and semantic structures may be realized vary sub-
stantially.
Recent syntactic theories characterize syntactic operations by mi-
nimalist principles, which are subject to directives of economy and
explicitness. In the minimalist framework (cf. e.g. Chomsky 1995)
lexical items enter the syntactic building process fully equipped with
their grammatically relevant features including categorial, semantic
argument structure, and thematic features. The relevant operation
select maps lexical items from a set of elements activated from the
lexicon onto the computational process. This process makes use of
two basic mechanisms, i.e. merge and move. Furthermore, procrasti-
nate regulates that syntactic movement has to take place as late as
possible in the derivation, if there is a choice, which differs from
language to language thus creating language-specific word order
variations. The underlying idea is that covert movement is 'less
costly', because it does not have to pied-pipe phonological features
(cf. e.g. Chomsky 1995, Wilder and Cavar 1994). In this fashion the
syntactic component produces structures that are compatible and
legible to the linguistic levels adjoining the syntactic level and also
to the levels adjacent to the linguistic system itself. The language
faculty has to meet specific interface conditions to allow for interac-
tion with the adjoining nonlinguistic components. This requirement
has led Chomsky (2000) to the conclusion that "language is an op-
tional solution to legibility conditions". These legibility conditions
have to involve principles of how syntactic material is to be mapped
onto phonological representations of the articulatory-perceptual sys-
tem on the one hand, and the semantic representations of the concep-
tual-intentional system on the other.
Developing a somewhat different approach to modeling the lexi-
con-syntax interface within the feature-checking framework of the
minimalist program (Chomsky 1995), van Hout (1996) proposes a
8 Heike Tappe and Holden Härtl

CHecking Event-Semantic Structure model (CHESS). She assumes

that the event structure of a predicate must be syntactically identified
(cf. Grimshaw 1990; Grimshaw and Vikner 1993) and defines the
mapping relation in terms of checking event-semantic features in
functional configurations. There are two structural argument posi-
tions: the specifier positions of AgrS and AgrO. An argument in ei-
ther of these positions identifies an event or subevent by referring to
an event participant that is involved in that (sub)event. Telic event
type features must be checked in AgrOP. Van Hout argues that the
CHESS model accounts for the event-semantic mapping generaliza-
tions in a natural way, explaining the phenomenon of lexical-
syntactic flexibility as a derivative of event-type shifting.
These current developments within syntactic theory are compati-
ble with semantically oriented approaches that assume specific link-
ing mechanisms operational between semantic and syntactic
structure. Here it is held that specific configurational constellations
in the semantic representation determine the syntactic realization of a
language. In Bierwisch (1986) and Wunderlich (1997) the mapping
of arguments onto syntactic structure is organized through the em-
bedding of the arguments in the semantic form representation, i.e. a
predicate-argument structure. Jackendoff (1990) advances a similar
approach with the difference that he assumes correspondence rules
to negotiate between syntactic and semantic-conceptual structure.
Moreover, he also claims that lexical syntactic representation of a
predicate can always be reduced to its lexical semantic representa-
tion. In the consequence he treats the semantic and syntactic infor-
mation of the lexicon as part of conceptual structure whereby
arguments correspond to ontological categories of conceptual struc-
ture.
This latter claim differentiates Jackendoff s account considerably
from most linking theories. Based on the observation that some pairs
of predicates like, e.g. ask and inquire have different syntactic subca-
tegorizations albeit their semantics are identical, Grimshaw (1979)
proposes that predicates select both syntactic objects (nouns phrases,
sentences and semantic objects (propositions, questions, exclama-
tions) with no correlation between the two. The linking between the
 Mediating between concepts and language - Processing structures 9

two distinct types of information is handled by thematic hierarchies

where semantic argument features like AGENT, BENEFACTIVE or
THEME organize the order of arguments to be realized in syntax (cf.
Baker 1997, Grimshaw 1990, Jackendoff 1972 among many others).
AGENTS, for example, surface in a hierarchically higher position (as
subject) than THEMES (as direct object in transitive verb complexes).
The very nature of argument structure is less than clear.6 'Linking
theoreticians' assume that argument structure not only contains the-
matic information but that it is also closely tied up with event struc-
ture, which contains aspectual information (cf. Grimshaw 1990).7
Tenny (1992, 1994) assumes that only aspectually relevant informa-
tion is mapped onto syntax {Aspectual Interface Hypothesis). In the
other extreme, researchers like Rappaport and Levin (1988) encode
no more than syntactically relevant information into argument struc-
ture, which thus does not contain any thematic role specifications.
As becomes evident from this discussion, most of the various exem-
plary conceptions of the mapping between syntax and semantics are
joined by the consistent assumption that there is an independent
level, where lexical properties such as predicate-argument structures
are calculated. However, the question of what kind of information
influences and/or is to be integrated into this structure during lan-
guage processing has not yet received a widely accepted mutual an-
swer. This is partly due to the fact that syntactic theories tend to
center around the outcome of the computation rather than a real time
piecemeal construction of syntactic strings. In this context, the ques-
tion of how information is weighted such that the salience of the
constituents has its reflexes in an incremental syntactic realization
gains central importance.

1.2. Semantics

As was already hinted at in the first paragraph, formal model-

theoretic approaches towards meaning assume a modular organiza-
tion of linguistic processes: A morpho-syntactic component generat-
ing overt linguistic sequences and a semantic component, which re-
lates the grammatical material to extra-linguistic structures. General-
10 Heike Tappe and Holden Härtl

ly the focus of investigation the pairing of syntactic categories and

semantic types and the subsequent model-theoretical interpretation
of the analyses (e.g. in the framework of categorial grammar,
Ajdukiewicz 1935). The prime target is to specify how linguistic
expressions fit the world. Therefore investigations center, first,
around referring expressions, (syntactically encoded in noun
phrases) and, second, around truth-conditions of propositions, in-
cluding the exploration of which inferences follow from a linguistic
expression (cf. e.g. Lewis 1972, Tarsky 1977). Under this perspec-
tive the linking between syntax and semantics the need to further
explicate the linking between syntax and semantics does not arise
because here syntactic structures are considered categorical com-
plexes, whose interpretation is derived compositionally from either
the syntactic parts or their fixed meaning, respectively.
Syntactic constellations are deemed relevant only if the modifica-
tion of a linguistic string results in different entailments such that the
truth conditions underlying the expression in question are altered.
Correlations between certain linguistic expressions are taken to be of
a logical rather than a grammatical nature (cf. Montague 1973, Par-
tee 1975, Dowty 1979). Grammatically different but logically identi-
cal sentences inducing parallel entailments like the three examples in
(1) are generally treated in a homogenous fashion. The differences
between them are ascribed to information structure and focus pack-
aging routines.

(1) a. Somebody killed the fly.

b. The fly was killed. Somebody did something.
c. The fly, somebody killed.

Decompositional approaches strive to grasp further entailments that

cannot be explicitly derived from overt form, but need to be inferred
from inherent meaning features. To this aim they employ the concept
of basic meaning components (cf. Katz and Fodor 1963, McCawley
1971 and many others). Under the assumption that complex mean-
ings are built up from smaller units such as CAUSE or NOT ALIVE,
 Mediating between concepts and language - Processing structures 11

more specific entailments can be logically derived from the sen-

tences in (1), cf. examples in (2).

(2) a. Somebody killed CAUSE[somebody,[BECOME[-iALiVE fly]]]

thefly.
b. The fly was killed.
c. The fly, somebody Somebody did something
killed. Somebody caused something to
happen
Something became not alive

Although purely logically oriented, decompositional approaches can

thus capture implicit entailments, they cannot address the issue of
contextually driven truth evaluations. Under the assumption that - in
order to adequately convey a message structure - such information
structural values determining an expression are to be defined as re-
flexes of the speaker's intention, a broader notion of what is meant
by the term propositional content is needed. Consequently, the truth
conditions underlying the example in (2c) have to imply that this
sentence can have been uttered only in a specific contextual (i.e. a
contrastive) situation: The respective discourse set needs to contain
at least one more object such that the contrastive function of the ex-
pression can be evaluated as true.
A further shortcoming of purely logically oriented semantic theo-
ries is that they have to define truth conditions that must hold in
every possible situation the corresponding expression occurs in. For
example, a semantic analysis for short passives - cf. examples in (3)
- has to explain the fact that passives can be accompanied by pur-
pose clauses, which imply that there is an implicit agent denoted in
the matrix clause. This leads to the conclusion that the truth condi-
tions underlying passives have to signify an (existentially bound)
individual (cf. Brody and Manzini 1988, Roeper 1987, Koenig and
Mauner 1999, for discussion).

(3) a. The letter was written in order to impress the duchess.

b. The letter was written but it never reached its addressee.
12 Heike Tappe and Holden Härtl

In (3 a) the implicit agent of the purpose clause (the one who im-
presses) and the demoted entity in the matrix clause (the one who
wrote the letter) are co-referential. Although this can surely be taken
as evidence for the conceptual existence of an implicit agent in short
passives, nothing prevents us from rejecting this assumption in cases
like (3b) where no purpose clause is added. However, the latter hy-
pothesis can merely be upheld if we assume a level of language
processing where only those pieces of information are provided
which are relevant for a successful realization of the communicative
act. Consequently, for a message like (3b) an implicit agent - as it
does not gain any referential salience - might not be present in the
semantic-conceptual structure underlying the message. Only in cases
where a conceptual activation of a corresponding entity becomes
relevant (as in (3a)) this knowledge has to be retrieved. Yet, in order
to cover cases where contextual constellations indeed require the
conceptualization of an entity, truth-theoretic analyses over-generate
and represent both sentences alike. Obviously, this problem concerns
the notion of conceptual activeness and here empirical and proce-
dural evidence may provide a solution by indicating the concrete
conceptual constellations holding during actual language processing
in real time. Against this background, it is apparent that experimental
results can not only help to reveal stages of language processing and
to define an adequate processing model, but also to indicate how
linguistic expressions be analyzed and to determine corresponding
representations.
Generally, semantic theories are, of course, not oblivious of the
importance of context- and situation-dependent aspects of meaning
construction, as is emphasized, e.g. in situation semantics (cf. Bar-
wise and Perry 1983, among others). Here, sentence meanings are
built up compositionally as functions from reference situations to
described situations. Thereby contextual factors reflecting specific
speech situations are incorporated into the study of meaning such
that expressions like I, this, and yesterday in I saw this plate on the
table yesterday are evaluated against the context of the actual speech
event. In this way, adequate means to determine corresponding truth-
values are provided. In a similar fashion, Kaplan (1977) distin-
 Mediating between concepts and language - Processing structures 13

guishes between fixed context-independent character of an expres-

sion and its content evaluation. The latter concept accounts for the
fact that the meaning of linguistic units is adapted to contextual re-
quirements and acknowledges that the interpretation of indexical
expressions like (demonstrative) pronouns is dependent on time.
In contrast to the model-theoretic approaches sketched above,
semantic theories that include grammatical aspects into the analysis
of linguistic expressions are enabled to explain entailment relations
between sentences that are based on lexical and morpho-syntactic
constellations. Consider the following examples:

(4) a. John broke the mirror. The mirror broke.

b. John destroyed the mirror. 'The mirror destroyed.

The difference between destroy and break can be put down to inher-
ο

ent features of the respective lexical entries. Levin and Rappaport

Hovav (1995), for example, argue that only those verbs detransitiv-
ize which can express a change of state coming about without the
intervention of a volitional agent, i.e. which can instead denote an
effect of a natural force. In this sense, lexical semantics seeks to de-
fine predictable relations between semantic features and overt
grammatical behavior, which, at the same time, allows to predict
possible semantic relations between sentences such that a transitive
verbal complex entails the corresponding intransitive one and vice
versa.
Likewise, decompositional lexico-semantic approaches control
the mapping of grammatically relevant aspects of meaning structures
onto linguistic form by encoding grammatically visible differences
in meaning by way of decompositonal representations, which are
linked to morpho-syntactic representations. As we pointed out in
paragraph 1.1, for now there is still no agreement on the question
whether meaning aspects visible in grammar are to be defined as a
subset of the conceptual, non-linguistic level of language processing
or rather as part of the linguistic system. The former assumption im-
plies that conceptual structures are directly linked to syntactic struc-
ture - a view that is employed by conceptual semanticists like
14 Heike Tappe and Holden Härtl

Jackendoff (cf. Jackendoff 1992, 1997). Here, conceptual structures

that constitute the non-linguistic message have to be compatible with
both the linguistic system with its independent language-specific
requirements on the one hand, and the conceptual knowledge base
organizing information from the several sensory and memory sys-
tems on the other hand. In contrast, a more modular conception of
the encoding of grammatically relevant aspects of meaning is incor-
porated in theories that assume a separate, lexico-semantic level,
which is organized by strictly linguistic principles. This grammati-
cally determined level - the semantic form - of meaning representa-
tion is distinguished from a non-verbal, conceptual level comprising
propositional information of a message level by semanticists like
Bierwisch (1983), Dölling (1998), Ehrich (1992), Härtl (2001), Lang
(1994), Olsen (1998), and Wunderlich (1997). Similar distinctions
have been formulated in Mohanan and Wee (1999), who differenti-
ate a semantic structure from a conceptual structure, or Grimshaw
(1993) who distinguishes between the semantic content of an expres-
sion and its semantic structure. Similarly, the logical form level (LF)
of syntactically reflected meaning aspects such as the scope of quan-
tifiers or of negations in the Government & Binding program and its
successors (Chomsky 1981, 1993 and many others) can be consid-
ered a reflex of the need for a linguistically determined level of se-
mantic information. These rules generate semantically adapted
structures, which then are mapped onto representations of the con-
ceptual-intentional system of the conceptual knowledge base inter-
facing the several conceptual subsystems that organize the world-
knowledge of an individual.9 While these conceptions are in them-
selves quite elaborated, they are still largely oriented towards the
linguistic representations as outcome of processing stages, while the
processing aspects themselves are largely ignored. In language pro-
duction research, however, it is of prime importance to clarify how
conceptual structures might influence the construction of linguistic
representations and thus also the variability of semantic and syntactic
structures.
 Mediating between concepts and language - Processing structures 15

1.3. Conceptualization

As has become apparent in paragraph 1.2 cognitively oriented se-

mantic theories are primarily concerned with the question of how
semantic representations systematically interface non-linguistic and
syntactic representations. More broadly considered this is a common
goal in the interdisciplinary research aiming at understanding the
language faculty and its interaction with other cognitive capacities.
In this context the basic ontological categories, i.e. objects and
events, and how their respective conceptualization relates to verbali-
zation and comprehension are of prime importance. Growing evi-
dence from psychological and neurological research indicates that
objects and events cannot only be differentiated on philosophical and
theoretical grounds, but that the neural processing of these two basic
entity types engages discriminable sub-parts of semantic memory
(cf. e.g. Caramazza 1997).
Being able to talk about an object or to decode a specific object
reference has as its prerequisite object recognition. This complex
mental operation involves two more basic processes concerning ob-
ject constancy and object categorization. The first one relates to sta-
bility of object recognition independent of spatial transformations,
i.e. regardless of a given object's orientation, size and position.10 The
second one-object categorization- involves the ability to perceive
and categorize different objects as members of the same category. In
order to be able to tackle the second task, perceptual or conceptual
equivalences among the objects within a given class have to be de-
tected (cf. e.g. Anderson 1991, Bloom 1998, Medin 1989).
For the contributions to the current volume these two cognitive
processes are less important than the fact that humans generally ex-
perience objects in various locations and in many different spatial
arrangements. Consequently the spatial configurations in which ob-
jects occur and the spatial relations that hold between different ob-
ject become essential for linguistic encoding of situations. In
verbalization and in comprehension spatial relations between ob-
jects, which may freely employ the multidimensionality of space,
have to be linked to a linear string of linguistic expressions. Verbal
16 Heike Tappe and Holden Härtl

expressions typically contain projective expressions (e.g. left, right)

that are dependent on a specific perspective reflecting a view point
on the described situation. Perspectives are linguistically encoded by
utilizing reference systems, i.e. systematically structured fields of
linguistic expressions. Spatial reference systems are usually subdi-
vided into two major classes. Egocentric reference systems are those
in which relations between objects are specified in relation to body
coordinates of an observer (most prominently body-axes or retinal
coordinates). In environmental reference systems, on the other hand,
locations are characterized via objects other than the speaker; exam-
ples are absolute reference frames employing cardinal directions
{North, South, East, West), or, reference systems making use of
prominent landmarks (e.g. 'hillwards') (cf. Levinson 1996). As has
been pointed out in the literature, the employment of spatial perspec-
tives on a given situation is influenced by various parameters and
often is not maintained throughout a description (cf. e.g. Taylor and
Tversky 1996, Tappe 2000).
Object conceptualization also plays an essential role in event con-
ceptualizations, as in events entities figure as event participants.
Fundamental features of event structure must be accessed to assure
language processing, which e.g. determine during comprehension
which syntactic structure is projected. Depending on whether or not
the speaker/hearer identifies an initiator of the event, the verb class
will vary. A verb like, e.g. push, requires an initiator (which means
at the same time that it is always transitive), whereas break may or
may not encode an event with an initiator (i.e. may also be intransi-
tive). Another feature concerns whether there is an endpoint of the
event (telicity). Telic events must have an underlying direct object
(cf. O'Brian, Folli, Harley and Bever, in prep.).
In the larger context of event conceptualization the influence of
conceptual features like-most prominently-animateness on linguistic
processing and on linguistic encoding are investigated. A feature like
[+animate] is reflected e.g. in sortal preferences for argument roles.
An animated entity is preferably identified as the initiator of an event
and therefore assigned the agent role.
 Mediating between concepts and language - Processing structures 17

The assignment of thematic roles is part of the conceptual struc-

turing of situations, which is a complex process encompassing a va-
riety of conceptual operations. As our environment consists of a
continuous flow of activity, the perceptual and conceptual segmenta-
tion this continuation into meaningful units is a precondition to lin-
guistic encoding. This insight leads to a modification of Level's
principle of natural order, which assumes a strict correspondence of
chronological order and ordering of events.

What counts as natural ordering is different for different domains of dis-

course, and there is no general definition. Still, for certain important cases
the notion is obvious. For event structures, the natural order is the chrono-
logical order of events (Levelt 1989: 138).

That this assumption is not tenable in a strict sense has been demon-
strated in a variety of empirical investigations suggesting that, as
Zacks puts it, "events arise in the perception of observers" (Zacks
1997). Thus, for conceptualizing of event structures some additional
processes like segmentation, structuring, and selection have to be
applied prior to linearization, which transform a continuous stream
of experiences into a highly structured, often non-sequential event
structures.11
Hierarchically organized event types are sometimes held to be
stored in special sub part of the conceptual knowledge base, namely
semantic memory (cf. Kintsch 1980). Semantic memory comprises
an individual's ontological knowledge about the world at large in the
10
format of rather abstract types. The adjective semantic is ambigu-
ous in the given context. In psychological literature a distinction be-
tween general conceptual ontological knowledge and genuine
linguistic semantic knowledge is often either neglected, or, ignored.
In some linguistic approaches, however, semantic and conceptual
knowledge is systematically differentiated (cf. Lang, 1994 for exten-
sive arguments in favor for this distinction).
18 Heike Tappe and Holden Härtl

1.4. Interface in action

In the previous sections we have provided an overview of prominent

approaches to the interfacing of conceptual and linguistic representa-
tions. We have shown that from both sub-disciplines the linking be-
tween syntactic and semantic structures is either approached via
intermediate representations such as argument structure, or taken as
more or less given; e.g. in approaches that advocate quite a direct
coupling between the two as in model-theoretical theories. We have
pointed out persistent problems as how to model the different inter-
face representations or linking mechanisms and some limitations of
the respective approaches.
In the adjacent disciplines psychology and computational linguis-
tics the problem also exists but in a somewhat different fashion. In
both disciplines the processing aspect has been in greater focus as
they do not generally treat language as a formal system in its own
right. Either the overall research interest does not encompass this
aspect - as in psychology for the most part - or, is back-grounded in
the interest of building running systems.
In psychology the interfacing between different components of
the language system is for the most part regarded from the perspec-
tives of the three areas of psycholinguistic inquiry, that is acquisi-
tion, comprehension, and production. With reference to the latter
two areas the main body of research focuses on language compre-
hension, since it is of prime importance to psychological researchers
to make empirical data controllable and subject to experimental
methods. Language production research is judged less manageable in
these respects, especially concerning the production of longer strings
of language, i.e. whole utterances and texts, because it is almost im-
possible to define dependent variables in these cases. Either the ver-
balization situation has to be highly restricted,13 which then leaves
speakers no choices in how to communicate the contents in question
(and renders the whole endeavor pointless), or, the language data
become too variant to pin down the more fine grained aspects of
conceptualization and formulation.14 Thus, psycholinguistic language
production research mainly concentrates on impaired language pro-
 Mediating between concepts and language - Processing structures 19

duction (e.g. in aphasics), analyses of slips of the tongue and speech

pauses, and lexical access studies. Especially in the latter field, intri-
cate experimental paradigms have been developed to tease apart
stages during which different features of a target word become ac-
cessible: A first stage of a preverbal conceptual representation. A
second stage, during which an abstract representation of semantic
and syntactic information is retrieved (i.e. lemma selection, ibid).
And, a third stage, which eventually involves activation of the word's
phonological representation (or lexeme activation, ibid), that will
initiate articulatory encoding (cf. e.g. Jescheniak and Levelt 1994).
As becomes apparent the interface problem is thus tackled in the
transition from the conceptual component to the formulation compo-
nent, as syntactic and semantic features of the target word are acti-
vated in parallel. The utterance formulation is conceived of as being
driven via the selected lexical entries. However, the very nature of
the conceptual representation is usually not addressed as in lexical
access studies the probes for lemma and lexeme activation are either
phonetically or graphically presented word or pictures. Thus, ques-
tions of choice of open class words, collocations, connotations, and
sub-lexical relations and the like are not addressed.
This is akin to the common practice in the computer science,
where lexicalization (or lexical choice) has also become the focal
domain for a variety of sub-problems associated with the transition
from conceptual (what-to-say) to lexical representation and formula-
tion (ihow-to-say) levels (cf. Busemann 1993). Here, too, correspon-
dences between conceptual and lexical entities deviating from the
simple one-to-one pattern are not frequently encountered. In fact,
very few existing NLG systems make a distinction between concep-
tual and semantic representations in any explicit way. Typically, they
strive to reliably express their input from a well defined and limited
domain - and succeed in doing so. In parallel, the syntax-semantics
interface has been shifted into the lexicon: Most theories adhere to a
compositional semantic conception, meaning in this context the con-
struction of utterance meaning (and in the consequence utterance
structure) from the meaning of constituents and phrases. The role of
the other components has been considerable decreased in the conse-
20 Heike Tappe and Holden Härtl

quence and syntax is often reduced to one or two general principles.

Information concerning the categorical identity and combinatorial
constraints are projected from individual lexical entries. Lexical-
Functional Grammar (LFG, Kaplan and Bresnan 1982), Generalized
Phrase Structure Grammar (GPSG, Gazdar et al. 1985), Head
Driven Phrase Structure Grammar (HPSG, Pollard and Sag 1987)
and Unification Categorial Grammar (UCG, Zeevat et al. 1987) are
prominent examples for such monostratal and lexical theories of
grammar.15
In addition to being restricted to limited domains, existing NLG
systems encounter persistent problems in at least three fields: In the
appropriate tackling of synonyms and near-synonyms, in machine
translation and in artificial life applications. These have in common
the fact that a mere one-to-one mapping between the conceptual
level and the linguistic levels does not yield appropriate results.
The solution to these problems is for the most part sought in
modification of the system-architectures. The standard versions of
NLG systems today are modular, relying on a strictly sequential
architecture and a one-way information flow. Sequentiality and
modularity yield stability, but they also result in rigidness of the
system. The antipode to this conception is an integrated architecture,
in which knowledge at all levels acts together. Interactive
architectures are extremely flexible, albeit prone to system break-
down. Between these two extremes, we find architectures that

sequential ( 3 "C O

integrated I 1

interactive (feedback) ( )- " ( 1

blackboard C _ J - -
I J- Η
revision-based ^ ^

Figure 1. Schemes for control of information flow (ibid)

 Mediating between concepts and language - Processing structures 21

allow for various kinds of interaction between the modules.

Interactive architectures allow for feedback processes between
modules, whereas in blackboard architectures every module has
access to common information that is shared between modules and
laid down in a mutual data structure. Revision based achitectures
allow for a limited range of feedback via monitoring components.
(For extended description of the architecture types viz. DeSmedt,
Horacek and Zock 1996).
Apart from these conceptions, there is a growing endeavor to
build hybrid models that combine advantages of different model
types. Most prominently in the revised version of Levelt's model
combines a modular architecture with interactive (connectionist)
substructures - the latter are to be found within the formulator. More
concretely the lemma-model is implemented within a spreading acti-
vation framework (WEAVER++, see Levelt, Roelofs and Meyer
1999).
In sum, the overview presented in this introduction shows that the
processing problem is tightly intertwined with, first, the kinds of
structures and processes we assume at the different stages of proc-
essing and the way we model their interaction - especially so at the
transition from conceptual/semantic to syntactic representations.
And, second, with the underlying modularity assumption, i.e. the
proposed architecture of the language faculty, which also has a
strong impact on the respective interface conceptions. The contribu-
tions in this book address these issues from various viewpoints and
theoretical backgrounds. Either they take on a model-oriented per-
spective, or, concentrate on a specific phenomenon. One phenome-
non that has currently received growing interest in the disciplines
involved is the coupling between conceptualizations of events and
their grammatical realizations. This issue is notoriously complex
(viz. paragraphs 1.1 and 1.2) as the verbalization of events varies
significantly depending on the internal features (e.g. aspectual) and
external characteristics (e.g. the chronological order) of events.
From these starting points, the current volume contributes to the
ongoing discussion about the relevance of empirical and psychologi-
cal evidence for theoretical-linguistic research and vice versa. The
22 Heike Tappe and Holden Härtl

book is based on the assumption that any research on human lan-

guage - even from a heuristic perspective - should include insights
into procedural aspects in the computation of a linguistic expression.
This conception has its roots in the conviction that the ways of proc-
essing data from different levels have to be reflected in the linguistic
target representation. In reverse, even though theoretical explicitness
and fine grained analyses might appear neither manageable nor de-
sirable in the implementation of NLG systems, the integration of
more findings from theoretical linguistics into computer science may
turn out to be useful in more intricate language production domains.

2. The contributions

The mediating function between concepts and grammar is ap-

proached by the contributions to this volume from three interrelated
areas of emphasis: i.) the interplay between non-linguistic and lin-
guistic information in the grammaticalization and linearization of a
preverbal message, ii.) the mapping between non-linguistic, concep-
tual event representations and the ways of verbalizing them, and iii.)
the mediating function of the lexicon in the verbalization of different
types of events. First, questions of the general architecture including
the number of levels, specific ways the information is processed on
them, and the size and the format of the interface representations is
dealt with. Here, the persistence of extra-linguistic information, its
visibility for linguistic processes, and its realization in grammar is
explored. The interplay of the several types of information involved
becomes especially apparent with the issue of event conceptualiza-
tion and verbalization, which at the same time represents a useful
basis for an application of the model assumptions developed so far.
Specifically, the question of how event concepts are stored in mem-
ory and fractionized for language processing is addressed. In this
context, a main issue to be discussed is how grammatical require-
ments determine the verbalization of event concepts and how the
interface can mediate between corresponding informational conflicts.
This thematic complex joins together the contributions of the third
section. The morpho-syntactic realization of event structural features
 Mediating between concepts and language - Processing structures 23

and their effects on the assembly of verb complexes is projected

from principles organized in the lexicon, which are addressed in the
third group of papers. We organize the contributions according to
their main focus into the described three sections while at the same
time the interrelatedness of the issues dealt with allows for repeated
naming of one author in multiple sections. (Authors names appear in
bold letters to associate them to a respective section).

Mediating between non-linguistic and linguistic structures. The

contributions of the first section investigate the influence of different
types of extra-linguistic information on the verbalization of a linguis-
tic string. Here, affects on the linearization of a preverbal message
are of central interest. This requires a modeling of the incremental
realization of the preverbal message as well as a definition of those
meaning components which are reflected in grammar. Against this
background, FEMKE F . VAN DER MEULEN provides evidence from eye
tracking experiments that point to a close link between looking and
verbalization. Like Cummins, Gutbrod, and Weingarten, she uses
spatial configurations to elicit verbal descriptions. Her data shows
that the description of certain types of object arrays is preceded by a
preview, which interacts with the viewing times during the main pass
of the verbalization. Temporal aspect are of focal importance in the
contribution of PHILIP CUMMINS, BORIS GUTBROD, and RÜDIGER
WEINGARTEN also, where the complexity of phrasal structures is re-
lated to the time course of their production. To show also that addi-
tional conceptual information such as the size of the set of concepts
to choose from affects the verbalization of spatial configurations, the
authors provide evidence from eye-tracking and keyboard data to
underpin their hypothesis. The accessibility of conceptually differ-
ently weighted constituents is investigated by K A T H Y Y. VAN N I C E
and RAINER DIETRICH. They disentangle extra-linguistic features
such as animacy and agentivity effects in their impact on word order
and develop a model of how this information is carried down
through the language production system. The authors thus motivate
the incremental processing models as proposed by Guhe as well as
Kempen and Harbusch by pointing to the relevance of extra-
24 Heike Tappe and Holden Härtl

linguistic features that become information structurally relevant dur-

ing processing. MARKUS GUHE proposes an incremental construction
of the preverbal message. He explicates how these piecemeal struc-
tures link to the underspecified semantic representations (as they are
proposed e.g. by Johannes Dölling, Veronika Ehrich, Andreas Späth,
and Ladina Tschander). Here, a critical factor is determined, namely,
the criterion that need to be fulfilled in order for a conceptual entity
to function as a legitimate increment. The incremental processing of
information on the syntactic level is central to the work of GERARD
KEMPEN and KARIN HARBUSCH, which strongly relies on experimen-
tal evidence. They apply a probabilistic method in order to model
word order phenomena in the German midfield and indicate that -
besides syntactic constraints - information structural conditions are
crucial for scrambling. Thereby they mirror the order in which the
constituents become accessible for syntactic processing during com-
putation. Considering the referential status of nominal expressions in
discourse, aspects of word order are discussed by ANDREAS SPÄTH
also. Here, the lexical principles which relate to the syntactic base
generation of lexical entries are determined. By means of these
principles - as is discussed by Veronika Ehrich and Andrea Schalley
also - the link between argument structure and word order is
accounted for where informational structural features are included
into the computational routines at work between semantics and syn-
tax. With these means presuppositions to be derived from nominal
argument phrases can be associated with a current discourse model.
From a general architectural perspective, the interaction between
grammatically visible and invisible meaning components is investi-
gated by HEIKE WIESE in her tripartite model. Drawing on empirical
evidence, she integrates insights from two-level approaches to se-
mantics with conceptual semantics. She advocates semantics as the
interface level of the conceptual and the linguistic system, where it is
a particular SEM-function that makes visible conceptual information
to the linguistic system and generates an under-specified representa-
tion.
 Mediating between concepts and language - Processing structures 25

Mediating between event conceptualization and verbalization.

Spatial and temporal configurations are to be linearized during lan-
guage production. However, while with spatial configurations the
multidimensionality of space has to be transferred onto a linear lin-
guistic sequence, with temporal relations the knowledge about the
canonical sequential ordering of events such as SOIL-WASH can be
employed for structuring the message and thus enhances processing.
This latter hypothesis is supported by the findings of ELKE VAN DER
MEER, REINHARD BEYER, HERBERT HAGENDORF, DIRK STRAUCH,
and MATTHIAS KOLBE, who show in a series of priming experiments
that the disruption of the canonical sequence of events as with WASH-
—SOIL leads to processing difficulties. The authors thus show, how
world-knowledge about events has its reflexes in linguistic event
descriptions. RALF NÜSE approaches the interrelation between event
conceptualization and event verbalization by analyzing language
specific differences between English and German speakers. By com-
paring both linguistic descriptions of visually presented events and
the corresponding eye-movements of the speakers, he comes to the
conclusion that language specific grammatical features are already at
work in the conceptualizes While she also considers the event do-
main, a modular conception is supported by MARIA MERCEDES
PINANGO. She advocates the separation between a semantic and a
syntactic module on the basis of the processing event structural
variations. She holds that utterances, in which semantic meaning is
syntactically transparent are more easily processed than those which
are compositionally enriched and thus have to be aspectually coerced
into a derived interpretation. This perspective is rejected by
JOHANNES DÖLLING. Rather than suggesting a coercion operation for
event structurally shifted expressions like John broke a cup for
weeks, he introduces a parameter which is obligatorily inserted into
the semantic representation of any verb complex. Since the parame-
ter is contextually filled, 'coercion' is reinterpreted as contextual en-
richment. The idea of enriching linguistic representations by
contextual and conceptual information is shared by MARKUS EGG
and KRISTINA STRIEGNITZ. However, in the formal realization of this
mutual understanding the two approaches differ. For one thing in the
26 Heike Tappe and Holden Härtl

NLG conception of Egg and Striegnitz a context-sensitive type coer-

cion operator (TC) is added to the linguistic representation only in
specific cases, namely in order to derive a well-formed syntactic
structure for expressions containing sortally coerced verb arguments,
e.g. bottle in Every bottle froze.

The mediating function of the lexicon. The lexicon is the system

where information is stored of how to relate preverbal and linguistic
structures in an economic way such that the different communicative
requirements accompanying the speech act can be met. Here, a func-
tional perspective is adopted by HEIDRUN DORGELOH and ANJA
WANNER. The authors demonstrate that the internal structure of
event concepts and their lexical argument structure, respectively, can
be made use of to meet register specific requirements. They illustrate
how the expression of certain types of events in research articles re-
late to the degrees of implicitness text producers ascribe to agentive
entities. Lexical principles controlling the derivation of nominaliza-
tions from different types of verbs are discussed by VERONIKA
EHRICH. She shows how different event structural verb types relate
to the argument structural behavior of the corresponding nominaliza-
tions. While she acknowledges that the interpretation of event nomi-
nalizations draws on conceptual knowledge, she insists that the
nominal linking rules interfacing syntax and semantics are rooted in
the grammatical system, i.e. the lexicon. Lexicon internal event en-
coding principles are treated by ANDREA C . SCHALLEY, who shares
the aspect of language comparison with Ralf Nüse. By exploring
data from Walmajarri, Kalam, and German she identifies two com-
peting lexical principles, which are derived from the language spe-
cific chunking of event concepts and determine the grammatical
alternatives of coding complex events. In the context of motion verbs
LADINA B. TSCHANDER investigates the alternation between particle
verb constructions versus prepositional phrase constructions. She
holds that conceptual conditions associated with motion and path
concepts regulate the realization of the corresponding verb com-
plexes. Thereby she accounts for the requirement that goal concepts
need to be specified in certain contexts during language production
 Mediating between concepts and language - Processing structures 27

and shows which lexical properties can adequately realize the corre-
sponding conditions.

Notes

This volume is the outcome of the workshop The Syntax-Semantics-Interface:

Linguistic Structures and Processes at the DGfS conference Language and
Cognition in March 2001. The editors' work on this volume has been com-
pleted within the projects Conceptualization processes in language production
(HA 1237-10) and Conceptual transfer of situations into verbal meaning and
the status of thematic roles (OL 101-2) of the DFG priority program Language
production and the project Semantic interfaces: copula-predicative construc-
tions at ZAS (Berlin). For constructive comments we wish thank Susan Olsen
and we are grateful for the valuable suggestions for improvement that fol-
lowed from the anonymous review process. For their competent support in the
technical realization we are emdebted to Britta Gömy, Delia Herrn, and Tho-
mas Schulz. Many thanks go to the team of Mouton De Gruyter who were ef-
ficient and helpful.
1. There exists a vast body of empirical evidence that e.g. many perceptional
processes, e.g. in visual perception, are largely autonomous of other cognitive
processes (Pylyshyn 1999).
2. The remaining three criteria relate to the biological prerequisites of modules
and Fodor holds them to be important for discerning module-generated from
learned behavior: Modules are localized, i.e. mediated by dedicated neural
structures. They obey ontogenetic and pathological universals in that they
both mature and decay in distinctive sequences.
3. Compare e.g. Frazier (1987) for a strictly modular, and e.g. Bates (1994) for a
non-modular view.
4. The syntax of a given language is semantically and pragmatically arbitrary.
For example, there are no compelling arguments outside syntax for verb-end
position in German subordinate clauses.
5. A completely different viewpoint is presented by Elman et al. (1996) and
Marslen-Wilson & Tyler (1987).
6. Here, theories of a generative character like the Government and Binding The-
ory (Chomsky 1981) focus almost exclusively on the representation of argu-
ment structure, while there is no consensus on which kind of lexical
information is to be included.
7. This conviction is shared by theoreticians outside the linking theoretical
framework (e.g. Pustejovky (1992).
8. See Härtl (2003) for discussion.
28 Heike Tappe and Holden Härtl

9. Friederici (1997) discusses corresponding neuro-psychological implications of

the assumption that meaning construction is achieved in two steps in language
processing.
10. This process has most prominently been accounted for in the Recognition by
Components or Geon Theory (cf. Biederman 1995). It posits that objects and
scenes are represented as an arrangement of simple, viewpoint-invariant
volumetric primitives (e.g. bricks, cylinders, wedges, and cones) termed geons
that are recognizable even if parts are occluded. Geon theory has been exten-
sively tested and can elegantly account for the fact that objects become hardly
recognizable when viewed from a highly unfamiliar perspective. A leading al-
ternative view to recognition by components is proposed by View-Based
Recognition approaches (cf. e.g. Tarr & Bülthoff 1995).
11. These processes can be characterized as follows: Segmentation of states of
affairs is the distinction of those entities that are relevant within a current con-
ceptualization, especially temporal and spatial segmentation. Structuring of
states of affairs leads to the construction of hierarchical event structures. Se-
lection singles out the subclass of available entities that are to be verbalized
(cf. Habel & Tappe 1999).
12. Following Härtl (2001: 109) we assume that during language production the
first component of the language production system, the so called conceptual-
izer, has access to the currently activated information from both the semantic
and the episodic knowledge base. Thus concrete episodic information (includ-
ing temporal and spatial specifications) can be linked to global information
about abstract event types (including abstract temporal and spatial structures)
13. Here we find a striking analogy to computational language production models:
Computer linguists have so far been forced to content themselves with very re-
stricted domains in order to build running systems, in which a coupling be-
tween the to-be-verbalized contents and language output can be guaranteed.
14. Cf. Pechmann (in print) for an overview of experimental methods in language
production research.
15. Similar trends are witnessed in linguistics, e.g. in conceptions of the genera-
tive lexicon (Pustejovsky 1995) and also in the minimalist program (Chomsky
1995,2000).

References
Aduciewicz, Kazimierz
1935 Die syntaktische Konnexität. Studio Philosophica 1, 1-27.
 Mediating between concepts and language - Processing structures 29

Anderson, John R.
1991 The adaptive nature of human categorization. Psychological Review
98: 409-429.
Baker, Mark
1997 Thematic roles and syntactic structure. In: Liliane Haegeman (ed.),
Elements of Grammar, 73-137. Dordrecht, Kluwer..
Barsalou, Lawrence W., Wenchi Yeh, Barbara J. Luka, Karen L. Olseth, Kelly S.
Mix and Ling-Ling Wu
1993 Concepts and meaning. In: Katharine Beals, Gina Cooke, David
Kathman, Karl-Erik McCullough, Sotaro Kita and David Testen (eds.),
Chicago Linguistics Society 29: Papers from the parasession on con-
ceptual representations, 23-61. University of Chicago: Chicago Lin-
guistics Society.
Barwise, Jon and John Perry
1983 Situations and Attitudes. Cambridge, MA: MIT Press.
Bates, Elizabeth
1994 Modularity, domain specificity and the development of language. In:
Discussions in Neuroscience 10: 136-149.
Bates, Elizabeth and Brian MacWhinney
1984 Cross-linguistic Studies of Sentence Processing. Cambridge:
Cambridge University Press.
Bäuerle, Rainer
1985 Das Lexikon in der kompositionalen Semantik. In: Christoph Schwar-
ze and Dieter Wunderlich (eds.), Handbuch der Lexikologie, 199-229.
Königstein, Ts.: Athenäum.
Berwick, Robert C. and Amy S. Weinberg
1984 The Grammatical Basis of Linguistic Performance. Cambridge, MA:
MIT Press.
Β ickerton, Derek
1990 Language and Species. Chicago: University of Chicago Press.
Biederman, Irving
1995 Visual object recognition. In: Stephen M. Kosslyn and Daniel N.
Osherson (eds.), An Invitation to Cognitive Science, 2nd edition, Vol.
2, Visual Cognition, 121-165. Cambridge, MA: MIT Press.
Bierwisch, Manfred
1983 Semantische und konzeptionelle Repräsentationen lexikalischer Ein-
heiten. In: Rudolf Ruzicka and Wolfgang Mötsch (eds.), Untersuchun-
gen zur Semantik (Studia Grammatika 22), 61-89. Berlin: Akademie-
Verlag.
1986 On the nature of semantic form in natural language. In: Friedhart Klix
and Herbert Hagendorf (eds.), Human Memory and Cognitive Capa-
bilities - Mechanisms and Performances, 765-784. Vol. B. Amster-
dam: Elsevier.
30 Heike Tappe and Holden Härtl

Bierwisch, Manfred and Rob Schreuder

1992 From concepts to lexical items. Cognition 4: 23-60.
Bloom, Paul
1998 Theories of artifact categorization. Cognition 66: 87-93.
Brody, Michael and M. Rita Manzini
1988 On implicit arguments. In: Ruth Kempson (ed.), Mental
Representations: The Interfaces between Language and Reality, 105-
130. Cambridge: Cambridge University Press.
Busemann, Stephan
1993 A holistic view of lexical choice. In: Helmut Horacek (ed.), New
Concepts in Natural Language Generation: Planning, Realization, and
Systems, 302-308. London, New York: Frances Pinter.
Caramazza, Alfonso
1997 Brain and language. In: Michael S. Gazzaniga (ed.), Conversations in
the Cognitive Neurosciences, 131-151. Cambridge, MA: MIT Press.
Chomsky, Noam
1981 Lectures on Government and Binding. Dordrecht: Foris Publications.
1986 Knowledge of Language: Its Nature, Origin and Use. New York:
Praeger.
1993 A minimalist program for linguistic theory. In: Kenneth Hale and
Samuel J. Keyser (eds.), The view from Building 20, 1-52. Cambridge,
MA: MIT Press.
1995 The Minimalist Program . Cambridge, MA: MIT Press.
2000 Minimalist inquiries: The framework. In: Roger Martin, David
Michaels and Juan Uriagereka (eds.), Step by Step. Cambridge, MA:
MIT Press.
Cresswell, Maxwell J.
1978 Semantic competence. In: Franz Guenthner and M. Guenthner-Reutter
(eds.), Meaning and Translation, 9-27. London: Duckworth.
De Smedt, Koenraad, Helmut Horacek, and Michael Zock
1996 Architectures for natural language generation: Problems and perspec-
tives. In: Giovanni Adorni and Michael Zock (eds.), Trends in Natural
Language Generation: An Artificial Intelligence Perspective (Springer
lecture notes in artificial intelligence 1036), 17-46. Berlin: Springer.
Deane, Paul D.
1992 Grammar in Mind and Brain. Berlin, New York: Mouton de Gruyter.
Dölling, Johannes
1998 Semantic form and abductive fixation of parameters. Ms. Universität
Leipzig.
Dowty, David R.
1979 Word Meaning and Montague Grammar. Reidel, Dordrecht.
 Mediating between concepts and language - Processing structures 31

Ehrich, Veronika
1992 Hier undjetzt: Studien zur lokalen und temporalen Deixis im
Deutschen. Tübingen: Niemeyer
Elman, Jeffrey L., Elizabeth A. Bates, Mark H. Johnson, Annette Karmiloff-Smith,
Domenico Parisi and Kim Plunkett
1996 Rethinking Innateness. MIT Press.
Fodor, Jerry A.
1983 The Modularity of Mind. An Essay on Faculty Psychology. Cambridge,
MA: MIT Press
1998 The trouble with psychological Darwinism. London review of books.
Vol. 20, 2. [http://www.lrb.co.uk/v20n02/fodo2002.html]
Frazier, Lyn
1987a Sentence processing: A tutorial review. In: Max Coltheart (ed.),
Attention and Performance XII: The Psychology of Reading. Hillsdale:
Lawrence Erlbaum.
1987b Theorie of sentence processing. In: Jay L. Garfield (ed.), Modularity in
Knowledge Representation and Natural Language Understanding.
Cambridge, MA: MIT Press.
Frazier, Lyn, Charles Clifton and Janet Randall
1983 Filling gaps: Decision principles and structure in sentence comprehen-
sion. Cognition 13: 187-222.
Friederici, Angela D.
1997 Neurophysiological aspects of language processing. Clinical
Neuroscience 4: 64-72.
Gärdenfors, Peter
2000 Conceptual Spaces. Bradford Books. Cambridge, MA: MIT Press
Gazdar, Gerald, Ewan Klein, Geoffrey Pullum and Ivan Sag
1985 Generalized Phrase Structure Grammar. Oxford: Basil Blackwell.
Grimshaw, Jane
1979 Complement selection and the lexicon. Linguistic Inquiry 10: 279-
326.
1990 Argument Structure. Cambridge, MA: MIT Press.
1993 Semantic structure and semantic content in lexical representation. Un-
published ms., Rutgers University, New Brunswick, NJ.
Grimshaw, Jane and Sten Vikner
1993 Obligatory adjuncts and the structure of events. In: Eric Reuland and
Werner Abraham (eds.), Knowledge and Language: Vol. II: Lexical
and Conceptual Structure, 145-159. Dordrecht: Kluwer.
Habel, Christopher and Heike Tappe
1999 Processes of segmentation and linearization in describing events. In:
Ralf Klabunde and Christiane v. Stutterheim (eds.), Representations
and processes in language production, 117-153. Wiesbaden: Deut-
scher Universitätsverlag.
32 Heike Tappe and Holden Härtl

Härtl, Holden
2001 CA USE und CHANGE: Thematische Relationen und Ereignisstruktu-
ren in Konzeptualisierung und Grammatikalisierung. (Studia Gram-
matica 50). Berlin: Akademie-Verlag.
2003 Conceptual and grammatical characteristics of argument alternations:
The case of decausative verbs. Appears in: Linguistics 41(5).
Jackendoff, Ray
1972 Semantic Interpretation in Generative Grammar. Cambridge, MA:
MIT Press.
1990 Semantic Structures. Cambridge, MA: MIT Press
1992 Languages of the Mind. Cambridge, MA: MIT Press.
1997 The Architecture of the Language Faculty. Cambridge, MA: MIT
Press.
Jescheniak, Jörg D. and Willem J. M. Levelt
1994 Word frequency in speech production: Retrieval of syntactic informa-
tion and phonological form. Journal of Experimental Psychology:
Learning, Memory and Cognition 20: 824—843.
Kaplan, David
1977 The logic of demonstratives. Journal of Philosophical Logic 8: 81-98.
Kaplan, Ronald M. and Joan Bresnan
1982 Lexical-functional grammar: A formal system for grammatical repre-
sentation. In: Joan Bresnan (ed.), The Mental Representation of
Grammatical Relations. Cambridge, MA: MIT Press.
Karmiloff-Smith, Annette
1992 Beyond Modularity: A Developmental Perspective on Cognitive Sci-
ence. Cambridge, MA: MIT Press.
Katz, Jerrold J. and Jerry A. Fodor
1963 The structure of a semantic theory. Language 39: 170-210.
Koenig, Jean-Pierre and Gail Mauner
1999 A-definites and the semantics of implicit arguments. Journal of
Semantics 16: 207-236.
Lakoff, George
1991 Metaphor and war: The metaphor system used to justify war in the
Gulf. Distributed via electronic bulletin boards. Journal of Urban and
Cultural Studies 2.
Lang, Ewald
1994 Semantische und konzeptuelle Struktur: Unterschneidung und Über-
schneidung. In: Monika Schwarz (ed.), Kognitive Semantik: Ergebnis-
se, Probleme, Perspektiven, 25-41. Tübingen: Narr.
Langacker, Ronald W.
1987 Foundations of Cognitive Grammar I: Theoretical Prerequisites. Stan-
ford: Stanford University Press.
 Mediating between concepts and language - Processing structures 33

1990 Settings, participants, and grammatical relations. In: S. Tsohatzidis

(ed.), Meanings and Prototypes: Studies in Linguistic Categorization,
213-238. New York: Routledge, Chapman and Hall.
1991 Foundations of Cognitive Grammar II: Descriptive Applications.
Stanford: Stanford University Press.
Levelt, Willem J. M.
1989 Speaking: From Intention to Articulation. Cambridge, MA, London,
UK: MIT Press.
Levelt, Willem J. M., Ardi Roelofs and Antje S. Meyer
1999 A theory of lexical access in speech production. Behavioral and Brain
Sciences 22: 1-75.
Levinson, Stephen C.
1996 Frames of reference and Molyneaux's question: Cross-linguistic evi-
dence. In: Paul Bloom, Mary A. Peterson, Lynn Nadel and Merrill F.
Garrett (eds.), Language and Space, 109-169. MIT Press.
Levin, Beth and Malka Rappaport Hovav
1995 Unaccusativity: At the Syntax-Lexical Semantics Interface. (Linguistic
Inquiry Monograph 26). Cambridge, MA: MIT Press.
Lewis, David
1972 Psychophysical and theoretical identifications. Australasian Journal of
Philosphy 50: 249-258.
Marcus, Mitchell P.
1982 Consequences of functional deficits in a parsing model: Implications
for Broca's aphasia. In: Michael A. Arbib, David Caplan, John C. Mar-
shall (eds.), Neural Models of Language Processes (Perspectives in
Neurolinguistics, Neuropsychology, and Psycholinguistics), 115-134.
New York: Academic Press.
Marshall, John C.
1984 Multiple perspectives on modularity. Cognition 17: 209-242.
Marslen-Wilson, William D. and Lorraine K. Tyler
1987 Against modularity. In: Jay L. Garfield (ed.), Modularity in
Knowledge Representation and Natural-Language Understanding.
Cambridge, MA: MIT Press.
McCawley, James D.
1971 Tense and time reference in English. In: Charles Fillmore and D.
Terence Langendoen (eds.), Studies in Linguistic Semantics, 97-113.
New York: Holt Rinehart.
McShane, Marjorie and Ron Zacharski
2001 Modularity in knowledge elicitation and language processing. Pro-
ceedings of the Third Annual High Desert Linguistics Conference.
University of New Mexico, Albuquerque NM, April 7-9, 2000. 93-
104.
34 Heike Tappe and Holden Härtl

Medin, Douglas L.
1989 Concepts and conceptual structure. American Psychologist 44: 1469-
1481.
Meisel, Jürgen M. (ed.)
1990 Two First Languages: Early Grammatical Development in Bilingual
Children. Dordrecht: Foris.
Mohanan, Tara and Lionel Wee (eds.)
1999 Grammatical Semantics: Evidence for Structure in Meaning. Stanford,
CA: CSLI Lecture Notes.
Montague, Richard
1970 Universal grammar. Theoria 36, 373-398.
Montague, Richard
1973 The proper treatment of quantification in ordinary English. In: Jaakko
Hintikka, Julius M. Moravcsik and Patrick Suppes (eds.), Approaches
to Natural-Language: Proceedings of the 1970 Stanford Workshop on
Grammar and Semantics, 221-242. Dordrecht: Reidel. [Reprinted in
Montague (1974). 247-270.].
Müller, Ralph-Axel
1996 Innateness, autonomy, universality? Neurobiological approaches to
language. Behavioral and Brian Sciences 19: 611-675.
Newmeyer, Frederick J.
1998 Language Form and Language Function. Cambridge, MA: MIT Press.
O'Bryan, Erin, Raffaella Folli, Heidi Harley and Thomas Bever
in prep. The Role of Event Structure in Language Comprehension,
http://www.ling-phil.ox.ac.uk/research/folli/research_files/amlap.doc
Olsen, Susan
1998 Prädikative Argumente syntaktischer und lexikalischer Köpfe: Parti-
kelverben im Deutschen und Englischen. Folia Linguistica 31: 301—
329.
Partee, Barbara
1975 Montague grammar and transformational grammar. Linguistic Inquiry
6(2): 203-300.
Pechmann, Thomas
in print Experimentelle Methoden in der Sprachproduktionsforschung. In:
Theo Herrmann and Joachim Grabowski (eds.), Enzyklopädie der Psy-
chologie, Band C III/1: Sprachproduktion. Göttingen: Verlag für Psy-
chologie.
Pollard, Carl and Ivan Sag
1987 An Information-based Approach to Syntax and Semantics: Vol. 1 Fun-
damentals. Stanford, CA: CSLI Lecture Notes.
Pustejovsky, James
1995 The Generative Lexicon. Cambridge, MA: MIT Press.
 Mediating between concepts and language - Processing structures 35

Pylyshyn, Zenon W.
1999 Is vision continuous with cognition? The case for cognitive
impenetrability of visual perception. Behavioral and Brain Sciences
22(3): 341—423.
Rappaport, Malka and Beth Levin
1988 What to do with theta-roles. In: Wendy Wilkins (ed.), Syntax and
Semantics 21: Thematic Relations, 7-36. New York, NY: Academic
Press.
Roeper, Thomas
1987 Implicit arguments and the head-complement relation. Linguistic
Inquiry 18: 267-310.
Saussure, Ferdinand de
1916 Cours de linguistique generale. In: Charles Bally and Albert
Sechehaye (eds.), Paris: Payot.
Smolensky, Paul
1988 On the proper treatment of connectionism. Behavioral and Brain Sci-
ences 11: 1-74.
Stillings, Neil Α.; Steven E. Weisler, Christopher H. Chase, Mark H. Feinstein, Jay
L. Garfield and Edwina L. Rissland
1998 Cognitive Science: An Introduction (2nd ed.). Cambridge, MA: MIT
Press.
Stromswold, Karin
1992 Leamability and the acquisition of auxiliary and copula be. In:
ESCOL '91. Columbus State University.
Tappe, Heike
1999 Der Spracherwerb bei Corpus-Callosum-Agenesie: Eine explorative
Studie. Tübingen: Narr.
2000 Perspektivenwahl in Beschreibungen dynamischer und statischer We-
geskizzen. In: Christopher Habel and Christiane v. Stutterheim (eds.),
Räumliche Konzepte und sprachliche Strukturen, 69-97. Tübingen:
Niemeyer.
Tarsky, Alfred
1977 Einföhrung in die Mathematische Logik. Göttingen: Vandenhoeck and
Ruprecht.
Tarr, Michael J. and Heinrich H. Bülthoff
1995 Is human object recognition better described by geon structural de-
scriptions or by multiple views? Comment on Biederman and
Gerhardstein (1993). Journal of Experimental Psychology: Human
Perception and Performance. 21(6): 1494-1505.
Taylor, Holly A. and Barbara Tversky
1996 Perspective in spatial descriptions. Journal of Memory and Language
35,371-391.
36 Heike Tappe and Holden Härtl

Tenny, Carol
1992 The Aspectual Interface Hypothesis. In: Ivan A. Sag and Anna
Szabolcsi (eds.), Lexical Matters (CSLI lecture notes 24), 1-27. Palo
Alto: Stanford University Press.
1994 Aspectual Roles and the Syntax-Semantics Interface. (Studies in Lin-
guistics and Philosophy 52). Dordrecht: Kluwer.
Tomasello, Michael
2000a Do young children have adult syntactic competence? Cognition 74:
209-253.
2000b The item based nature of children's early syntactic development.
Trends in Cognitive Sciences 4: 156-163.
van Hout, Angel iek
1996 Event semantics of verb frame alternations: A case study of Dutch and
its acquisition. Ph.D. dissertation, Tilburg University.
Wilder, Chris and Damir Cavar
1994 Long head movement? Verb movement and cliticzation in Croatian.
Lingua 93: 1-58
Wunderlich, Dieter
1997 CAUSE and the structure of verbs. Linguistic Inquiry 28(1): 27-68.
Zacks, Jeffrey M.
1997 Seeing the structure in events. Ms., Stanford University.
Zeevat, Henk, Ewan Klein and Jo Calder
1987 Unification categorial grammar. In: Nicholas Haddock, Ewan Klein
and Glyn Morrill (eds.), Categorial Grammar, Unification Grammar,
and Parsing. University of Edinburgh, Scotland.
Mediating between
non-linguistic and linguistic structures
Coordination of eye gaze and speech in sentence
production

Femke F. van der Meulen

1. Introduction

In recent years, experiments on eye movements and object naming

demonstrated a link between looking at an object and linguistically
processing the object's name. Speakers tend to look at the objects
they are about to find words for in the same order in which the object
names were mentioned in the utterance. They not only looked in or-
der to recognize an object, but they kept looking until they had proc-
essed the object's appropriate name up until the level of
phonological encoding (Meyer, Sleiderink and Levelt 1998; Meyer
and van der Meulen 2000). When pronouns where used instead of
noun phrases to describe action scenes or repeated objects, speakers
looked less frequent and more briefly at the objects they referred to
than when noun phrases were used (van der Meulen and Meyer
2001). These results confirmed the link between looking and nam-
ing.
Another important result followed from an experiment, in which
speakers named two objects and, in addition, two properties of the
first one. In different blocks, speakers used different utterance types:
"The large, red ball is next to the mouse" or "The ball, next to the
mouse, is large and red". In the first utterance type, speakers kept
their eyes on the large red ball for a very long time, until right before
they produced the word "mouse". Interestingly, in the second utter-
ance type, where the adjectives were named later in the sentence,
speakers moved their eyes from /ball/ to /mouse/ and back to /ball/,
with a tight alignment to the produced speech: They returned their
gaze to the first object right before they started to name the adjec-
tives. Even though one might assume that speakers have taken in the
40 Femke F. van der Meulen

conceptual information regarding color and size of an object during

the first gaze, they apparently prefer to allocate their visual attention
to the information on the screen that is to be verbalized (van der
Meulen 2001).
In all experiments, speakers looked at the objects and sometimes
returned their gaze to them in the same order of subsequent naming.
This indicates that speakers preferred to view each object and proc-
ess each object's name in serial order. However, in all of these ex-
periments speakers were told which utterance structure they should
use. Speakers were therefore able to put the object names in prede-
fined syntactic structures, specifying the order of fixation even be-
fore a picture appeared. The processing of the first part of the
utterance was allowed to start without any delay or any kind of vis-
ual overview of the complete scene. The participants in the experi-
ments were likely to create a looking order strategy that enabled
them to work through each experimental trial as fast and as effi-
ciently as possible.
When, as in the experiments describes above, the speakers al-
ready have a sentence structure in mind, it can safely be assumed
that they view the objects to recognize them and then activate lexical
concepts. This is called conceptual preparation, and it includes a
decision on how to name a specific object in a specific situation.
When the appropriate lexical concept is found, it gives access to its
lemma and word form (Levelt, Roelofs and Meyer 1999). In every-
day language use, a lexical concept is often activated as part of a lar-
ger message that captures the speaker's communicative intention
(Levelt 1989). The order of words within an utterance is (in part)
determined by this intention. When the experimenter takes this deci-
sion, the speaker does not have to include this high level processing.
A related study I know of in which the speakers were not in-
structed to use a pre-described sentence structure, was an eye gaze
study by Griffin and Bock (2000). Speakers viewed and spontane-
ously described simple action events while their eye movements
were monitored. The cognitive processing necessary to understand
the action scene and planning an appropriate sentence structure was
thereby added to speaking processes. Four groups of subjects par-
 Coordination of eye gaze and speech in sentence production 41

ticipated in four different tasks: free viewing, scene comprehension,

preparation of a sentence to be spoken later and description of the
scene online. Comparison of the subjects' eye movements between
the groups showed that in the online speaking task, speakers began
with an effort to comprehend the scene and then fixated the partici-
pants in the event in the same order in which they were subsequently
named.
The present object naming experiment used on the one hand a
fixed utterance situation as in the earlier experiments. On the other
hand, a more variable situation was introduced, in which speakers,
like in the Griffin and Bock (2000) experiment, needed to retrieve
some visual information from the picture, before being able to start
speaking an appropriate utterance.
Speakers had to name four objects presented on the screen in a
fluent utterance. The bottom objects were either identical or differ-
ent. When they were identical, speakers had to use a conjoined NP
structure to describe the picture: The fork and the pen are above a
cup. When the bottom two objects were different, a conjoined clause
structure was to be produced: The fork is above a cup and the pen is
above a key. Presentation of the pictures took place in four blocks. In
one of those blocks, all pictures had identical bottom objects and
therefore a conjoined NP structure was required for each picture. In
another block, all pictures had bottom objects that differed from each
other, thereby requiring a conjoined clause structure. These two
blocks were labeled fixed blocks. In the other two blocks, pictures
with different and identical bottom objects were mixed, creating
variable blocks. In these variable blocks, speakers needed to com-
pare the bottom objects to decide on the appropriate utterance struc-
ture, before being able to start that utterance. Therefore, visual
attention to the bottom objects was necessary. Records of eye
movements were used to compare gaze patterns between the variable
and fixed production situation.
Based on the study by Griffin and Bock (2000), it was expected
that in the variable condition speakers would scan the objects (apre-
view), decide which utterance structure was appropriate, and go back
to look at each object in the order of mention (a main pass).
42 Femke F. van der Meuten

A basic interest concerns the order of gaze in this main pass. Will
the speakers indeed look at all objects when naming them, after they
have seen them already?

2. Method

2.1. Participants

Eight speakers participated in the experiment. They were under-

graduate students of Nijmegen University, native speakers of Dutch
and had normal or corrected-to-normal vision. They were paid for
participation.

2.2. Materials and design

Top screen pictures: 48 line drawings of common objects with mono

or bisyllabic names were selected from a picture pool available at the
Max Planck Institute for Psycholinguistics, Nijmegen, The Nether-
lands (MPI), to appear on the top half of the screen. They were
paired, resulting in 12 pairs of monosyllabic and 12 pairs of bisyl-
labic names.
Bottom screen pictures: 24 line drawings of common objects with
monosyllabic names were paired, resulting in twelve pairs of objects.
A complete list of the materials is presented in the Appendix.
Each pair of bottom objects was presented twice, once with
monosyllabic and once with bisyllabic top-objects, creating 24 object
scenes. These 24 scenes were presented twice as well, resulting in 48
target items.1
In the conjoined clause condition, the 48 target items were used.
In the conjoined NP condition, the right bottom object was replaced
with a copy of the left bottom one, resulting in two identical objects
on the bottom half of the screen. Figures la and lb show examples
of the items.
Exploring the Variety of Random
Documents with Different Content
The Project Gutenberg eBook of Pride and His
Prisoners
 This ebook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this ebook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.

Title: Pride and His Prisoners

Author: A. L. O. E.

Illustrator: T. W. Holgate

Release date: August 21, 2019 [eBook #60149]

Language: English

Credits: Produced by Richard Hulse and the Online Distributed

Proofreading Team at http://www.pgdp.net (This file
was
produced from images generously made available by
The
Internet Archive)

*** START OF THE PROJECT GUTENBERG EBOOK PRIDE AND HIS

PRISONERS ***
PRIDE AND HIS PRISONERS
A Terrible Danger.

Page 230.
Pride and his
prisoners BY
A. L. O. E.
LONDON, EDINBURGH,
AND NEW YORK
THOMAS NELSON
AND SONS
 CONTENTS
I. The Haunted Dwelling 5
II. Resisted, yet Returning 16
III. Snares 26
IV. A Glance into the Cottage 33
V. Both Sides 43
VI. The Visit to the Hall 51
VII. A Misadventure 60
VIII. A Brother’s Effort 75
IX. Disappointment 88
X. On the Watch 96
XI. The Quarrel 102
XII. The Unexpected Guest 111
XIII. The Friend’s Mission 119
XIV. A Fatal Step 128
XV. The Deserted Home 140
XVI. Pleading 147
XVII. Conscience Asleep 157
XVIII. The Magazine 162
XIX. Expectation 170
XX. A Sunny Morn 178
XXI. The Ascent 187
XXII. In the Clouds 193
XXIII. Regrets 201
XXIV. Soaring above Pride 208
XXV. A Broken Chain 217
XXVI. The Awful Crisis 222
XXVII. Tidings 234
XXVIII. The Wheel Turns 242
XXIX. Two Words 252
XXX. The Spirit Laid 263
 LIST OF ILLUSTRATIONS
A Terrible Danger Frontispiece
Her greeting to Dr. and Miss Bardon was most
gracious and cordial 57
Tearing the Manuscript 107
An Unwelcome Surprise 168

PRIDE AND HIS PRISONERS.

 CHAPTER I.
THE HAUNTED DWELLING.

“He who envies now thy state,

Who now is plotting how he may seduce
Thee also from obedience; that with him,
Bereaved of happiness, thou mayst partake
His punishment,—eternal misery!”

Milton.

Bright and joyous was the aspect of nature on a spring morning in

the beautiful county of Somersetshire. The budding green on the
trees was yet so light, that, like a transparent veil, it showed the
outlines of every twig; but on the lowlier hedges it lay like a rich
mantle of foliage, and clusters of primroses nestled below, while the
air was perfumed with violets. Already was heard the hum of some
adventurous bee in search of early sweets, the distant low of cattle
from the pasture, the mellow note of the cuckoo from the grove,—
every sight and sound told of enjoyment on that sunny Sabbath
morn.
Yet let me make an exception. There was one spot which reserved
to itself the unenviable privilege of looking gloomy all the year
round. Nettleby Tower, a venerable edifice, stood on the highest
summit of a hill, like some stern guardian of the fair country that
smiled around it. The tower had been raised in the time of the
Normans, and had then been the robber-hold of a succession of
fierce barons, who, from their strong position, had defied the power
of king or law. The iron age had passed away. The moat had been
dried, and the useless portcullis had rusted over the gate. The loop-
holes, whence archers had pointed their shafts, were half filled up
with the rubbish accumulated by time. Lichens had mantled the grey
stone till its original hue was almost undistinguishable; silent and
deserted was the courtyard which had so often echoed to the clatter
of hoofs, or the ringing clank of armour.
Silent and deserted—yes! It was not time alone that had wrought
the desolation. Nettleby Tower had stood a siege in the time of the
Commonwealth, and the marks of bullets might still be traced on its
walls; but the injuries which had been inflicted by the slow march of
centuries, or the more rapid visitation of war, were slight compared
to those which had been wrought by litigation and family dissension.
The property had been for years the subject of a vexatious lawsuit,
which had half ruined the unsuccessful party, and the present owner
of Nettleby Tower had not cared to take personal possession of the
gloomy pile. Perhaps Mr. Auger knew that the feeling of the
neighbourhood would be against him, as the sympathies of all would
be enlisted on the side of the descendant of that ancient family
which had for centuries dwelt in the Tower, who had been deprived
of his birthright by the will of a proud and intemperate father.
The old fortress had thus been suffered to fall into decay. Grass
grew in the courtyard; the wallflower clung to the battlements; the
winter snow and the summer rain made their way through the
broken casements, and no hand had removed the mass of wreck
which lay where a furious storm had thrown down one of the ancient
chimneys. Parties of tourists occasionally visited the gloomy place,
trod the long, dreary corridors, and heard from a wrinkled woman
accounts of the moth-eaten tapestry, and the time-darkened family
portraits that grimly frowned from the walls. They heard tales of the
last Mr. Bardon, the proud owner of the pile; how he had been wont
to sit long and late over his bottle, carousing with jovial companions,
till the hall resounded with their oaths and their songs; and how,
more than thirty years back, he had disinherited his only son for
marrying a farmer’s daughter. Then the old woman would, after
slowly showing the way up the worn stone steps which led round
and round till they opened on the summit of the tower, direct her
listener’s attention to a small grey speck in the wide-spreading
landscape below, and tell them that Dr. Bardon lived there in needy
circumstances, in actual sight of the place where, if every man had
his right, he would now be dwelling as his fathers had dwelt. And
the visitors would sigh, shake their heads, and moralize on the
strange changes in human fortunes.
The old woman who showed strangers over Nettleby Tower lived
in a cottage hard by; neither she nor any other person was ever to
be found in the old halls after the sun had set. The place had the
repute of being haunted, and was left after dark to the sole
possession of the rooks, the owls, and the bats. I must tax the faith
of my readers to believe that the old tower was actually haunted;
not by the ghosts of the dead, but by the spirits of evil that are ever
moving amongst the living. I must attempt with a bold hand to draw
aside the mysterious veil which divides the invisible from the visible
world, and though I must invoke imagination to my aid, it is
imagination fluttering on the confines of truth. Bear with me, then,
while I personify the spirits of Pride and Intemperance, and
represent them as lingering yet in the pile in which for centuries they
had borne sway over human hearts.
Standing on the battlements of the grey tower, behold two dim,
but gigantic forms, like dark clouds, that to the eye of fancy have
assumed a mortal shape. The little rock-plant that has found a cradle
between the crumbling stones bends not beneath their weight,—and
yet how many deep-rooted hopes have they crushed! Their
unsubstantial shapes cast no shadow on the wall, and yet have
darkened myriads of homes! The natural sense cannot recognise
their presence; the eye beholds them not, the human ear cannot
catch the low thunder of their speech; and yet there they stand,
terrible realities,—known, like the invisible plague, by their effects
upon those whom they destroy!
There is a wild light in the eyes of Intemperance, not caught from
the glad sunbeams that are bathing the world in glory; it is like a red
meteor playing over some deep morass, and though there is often
mirth in his tone, it is such mirth as jars upon the shuddering soul
like the laugh of a raving maniac! Pride is of more lofty stature than
his companion, perhaps of yet darker hue, and his voice is lower and
deeper. His features are stamped with the impress of all that piety
abhors and conscience shrinks from, for we behold him without his
veil. Human infirmity may devise soft names for cherished sins, and
even invest them with a specious glory which deceives the dazzled
eye; but who could endure to see in all their bare deformity those
two arch soul-destroyers, Intemperance and Pride?
“Nay, it was I who wrought this ruin!” exclaimed the former,
stretching his shadowy hand over the desolated dwelling. “Think you
that had Hugh Bardon possessed his senses unclouded by my spell,
he would ever have driven forth from his home his own—his only
son?”
“Was it not I,” replied Pride, “who ever stood beside him, counting
up the long line of his ancestry, inflaming his soul with legends of
the past, making him look upon his own blood as something
different from that which flows in the veins of ordinary mortals, till
he learned to regard a union with one of lower rank as a crime
beyond forgiveness?”
“I,” cried Intemperance, “intoxicated his brain”—
“I,” interrupted Pride, “intoxicated his spirit. You fill your deep cup
with fermented beverage; the fermentation which I cause is within
the soul, and it varies according to the different natures that receive
it. There is the vinous fermentation, that which man calls high spirit,
and the world hails with applause, whether it sparkle up into
courage, or effervesce into hasty resentment. There is the acid
fermentation; the sourness of a spirit brooding over wrongs and
disappointments, irritated against its fellow-man, and regarding his
acts with suspicion. This the world views with a kind of
compassionate scorn, or perhaps tolerates as something that may
occasionally correct the insipidity of social intercourse. And there is
the third, the last stage of fermentation, when hating and hated of
all, wrapt up in his own self-worship, and poisoning the atmosphere
around with the exhalations of rebellion and unbelief, my slave
becomes, even to his fellow-bondsmen, an object of aversion and
disgust. Such was my power over the spirit of Hugh Bardon. I
quenched the parent’s yearning over his son; I kept watch even by
his bed of death; and when holy words of warning were spoken, I
made him turn a deaf ear to the charmer, and hardened his soul to
destruction!”
“I yield this point to you,” said Intemperance, “I grant that your
black badge was rivetted on the miserable Bardon even more firmly
than mine. And yet, what are your scattered conquests to those
which I hourly achieve! Do I not drive my thousands and tens of
thousands down the steep descent of folly, misery, disgrace, till they
perish in the gulf of ruin? Count the gin-palaces dedicated to me in
this professedly Christian land; are they not crowded with my
victims? Who can boast a power to injure that is to be compared to
mine?”
“Your power is great,” replied Pride, “but it is a power that has
limits, nay, limits that become narrower and narrower as civilization
and religion gain ground. You have been driven from many a stately
abode, where once Intemperance was a welcome guest, and have to
cower amongst the lowest of the low, and seek your slaves amongst
the vilest of the vile. Seest thou yon church,” continued Pride,
pointing to the spire of a small, but beautiful edifice, embowered
amongst elms and beeches; “hast thou ever dared so much as to
touch one clod of the turf on which falls the shadow of that
building?”
“It is, as you well know, forbidden ground,” replied Intemperance.
“To you—to you, but not to me!” exclaimed Pride, his form dilating
with exultation. “I enter it unseen with the worshippers, my voice
blends with the hymn of praise; nay, I sometimes mount the pulpit
with the preacher,[1] and while a rapt audience hang upon his
words, infuse my secret poison into his soul! When offerings are
collected for the poor, how much of the silver and the gold is
tarnished and tainted by my breath! The very monuments raised to
the dead often bear the print of my touch; I fix the escutcheon,
write the false epitaph, and hang my banner boldly even over the
Christian’s tomb!”
“Your power also has limits,” quoth Intemperance. “There is an
antidote in the inspired Book for every poison that you can instil.”
“I know it, I know it,” exclaimed Pride, “and marks it not the
extent of my influence and the depth of the deceptions that I
practise, that against no spirit, except that of Idolatry, are so many
warnings given in that Book as against the spirit of Pride? For every
denunciation against Intemperance, how many may be found
against me! Not only religion and morality are your mortal
opponents, but self-interest and self-respect unite to weaken the
might of Intemperance; I have but one foe that I fear, one that
singles me out for conflict! As David with his sling to Goliath, so to
Pride is the Spirit of the Gospel!”
“How is it, then,” inquired Intemperance, “that so many believers
in the Gospel fall under your sway?”
“It is because I have so many arts, such subtle devices, I can
change myself into so many different shapes; I steal in so softly that
I waken not the sentinel Conscience to give an alarm to the soul!
You throw one broad net into the sea where you see a shoal within
your reach; I angle for my prey with skill, hiding my hook with the
bait most suited to the taste of each of my victims. You pursue your
quarry openly before man; I dig the deep hidden pit-fall for mine.
You disgust even those whom you enslave; I assume forms that
rather please than offend. Sometimes I am ‘a pardonable weakness,’
sometimes ‘a natural instinct,’ sometimes,” and here Pride curled his
lip with a mocking smile, “I am welcomed as a generous virtue!”
“It is in this shape,” said Intemperance angrily, “that you have
sometimes even taken a part against me! You have taught my slaves
to despise and break from my yoke!”
“Pass over that,” replied Pride; “or balance against it the many
times when I have done you a service, encouraging men to be
mighty to mingle strong drink.”
“Nay, you must acknowledge,” said Intemperance, “that we now
seldom work together.”
“We have different spheres,” answered Pride. “You keep
multitudes from ever even attempting to enter the fold; I put my
manacles upon tens of thousands who deem that they already have
entered. I doubt whether there be one goodly dwelling amongst all
those that dot yonder wide prospect, where one, if not all of the
inmates, wears not my invisible band round the arm.”
“You will except the pastor’s, at least,” said Intemperance.
“Yonder, on the path that leads to the school, I see his gentle
daughter. She has warned many against me; and with her words,
her persuasions, her prayers, has driven me from more than one
home. I shrink from the glance of that soft, dark eye, as if it carried
the power of Ithuriel’s spear. Ida seems to me to be purity itself;
upon her, at least, you can have no hold.”
“Were we nearer,” laughed the malignant spirit, “you would see
my dark badge on the saint! Since her childhood I have been striving
and struggling to make Ida Aumerle my own. Sometimes she has
snapped my chain, and I am ofttimes in fear that she will break
away from my bondage for ever. But methinks I have a firm hold
over her now.”
“Her pride must be spiritual pride,” observed Intemperance.
“Not so,” replied his evil companion; “I tried that spell, but my
efforts failed. While with sweet voice and winning persuasion Ida is
now guiding her class to Truth, and warning her little flock against
us both, would you wish to hearken to the story of the maiden, and
hear all that I have done to win entrance into a heart which the
grace of God has cleansed?”
“Tell me her history,” said Intemperance; “she seems to me like
the snowdrop that lifts its head above the sod, pure as a flake from
the skies.”
 “Even the snowdrop has its roots in the earth,” was the sardonic
answer of Pride.

[1] “What a beautiful sermon you gave us to-day!” exclaimed a

lady to her pastor. “The devil told me the very same thing while I
was in the pulpit,” was his quaint, but comprehensive reply.
 CHAPTER II.
RESISTED, YET RETURNING.

“Mount up, for heaven is won by prayer;

Be sober—for thou art not there!”

Keble.

“The sacred pages of God’s own book

Shall be the spring, the eternal brook,
In whose holy mirror, night and day,
Thou’lt study heaven’s reflected ray.
And should the foes of virtue dare
With gloomy wing to seek thee there,
Thou will see how dark their shadows lie,
Between heaven and thee, and trembling fly.”

Moore.

“Ida Aumerle,” began the dark narrator, “at the age of twelve had
the misfortune to lose her mother, and was left, with a sister several
years younger than herself, to the sole care of a tender and
indulgent father. Ever on the watch to strengthen my interests
amongst the children of men, I sounded the dispositions of the
sisters, to know what chance I possessed of making them prisoners
of Pride. Mabel, clever, impulsive, fearless in character, with a mind
ready to receive every impression, and a spirit full of energy and
emulation, I knew to be one who was likely readily to come under
the power of my spell. Ida was less easily won; she was a more
thoughtful, contemplative girl, her temper was less quick, he
passions were less easily roused, and I long doubted where lay the
weak point of character on which Pride might successfully work.
“As Ida grew towards womanhood my doubts were gradually
dispelled. I marked that the fair maiden loved to linger opposite the
mirror which reflected her tall, slight, graceful form, and that the
gazelle eyes rested upon it with secret satisfaction. There was much
time given to braiding the hair and adorning the person; and the
fashion of a dress, the tint of a ribbon, became a subject for grave
consideration. There are thousands of girls enslaved by the pride of
beauty with far less cause than Ida Aumerle.”
“But this folly,” observed Intemperance, “was likely to give you but
temporary power. Beauty is merely skin-deep, and passes away like
a flower!”
“But often leaves the pride of it behind,” replied his companion.
“There is many a wrinkled woman who can never forget that she
once was fair,—nay, who seems fondly to imagine that she can never
cease to be fair; and who makes herself the laughing-stock of the
world by assuming in age the attire and graces of youth. It will never
be thus with Ida Aumerle.
“I thought that my chain was firmly fixed upon her, when one
evening I found it suddenly torn from her wrist, and trampled
beneath her feet! The household at the Vicarage had retired to rest;
Ida had received her father’s nightly blessing, and was sitting alone
in her own little room. The lamp-light fell upon a form and face that
might have been thought to excuse some pride, but Ida’s reflections
at that moment had nothing in common with me. She was bending
eagerly over that Book which condemns, and would destroy me,—a
book which she had ofttimes perused before, but never with the
earnest devotion which was then swelling her heart. Her hands were
clasped, her dark eyes swimming in joyful tears, and her lips
sometimes moved in prayer,—not cold, formal prayer, such as I
myself might prompt, but the outpouring of a spirit overflowing with
grateful love. That was the birthday of a soul! I stood gloomily
apart; I dared not approach one first conscious of her immortal
destiny, first communing in spirit with her God!”
“You gave up your designs, then, in despair?”
“You would have done so,” answered Pride with haughtiness; “I do
not despair, I only delay. I found that pride of beauty had indeed
given way to a nobler, more exalting feeling. Ida had drunk at the
fountain of purity, and the petty rill of personal vanity had become to
her insipid and distasteful. She was putting away the childish things
which amuse the frivolous soul. Ida’s time was now too well filled up
with a succession of pious and charitable occupations, to leave a
superfluous share to the toilette. The maiden’s dress became simple,
because the luxury which she now esteemed was that of assisting
the needy. Many of her trinkets were laid aside, not because she
deemed it a sin to wear them, but because her mind was engrossed
by higher things. One whose first object and desire is to please a
heavenly Master by performing angels’ offices below, is hardly likely
to dwell much on the consideration that her face and her figure are
comely.”
“Ida is, I know, reckoned a model of every feminine virtue,” said
Intemperance. “I can conceive that your grand design was now to
make her think herself as perfect as all the rest of the world thought
her.”
“Ay, ay; to involve her in spiritual pride! But the maiden was too
much on her knees, examined her own heart too closely, tried
herself by too lofty a standard for that. When the faintest shadow of
that temptation fell upon her, she started as though she had seen
the viper lurking under the flowers, and cast it from her with
abhorrence! ‘A sinner, a weak, helpless sinner, saved only by the
mercy, trusting only in the strength of a higher power;’ this Ida
Aumerle not only calls herself, but actually feels herself to be. The
power of Grace in her heart is too strong on that point for Pride.”
“And yet you hope to subject her to your sway?
 “About two years after the night which I have mentioned,”
resumed Pride, “after Ida had attained the age of eighteen, she
resided for some time at Aspendale, the home of her uncle,
Augustine Aumerle.”
“One of your prisoners?” inquired Intemperance.
“Of him anon,” replied the dark one, “our present subject is his
niece. At his dwelling Ida met with one who had been Augustine’s
college companion, Reginald, Earl of Dashleigh. You can just discern
the towers of his mansion faint in the blue distance yonder.”
“I know it,” replied Intemperance; “I frequented the place in his
grandfather’s time. The present earl, as I understand, is your votary
rather than mine.”
“Puffed up with pride of rank,” said the stern spirit; “but pride of
rank could not withstand a stronger passion, or prevent him from
laying his fortune and title at the feet of Ida Aumerle.”
“An opportunity for you!” suggested Intemperance.
“A golden opportunity I deemed it. What woman is not dazzled by
a coronet? what girl is insensible to the flattering attentions of him
who owns one, even if he possess no other recommendation, which,
with Dashleigh, is far from being the case? There was a struggle in
the mind of Ida. I whispered to her of all those gilded baubles for
which numbers have eagerly bartered happiness here, and forfeited
happiness hereafter. I set before her grand images of earthly
greatness, the pomp and trappings of state, the homage paid by the
world to station. I strove to inflame her mind with ambition. But
here Ida sought counsel of the All-wise, and she saw through my
glittering snare. The earl, though of character unblemished in the
eyes of man, and far from indifferent to religion, is not one whom a
heaven-bound pilgrim like Ida would choose as a companion for life.
Dashleigh’s spirit is too much clogged with earth; he is too much
divided in his service; he wears too openly my chain, as if he
deemed it an ornament or distinction. Ida prayed, reflected, and
then resolved. She declined the addresses of her uncle’s guest, and
returned home at once to her father.”
“The wound which she inflicted was not a deep one,” remarked
Intemperance. “Dashleigh was speedily consoled, without even
seeking comfort from me.”
“I poisoned his wound,” exclaimed Pride, “and drove him to seek
instant cure. Dashleigh’s rejection aroused in his breast as much
indignation as grief; and I made the disappointed and irritated man
at once offer his hand to one who was not likely to decline it,
Annabella, the young cousin of Ida.”
“And what said the high-souled Ida to the sudden change in the
object of his devotion?”
“I breathed in her ear,” answered Pride, “the suggestion, ‘He might
have waited a little longer.’ I called up a flush to the maiden’s cheek
when she received tidings of the hasty engagement. But still I met
with little but repulse. With maidenly reserve Ida concealed even
from her own family a secret which pride might have led her to
reveal, and none more affectionately congratulated the young
countess on her engagement, than she who might have worn the
honours which now devolved upon another.”
“Ida Aumerle appears to be gifted with such a power of resisting
your influence and repelling your temptations, that I can scarcely
imagine,” quoth Intemperance, “upon what you can ground your
assurance that you hold her captive at length. Pride of beauty, pride
of conquest, pride of ambition, she has subdued; to spiritual pride
she never has yielded. What dart remains in your quiver when so
many have swerved from the mark?”
“Or rather, have fallen blunted from the shield of faith,” gloomily
interrupted Pride. “Ida’s real danger began when she thought the
dart too feeble to render it needful to lift the shield against it. Ida,
on her return home, found her father on the point of contracting a
second marriage with a lady who had been one of his principal
assistants in arranging and keeping in order the machinery of his
parish. Miss Lambert, by her activity and energy, seemed a most
fitting help-meet for a pastor. She was Aumerle’s equal in fortune
and birth, and not many years his junior in age. She had been
always on good terms with his family, and the connection appeared
one of the most suitable that under the circumstances could have
been formed. And so it might have proved,” continued Pride, “but for
me!”
“Is Mrs. Aumerle, then, under your control?”
“She is somewhat proud of her good management, of her clear
common sense, of her knowledge of the world,” was the dark one’s
reply; “and this is one cause of the coldness between her and the
daughters of her husband. Ida, from childhood, had been
accustomed to govern her own actions and direct her own pursuits.
Steady and persevering in character, she had not only pursued a
course of education by herself, but had superintended that of her
more impetuous sister. Since her mother’s death Ida had been
subject to no sensible control, for her father looked upon her as
perfection, and left her a degree of freedom which to most girls
might have been highly dangerous. Thus her spirit had become more
independent, and her opinions more formed than is usual in those of
her age. On her father’s marriage Ida found herself dethroned from
the position which she so long had held. She was second where she
had been first,—second in the house, second in the parish, second in
the affections of a parent whom she almost idolatrously loved. I saw
that the moment had come for inflicting a pang; you will believe that
the opportunity was not trifled away! Ida had been accustomed to
lead rather than to follow. She exercised almost boundless influence
over her sister Mabel, and was regarded as an oracle by the poor.
Another was now taking her place, and another whose views on
many subjects materially differed from her own, who saw various
duties in a different light, and whose character disposed her to act in
petty matters the part of a zealous reformer. I marked Ida’s
annoyance at changes proposed, improvements resolved on, and I
silently pushed my advantage. I have now placed Ida in the position
of an independent state, armed to resist encroachments from, and
owning no allegiance to a powerful neighbour. There is indeed no
open war; decency, piety, and regard for the feelings of a husband
and father alike forbid all approach to that; but there is secret,
ceaseless, determined opposition. I never suffer Ida to forget that
her own tastes are more refined, her ideas more elevated than those
of her step-mother; and I will not let her perceive that in many of
the affairs of domestic life, Mrs. Aumerle, as she had wider
experience, has also clearer judgment than herself. I represent
advice from a step-mother as interference, reproof from a step-
mother as persecution, and draw Ida to seek a sphere of her own as
distinct as possible from that of the woman whom her father has
chosen for his wife.”
“Doubtless you occasionally remind the fair maid,” suggested
Intemperance, “that but for her own heroic unworldliness she might
have been a peeress of the realm.”
“I neglect nothing,” answered Pride, “that can serve to elevate the
spirit of one whom I seek to enslave. I have need of caution and
reserve, though hitherto I have met with success, for it is no easy
task thoroughly to blind a conscience once enlightened like that of
Ida. She does even now in hours of self-examination reproach
herself for a feeling towards Mrs. Aumerle which almost approaches
dislike. She feels that her own peace is disturbed; for the lightest
breath of sin can cloud the bright mirror of such a soul. But in such
hours I hover near. I draw the penitent’s attention from her own
faults to those of the woman she loves not, till I make her pity
herself where she should blame, and account the burden which I
have laid upon her as a cross appointed by Heaven.”
“O Pride, Pride!” exclaimed Intemperance with a burst of
admiration, “I am a child in artifice compared with you!”
“Rest assured that when any young mortal is disposed to look
down upon one placed above her by the will of a higher power, that
pride is lingering near.”
 “And by what name may you be known in this particular phase of
your being?” inquired Intemperance.
“The pride of self-will in the language of truth; but Ida would call
me sensitiveness,” replied the dark spirit with a gloomy smile.
 CHAPTER III.
SNARES.

“But what are sun and moon, and this revolving ball
Compared with Him who thus supports them all;
Whose attributes, all-infinite, transcend
Whate’er the mind can reach, or mortal apprehend!
Whose words drew light from chaos drear and dark,
Whose goodness smoothes this state of toil and trouble,
Compared with it—the sun is as a spark—
The boundless ocean a mere empty bubble!”

Henry St. George Tucker.

“The pastor and his wife I see approaching the church,” observed
Intemperance, glancing down in the direction of the path along
which advanced a rather stout lady, with large features and high
complexion, who was leaning on the arm of a tall, handsome, but
rather heavily-built man, in whose mild, dark eyes might be traced a
resemblance to those of his daughter.
“They come early,” said Pride; “he, to prepare for service; his wife,
to hear the school children rehearse the hymns appointed for the
day. This was once Ida’s weekly care; she is far more qualified for
the charge than her step-mother, and the music has suffered from
the change.”
“Ida showed humility, at least, in yielding up that charge,”
remarked Intemperance.
“Humility,” exclaimed Pride, an expression of ineffable scorn
convulsing his shadowy features as the word was pronounced. “I
should not marvel if Ida thought so; but hear the real state of the
case. The maiden had taken extreme pains to teach her choir a
beautiful anthem, in which a trio is introduced, which she instructed
three of the girls who had the finest voices and the most perfect
taste to sing. Mrs. Aumerle, on hearing the anthem, at once
condemned it. It was time wasted, she averred, to teach cottage-
children to sing like choristers in a cathedral; and to make a whole
congregation cease singing in order to listen to the voices of three,
was to turn the heads of the girls, and make them fancy themselves
far above the homely duties of the state in which Providence had
been pleased to place them. There was common sense in the
observations; but Ida saw in it simply want of taste, and at my
suggestion,—at my suggestion,” repeated Pride in triumph, “she
gave up charge of the music altogether, because she was offended
at any fault having been found in it by one who knew so little of the
subject.”
“Is the minister himself a good man?” inquired Intemperance.
“Good! yes, good, if any of the worms of earth can be called so,”
replied Pride, with gloomy bitterness, “for he does not regard himself
as good. Naturally weak and corrupt are the best of mortals, prone
to fall, and liable to sin, yet I succeed in persuading many that the
gold which is intrusted to their keeping imparts some intrinsic merit
to the clay vessel which contains it; that the cinder, glowing bright
from the fire which pervades it, is in itself a brilliant and beautiful
thing!”
“But Lawrence Aumerle was never your captive?”
“I thought once that he would be so,” replied Pride, his features
darkening at the recollection of disappointment and failure. “Aumerle
had been a singularly prosperous man—his life had appeared one
uninterrupted course of success. Easy in circumstances, cherished in
his family, a favourite in society, beloved by the poor, with a
disposition easy and tranquil, disturbed by no violent passion,—the
lot of Aumerle was one which might well render him a subject of
envy. In the pleasantness of that lot lay its peril. Aumerle was not
the first saint who in prosperity has thought that he should never be
moved, who has been tempted to regard earthly blessings as tokens
of Heaven’s peculiar favour. He knew little of the burden and heat of
the day, still less of the strife and the struggle. Self-satisfaction was
beginning to creep over his soul, as vegetation mantles a standing
pool over which the rough winds never sweep. ‘He is mine!’ I
thought, ‘mine until death, and indolence and apathy shall soon add
their links to the chain forged by pride of prosperity.’ But mine was
not the only eye that was watching the Vicar of Ayrley. There is an
ever-wakeful Wisdom which ofttimes defeats my most subtle
schemes, leading the blind by a way they know not, drawing back
wandering souls to the orbit of duty, even as that same Wisdom
hangs the round world upon nothing, and guides the stars in their
courses! My chain was suddenly snapped asunder by a blow which
came from a hand of love, but which, in its needful force, laid
prostrate the soul which it saved. Aumerle’s loved partner was
smitten with sickness, smitten unto death, and the doating husband
wrestled in agonizing prayer for her who was dearer to him than life.
The prayer was not granted, for the wings of the saint were fledged.
She escaped, like a freed bird, from the power of temptation, for
ever! Her husband remained behind,—Lawrence Aumerle was an
altered man. Earth had lost for him its alluring charm, and enchained
his affections no more. He was softened—humbled,” continued Pride,
with the bitterness of one who records his own defeat, “and in
another world he will reckon as the most signal mercy of his life the
tempest which scattered his joys, and dashed his hopes to the
ground! Let us not speak of him more,” continued the fierce spirit
with impatience; “his younger brother, the stately Augustine, will not
shake off my yoke so lightly.”
“His pride may well be personal pride,” said Intemperance,
following the direction of the glance of his stern companion, “if that
be he who, with the rest of the congregation, is now obeying the
summons of the church bells. Mine eyes never rested on a more
goodly man.”