Language In Use Georgetown University Round Table On Languages And Linguistics Andrea Tyler

Language In Use Georgetown University Round
Table On Languages And Linguistics Andrea Tyler
download
https://ebookbell.com/product/language-in-use-georgetown-
university-round-table-on-languages-and-linguistics-andrea-
tyler-2528096
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Language In Use A Reader Patrick Griffiths Andrew John Merrison
https://ebookbell.com/product/language-in-use-a-reader-patrick-
griffiths-andrew-john-merrison-30849136
Introducing Language In Use Aileen Bloomer
https://ebookbell.com/product/introducing-language-in-use-aileen-
bloomer-1658824
Introducing Language In Use Andrew John Merrison Aileen Bloomer
Patrick Griffiths Christopher J Hall Aileen Bloomer Patrick Griffiths
Christopher J Hall
https://ebookbell.com/product/introducing-language-in-use-andrew-john-
merrison-aileen-bloomer-patrick-griffiths-christopher-j-hall-aileen-
bloomer-patrick-griffiths-christopher-j-hall-57439946
English In Southeast Asia Features Policy And Language In Use
Varieties Of English Around The World Eeling Low Editor
https://ebookbell.com/product/english-in-southeast-asia-features-
policy-and-language-in-use-varieties-of-english-around-the-world-
eeling-low-editor-24781608

English In East And South Asia Policy Features And Language In Use 1st
Edition Ee Ling Low
https://ebookbell.com/product/english-in-east-and-south-asia-policy-
features-and-language-in-use-1st-edition-ee-ling-low-33356152
Textbased Research And Teaching A Social Semiotic Perspective On
Language In Use 1st Edition Peter Mickan
https://ebookbell.com/product/textbased-research-and-teaching-a-
social-semiotic-perspective-on-language-in-use-1st-edition-peter-
mickan-5843732
Language Use In Interlingual Familes A Japaneseenglish Sociolinguistic
Study Masayo Yamamoto
https://ebookbell.com/product/language-use-in-interlingual-familes-a-
japaneseenglish-sociolinguistic-study-masayo-yamamoto-51814688
Language Use In Interlingual Familes A Japaneseenglish Sociolinguistic
Study Masayo Yamamoto
https://ebookbell.com/product/language-use-in-interlingual-familes-a-
japaneseenglish-sociolinguistic-study-masayo-yamamoto-5596708
Rethinking Language Use In Digital Africa Technology And Communication
In Subsaharan Africa Pdf Leketi Makalela Editor
https://ebookbell.com/product/rethinking-language-use-in-digital-
africa-technology-and-communication-in-subsaharan-africa-pdf-leketi-
makalela-editor-49061954

Georgetown University Round Table on Languages and Linguistics series
Selected Titles
Discourse and Technology: Multimodal Discourse Analysis
PHILIP LEVINE AND RON SCOLLON, EDITORS
Language in Our Time: Bilingual Education and Official English, Ebonics
and Standard English, Immigration and Unz Initiative
JAMES E. ALATIS AND AI-HUI TAN, EDITORS
Linguistics, Language, and the Professions: Education, Journalism, Law,
Medicine, and Technology
JAMES E. ALATIS, HEIDI E. HAMILTON, AND AI-HUI TAN, EDITORS
Linguistics, Language, and the Real World: Discourse and Beyond
DEBORAH TANNEN AND JAMES E. ALATIS, EDITORS

LANGUAGE IN USE
Cognitive and Discourse Perspectives
on Language and Language Learning
Andrea Tyler, Mari Takada, Yiyoung Kim, and Diana Marinova, Editors
GEORGETOWN UNIVERSITY PRESS
Washington, D.C.

Georgetown University Press, Washington, D.C.
©2005 by Georgetown University Press. All rights reserved.
Printed in the United States of America
10 9 8 7 6 5 4 3 2 1 2005
This book is printed on acid-free paper meeting the requirements of the American
National Standard for Permanence in Paper for Printed Library Materials.
Library of Congress Cataloging-in-Publication Data
Language in use : cognitive and discourse perspectives on language and language learning /
Andrea E. Tyler . . . [et al.], editors.
p. cm. — (Georgetown university round table on languages and linguistics series)
Includes bibliographical references and index.
Summary: “This book explores how language is shaped by the nature of human cognition
and social-cultural activity, by studying how language is used in context in interactions
between at least two people in order to achieve some purpose. It brings together
perspectives from cognitive linguistics, discourse analysis, and first and second language
acquisition research”—Provided by the publisher.
ISBN 1-58901-044-2 (pbk. : alk. paper)
1. Language acquisition. 2. Cognitive grammar. 3. Discourse analysis. I. Tyler, Andrea. II.
Georgetown University round table on languages and linguistics series (2004).
P118.L3638 2005
401⬘.93—dc22 2004023166

Contents
Figures and Tables vii
Acknowledgments ix
Introduction xi
Andrea Tyler
PART I: LANGUAGE PROCESSING AND
FIRST-LANGUAGE LEARNING
1. Support from Language Processing for a Constructional Approach to
Grammar 3
Adele E. Goldberg and Giulia M. L. Bencini, Princeton University and New
York University
2. Homonyms and Functional Mappings in Language Acquisition 19
Devin M. Casenhiser, Princeton University
3. Little Persuaders: Japanese Children’s Use of Datte (but-because) and
Their Developing Theories of Mind 36
Tomoko Matsui, Peter McCagg, and Taeko Yamamoto, International Christian
University, Japan
4. “Because” as a Marker of Collaborative Stance in Preschool
Children’s Peer Interactions 50
Amy Kyratzis, University of California, Santa Barbara
PART II: ISSUES IN SECOND-LANGUAGE LEARNING
5. Contextualizing Interlanguage Pragmatics 65
Kathleen Bardovi-Harlig, Indiana University
6. Learning the Discourse of Friendship 85
Catherine Evans Davies, University of Alabama
7. Applied Cognitive Linguistics and Newer Trends in Foreign
Language Teaching Methodology 100
Susanne Niemeier, University of Koblenz-Landau, Germany
v

8. Language Play and Language Learning: Creating Zones of Proximal
Development in a Third-Grade Multilingual Classroom 112
Ana Christina DaSilva Iddings and Steven G. McCafferty, Vanderbilt
University and University of Nevada at Las Vegas
9. Cognates, Cognition, and Writing: An Investigation of the Use of
Cognates by University Second-Language Learners 123
Robin Cameron Scarcella and Cheryl Boyd Zimmerman, University of
California at Irvine and California State University, Fullerton
PART III: DISCOURSE RESOURCES AND MEANING
CONSTRUCTION
10. Intonation, Mental Representation, and Mutual Knowledge 139
Ann Wennerstrom, University of Washington
11. Linguistic Variation in the Lexical Episodes of University
Classroom Talk 150
Eniko Csomay, San Diego State University
12. The Unofficial Businesses of Repair Initiation: Vehicles for Affiliation
and Disaffiliation 163
Hansun Zhang Waring, Teachers College, Columbia University
13. Pragmatic Inferencing in Grammaticalization: A Case Study of
Directional Verbs in Thai 176
Kingkarn Thepkanjana and Satoshi Uehara, Chulalongkorn University,
Thailand, and Tohoku University, Japan
PART IV: LANGUAGE AND IDENTITY
14. “Trying on” the Identity of “Big Sister”: Hypothetical Narratives in
Parent-Child Discourse 191
Cynthia Gordon, Georgetown University
15. The Discourse of Local Identity in Postwar Bosnia-Herzegovina 202
Aida Premilovac, Georgetown University
16. Immigration Geographies, Multilingual Immigrants, and the
Transmission of Minority Languages: Evidence from the Igbo Brain
Drain 214
Rachel R. Reynolds, Drexel University
vi Contents

Figures and Tables
Figures
Figure 2.1 Three types of form-to-form meaning mappings 20
Figure 2.2 Comparison of many-to-one mapping with one-to-many mapping 23
Figure 2.3 Sketches of the illustrations used in experiment one 25
Figure 2.4 Mean number of correct responses 29
Figure 2.5 Mean number of no-go responses 32
Figure 2.6 Mean number of incorrect responses 33
Figure 4.1 Motivation of justifications by group 54
Figure 4.2 Allocation of validating and nonvalidating justifications to
“because” 54
Figure 10.1 An ad hoc contrast on live without lexical antecedent 142
Figure 10.2 Parents has given intonation but no antecedent 143
Figure 10.3 Mental space network triggered by contrast intonation 145
Figure 11.1 VMP patterns for a 1,500-word text segment in university
classroom discourse 152
Figure 11.2 Three lexical episode types on three dimensions of academic
language 155
Tables
Table 1.1 English argument structure constructions 6
Table 1.2 Stimuli for sorting experiment 9
Table 1.3 Example of priming sentences in Bock and Loebell (1990) 12
Table 1.4 Example of priming sentences in Bencini, Bock,
and Goldberg (n.d.) 13
Table 1.5 Example of priming sentences in Hare and Goldberg (1999) 14
Table 2.1 Mean number of correct responses from Mazzocco (1997) 22
Table 2.2 Target words and corresponding objects 26
Table 2.3 List of words and landmarks 27
Table 2.4 Participant responses as a function of condition and response type 29
Table 2.5 Participants performing significantly above chance (N = 16) 30
Table 3.1 Number of utterances containing datte spoken by Tai 41
Table 3.2 Number of Tai’s datte-responses by adult challenge types 42
Table 3.3 False-belief test results and datte use 43
Table 3.4 Number of challenge types addressed to participants 44
vii

Table 3.5 Types of children’s responses to adult challenges 44
Table 3.6 Number of response types addressed to challenges 45
Table 3.7 Number of datte-utterances produced according to types of
challenges 45
Table 3.8 Types of child utterances that followed their use of datte 46
Table 3.9 Number of utterance types that followed datte 46
Table 9.1 Background characteristics of participants in experiment 1 129
Table 9.2 Typical native-speaker results 130
Table 9.3 Examples of student errors on test of derivatives 132
Table 11.1 Selected linguistic features on four dimensions of academic
language 154
Table 11.2 Partial taxonomy of communicative tasks as found in lexical
episodes of university classroom talk 159
Table 16.1 Languages spoken by Igbo immigrants in ONI group 216
Table 16.2 Key sites for bringing together Igbo communities among ONI
immigrants, arranged by use of language varieties 219
viii Figures and Tables

Acknowledgments
This volume contains a selection of papers from the 2003 Georgetown University
Round Table on Languages and Linguistics, widely known either as GURT or the
Round Table. The theme for GURT 2003 was “Language in Use: Cognitive and Dis-
course Perspectives on Language and Language Learning.” The papers were selected
by peer review from among more than 120 presentations and 5 plenary addresses.
The editors of this volume are Andrea Tyler, Mari Takada, Yiyoung Kim, and Diana
Marinova.
The chair for GURT 2003 was Andrea Tyler, professor of linguistics at
Georgetown University. Mari Takada, doctoral student in Georgetown’s Department
of Linguistics, was the conference coordinator. We also want to acknowledge the in-
valuable service of Ken Petersen, our webmaster, and Yiyoung Kim and Diana
Marinova, the assistant coordinators who helped ensure that the conference ran
smoothly. Our thanks also go to the members of Washington CogLink who enthusias-
tically provided intellectual support, in particular Joe Grady and Michael Israel. A
special thanks to the organizers of the four invited colloquia—Kendall King, Debby
Schiffrin, Deborah Tannen, and Sarah Taub—and the phenomenal group of scholars
who agreed to participate in the colloquia. We want to acknowledge the many gradu-
ate students and faculty members of the Department of Linguistics who volunteered
to assist in organizing and running the conference. Finally, we thank the faculty of
languages and linguistics for its generous financial support.
ix

Introduction
ANDREA TYLER
Georgetown University
IN RECENT YEARS there has been growing awareness of the importance of studying lan-
guage and language learning in its context of use. Researchers who identify them-
selves as taking a cognitive approach (broadly defined) and those who take various
discourse perspectives have sounded the theme, often independently of each other,
that an accurate understanding of the properties of language requires an understand-
ing of how language is used to create meaning. Moreover, an increasing number of
researchers in language learning have argued that in acquiring a language the
learner experiences the language in context. This perspective emphasizes the im-
portance of studying language learning as it is embedded in meaningful communi-
cation and recognition that language learning is crucially shaped by the particular
language patterns to which a learner is exposed. The aim of GURT 2003 was to
bring together research from various perspectives that emphasize the shared notions
that the properties of language and the process of language learning crucially in-
volve how language is used in context and how these patterns relate to cognition
more generally.
The presentations at GURT 2003 adhered to a shared set of tenets concerning
language as it occurs in natural contexts. These shared tenets include the following:
when humans use language, they do so for a purpose; with very few exceptions, the
purpose is to communicate with other humans beings; communication always occurs
in a context; language is created by humans who are unique not only in their language
using ability but also in their particular physical and neurological anatomy, as well as
many aspects of their social organization and culture making; and language is inevi-
tably shaped by the nature of human cognition and social-cultural activity. In spite of
the fact that these attributes stem from basic, commonsensical observations, for
many linguists and language acquisitionists they have not been of central concern.
Placing this particular perspective on language at the center of our inquiries has pro-
found consequences in terms of the questions we ask, the data we consider, the pat-
terns we discover, and our interpretation of the import of those patterns.
Although cognitive researchers, discourse analysis researchers, and language ac-
quisition researchers share the foregoing assumptions about language, the particular
areas of inquiry and emphases of these subfields are diverse enough that many of us
have tended to remain unaware of the interrelations among these approaches. Thus,
we also have remained unaware of the possibilities for research from each of these
perspectives to challenge, inform, and enrich the others. A key goal of GURT 2003,
the success of which is admirably reflected in this collection of papers, was to begin
to make these connections more transparent.
xi

The essays collected in this volume represent a rich range of frameworks within
a usage-based approach to language and language learning. They can be grouped
into four strands that were central to the conference.
Language Processing and First-Language Learning
The first strand of essays examines the nature of language through the lens of lan-
guage processing and first-language learning. Goldberg and Bencini present an im-
pressive body of language processing evidence in support of a construction grammar
model of language—that is, a model that represents syntactic patterns as indepen-
dently meaningful. From the area of language comprehension, they present evidence
that suggests that comprehenders recognize that basic sentence patterns are directly
linked to meaning, independently of the main verb. Reviewing priming studies,
Goldberg and Bencini also provide evidence that units of the type and size of con-
structions can be primed in language production. Thus, the psycholinguistic evi-
dence offers support for a constructional approach to grammar. Presenting new ex-
perimental findings, Casenhiser shows that young children tend to disprefer
homophony. He argues that patterns of one-to-many mappings between form and
meaning potentially reduce communicative efficiency and concludes that children’s
dispreference for homophones supports the hypothesis that communicative goals of
language are reflected in learning biases. Matsui, McCagg, and Yamamoto examine
the development of young Japanese children’s use of datte, a discourse marker that
roughly translates to but-because. Using both experimental and longitudinal observa-
tional data, they conclude that children as young as three years begin to use datte as a
justification marker in communicative situations in which they sense opposition to
their statements and only later in response to why questions. They argue that this de-
velopment of the suasive marker reflects the children’s growing awareness of the par-
ticular contexts in which adults use datte, in conjunction with their growing theory of
mind. Kyratzis also examines children’s use of a suasive discourse marker: the Eng-
lish because. Incorporating gender into her analysis of the discourse of preschool
play groups, Kyratzis finds that young girls tend to use because more often than boys
of the same age. Using the construct of participation network and considering both
the contextualized presence and absence of a linguistic feature, Kyratzis concludes
that when these young girls use because, it tends to work as a marker of collaborative
stance. In contrast, the absence of because that is characteristic of boy’s discourse is
used to convey either urgency or disagreement with the partner.
Issues in Second-Language Learning
The second group of essays addresses insights that discourse and usage-based mod-
els provide into issues of second-language learning (L2). Bardovi-Harlig offers a
comprehensive review of the literature on interlanguage pragmatics. She argues that
interlanguage pragmatics research would benefit from a “recontextualization” into
the larger framework of communication and communicative competence. She advo-
cates reorienting L2 pragmatic research to emphasize language learning embedded
in social interaction and the importance of contextual constraints on appropriate lan-
guage use and interpretation of utterances. Davies’ study offers a window into a
xii Introduction

natural process of language socialization—language coaching—with a focus (cre-
ated and articulated by the participants themselves) on particular cognitive dimen-
sions of the contextualized speech activity. Analyzing language-coaching discourse
between two friends who speak different varieties of English, Davies demonstrates
the need to incorporate the cognitive dimension of collaboratively constructed
intersubjectivity into a model of interactional competence. Niemeier provides a gen-
eral overview of key points of cognitive linguistics and how they might apply to sec-
ond-language teaching. In particular, she emphasizes the tenet that there are no dis-
crete boundaries between the lexicon and syntax; this perspective allows the L2
teacher to exploit the recurrent patterns, or operational uniformity, found at all “lev-
els” of linguistic organization. Taking a Vygotskyan perspective, Iddings and
McCafferty examine naturally occurring interactions of third-grade children who do
not share a mutual language. The interactions show how, through language play, the
children created a hybrid functional system for making meaning that afforded them
scaffolded opportunities to communicate, as well as to develop metalinguistic knowl-
edge of their first languages (L1s). Scarcella and Zimmerman present a series of ex-
periments focusing on L2 learners’ use of L1 cognates in L2 writing. They offer the
surprising finding that native Spanish speakers use relatively few academically ap-
propriate English lexemes that have cognates in Spanish. They conclude that cognate
knowledge does not transfer automatically. They suggest that explicit teaching of
cognates in the particular context of academic writing may be necessary before most
L1 Spanish speakers will be able to effectively produce English cognates in written
discourse.
Discourse Resources and Meaning Construction
Using a diverse array of methodologies, the third group of essays examines how speak-
ers employ various discourse-level resources to structure interaction and create mean-
ing. Wennerstrom considers the contribution of “contrast” and “given” intonation pat-
terns to meaning construction and how prosodic analysis might contribute to cognitive
models such as mental space theory (Fauconnier and Turner 2002) and Clark’s (1992)
model of community membership. Wennerstrom concludes that discourse-level
prosodics provide a rich resource for creating meaning and that analysis of their pat-
terns of use can shed new light on cognitive constructions and processes. Using multi-
dimensional analysis, Csomay provides a linguistic characterization of the lexically
coherent discourse units found in university classroom discourse. Combining ele-
ments of quantitative and qualitative analysis, the methodology allows classification
of lexical episodes into episode types (involved narrative, procedural, content-ori-
ented), based on their shared linguistic characteristics and association of these lexical
episode types with varying communicative purposes. Waring presents evidence that
speakers in graduate seminars use other-initiated repair as an interactional resource
both to advance their potentially disaffiliative claims and to find a way out of
interactional deadlock once a stalemate is reached. Rather than always employing re-
pair initiations to address problems in speaking, hearing, or understanding, English
speakers in this particular context use repairs as vehicles for conveying speaker stance.
In the final paper in this group, Thepkanjana and Uehara add a diachronic dimension
INTRODUCTION xiii

by considering how speakers’ contextualized uses of linguistic resources give rise to
implicatures that subsequently become entrenched (or grammaticalized) and eventu-
ally shape the language. They examine a set of polysemous Thai lexemes that
synchronically function as both directional verbs and success markers. They argue that
these lexemes express notions of motion and direction (coming from the earlier, core
meaning of the directional verbs), while also expressing the meaning of success that
arises from pragmatic inferences linked to the general human conceptualization of for-
ward motion along a path correlating with reaching a goal. The authors argue that the
multiple meanings associated with these directional verbs/success markers indicate
they are in an early stage of grammaticalization, in which meaning shift rather than se-
mantic bleaching has occurred.
Language and Identity
The final set of essays addresses issues of language use and creation of social iden-
tity. Gordon analyzes parent-child discourse involving a three-year-old’s narratives
concerning her yet-to-be-born brother. The analysis reveals how the interactions al-
low the young child to rehearse the future role of being a big sister and thus to ac-
tively shape her identity through imaginative (or hypothetical) discourse. This dis-
course is a stunning example of a young child, in collaboration with her parents,
actively creating multiple, complex conceptual blends. Premilovac investigates the
ways in which the discursive construction of local identity (i.e., identity tied to place
such as town versus country) is used at a reunion among old, ethnically diverse
friends from Bosnia-Herzegovina to reassert a multiethnic community in the wake of
radical, exclusionary nationalism. She argues that because the construction of local
identity can cut across ethnic and national boundaries, this discursive construct al-
lows accentuation of similarities among the groups and can serve as a basis for re-
building communities’ multiethnic composition. Reynolds offers an ethnographic
study of language maintenance and social identity among the contemporary Igbo di-
aspora living in the United States. Unlike many historical immigrant groups, the Igbo
have not settled in distinct neighborhoods. Nevertheless, through specific cultural or-
ganizations and special gatherings, which constitute “key sites” for language use, the
group creates contexts in which Igbo verbal arts and identity are performed and
transmitted to a new generation.
This volume presents a glimpse into the rich, intersecting lines of research repre-
sented at GURT 2003. For language researchers who are unfamiliar with usage-
based approaches to language, they offer a vibrant introduction to the range of
research currently being undertaken within this framework. For those working within
usage-based models, they demonstrate the challenges and potential rewards when—
to paraphrase Proust—we seek discovery not by simply searching for new landscapes
but in seeing the familiar with new eyes.
REFERENCES
Clark, Herbert H. 1992. Arenas of language use. Chicago: University of Chicago Press.
Fauconnier, Gilles, and Mark Turner. 2002. The way we think: Conceptual blending and the mind’s hidden
capacities. New York: Basic Books.
xiv Introduction

I
Language Processing and
First-Language Learning

1
Support from Language Processing for a
Constructional Approach to Grammar
ADELE E. GOLDBERG AND GIULIA M. L. BENCINI
Princeton University and New York University
A KEY TENET of Construction Grammar (CxG) (e.g., Goldberg 1995; Kay and Fillmore
1999; Michaelis and Lambrecht 1996) is that the basic units of language are learned
pairings of form and function: constructions. CxG strives to characterize the knowl-
edge that underlies a native speaker’s capacity to understand and produce an indefi-
nite number of sentences and discriminate between the acceptable and unacceptable
sentences in his or her native language. It departs from classical generative ap-
proaches in the Chomskian tradition in several crucial ways, however.1
Perhaps the most far-reaching difference stems from CxG’s additional commit-
ment to account for the entirety of a language. This commitment to full coverage en-
tails that problematic data cannot be set aside as irrelevant to the theory. Over the past
ten to fifteen years, linguists working within the CxG framework have provided anal-
yses of a large number of constructions traditionally relegated to the “periphery” (or
“residue”) of grammar. Systematic exploration of these seemingly noncore phenom-
ena has led to the discovery that these cases involve a greater degree of systematicity
and generalization than previously assumed (e.g., Goldberg 1995; Kay and Fillmore
1999; Michaelis and Lambrecht 1996). Importantly, CxG scholars have demon-
strated that the theoretical machinery that is used to account for seemingly more idi-
omatic cases (e.g., the WXDY construction that licenses expressions such as “What
are they doing resurrecting constructions?”) is the same as that needed to account for
more general patterns, (e.g., subject auxiliary inversion, which is found in yes/no and
main clause wh- questions, as well as several other more specific constructions). This
finding, in turn, has led to a blurring of the boundary between lexicon and grammar
and between “core” and “periphery.” The vision of grammar that emerges is a cline of
linguistic phenomena from the more idiomatic to the more abstract and general.
In the following section, we review some of the linguistic evidence for a con-
structional account of argument structure, which is uncontroversially part of tradi-
tional core grammar. We provide evidence for this approach from language compre-
hension. Specifically, we review evidence that suggests that comprehenders
recognize that basic sentence patterns such as the transitive (Pat threw the ball),
ditransitive (Pat gave Kim a ball) and resultative (Pat took the box apart) are directly
linked to meaning, independently of the main verb. We provide evidence from
3

language production in the third section. In particular, we report evidence that indi-
cates that units of the size and kind of constructions can be primed in language pro-
duction. Thus, the psycholinguistic evidence complements the growing body of tra-
ditional linguistic evidence accrued over the past fifteen years for adoption of a
constructional approach to grammar (e.g., Abbott-Smith, Lieven, and Tomasello
2004; Croft 2001; Gleitman et al. 1996; Goldberg 2003; Jackendoff 2002; Kay and
Fillmore 1999; Lambrecht 2001; Langacker 1988a, 1988b; Michaelis 2001; Zwicky
1994). Additional evidence comes from the area of child language (e.g., Bates and
Goodman 1997; Chang and Maia 2001; Childers and Tomasello 2001; Diessel and
Tomasello 2001; Tomasello 2003).
Theoretical Motivation for a Constructional Approach to
Argument Structure
What aspects of a sentence convey contentful meaning? Verbal predicates seem to
play a privileged role in determining a sentence’s meaning and overall form
(Chomsky 1981; Fillmore 1968; Lakoff 1970). For example, in the sentences in (1)
there seems to be a natural correspondence between the number and types of actors
in the scene and the number and types of actors typically associated with the
predicate.
(1) a. She sneezed.
b. She kicked the table.
c. She gave him a beer.
d. She threw her glass across the room.
Sneezing typically is conceived of as a one-argument predicate, with one partici-
pant role: “the sneezer.” A kicking event consists of two arguments—the “kicker”
and the “kickee”—whereas give is a trivalent predicate expressing a “giver,” a
“given,” and a “givee.” This observation has led to the traditional view that the overall
meaning of a sentence—the information about “who does what to whom”—is a pro-
jection of the lexical specifications of its verbal head. Under this lexical-projectionist
account, part of the lexical entry of give is that it requires three arguments: give [NP
[V NP NP]].
Unlike the predicates in formal logic however, natural language predicates typi-
cally occur in more than one (often many) alternative syntactic frames. For example,
give can occur in two alternate forms that seem to express roughly the same proposi-
tional meaning:
(2) NP V NP PP (dative)
a. Pat gave a cookie to the child.
NP V NP NP (ditransitive)
b. Pat gave the child a cookie.
This property is not exclusive to give, of course. Languages typically provide
more than one way of saying roughly the same thing, and accounting for these struc-
tural alternatives has been a central preoccupation of linguistic theory.
4 Part I: Language Processing and First-Language Learning

Early accounts of argument structure alternations relied on the rough
paraphrasability of the members of an alternation to posit the existence of a
transformational rule between them (e.g., Partee 1965, 1971; Fillmore 1968). Thus,
linguists assumed that if two sentences were semantically (truth-functionally) identi-
cal, they should be structurally identical at some level of syntactic representation—
specifically at the level of D-structure, where semantic relations were believed to be
read off of syntactic configurations. Yet even the early proponents of transformations
also noted that argument structure configurations are associated with subtle but sys-
tematic variations in meaning (e.g., Anderson 1971; Partee 1965; Fillmore 1968;
Wierzbicka 1988).
The recognition of differences in (3) and (4), for example, led Partee (1965) to
note that the ditransitive argument structure requires that the meaning be “X causes
Y to receive Z”:
(3) a. Pat sent a package to the boarder.
b. Pat sent the boarder a package. (ditransitive)
(4) a. Pat sent a package to the border.
b. *Pat sent the border a package. (ditransitive)
Additional semantic differences occur in the so-called spray/load alternation in
(5):
(5) a. Pat loaded hay onto the truck.
b. Pat loaded the truck with hay.
Sentence 5a differs semantically from 5b in that only 5b entails that the truck is
somehow affected by the hay-loading (Anderson 1971). A natural interpretation of
5b is that the truck is completely loaded; in contrast, 5a may refer to a situation in
which only one bale of hay has been placed onto the truck.
The recognition that differences in syntactic form are associated with subtle se-
mantic differences led other theorists to abandon a syntactic account of alternations.
In addition, some linguists also observed that many argument structure alternations
appear to be licensed by semantically defined classes of verbs (e.g., Levin 1993;
Pinker 1989). A broad variety of lexicalist accounts of alternations were therefore
proposed as alternatives to syntactic accounts (e.g., Pinker 1989; Pollard and Sag
1994). For example, in the lexicalist account of the dative alternation proposed by
Pinker (1989), a lexicosemantic rule is assumed to take as input a dative verb with the
meaning “X causes Y to go to Z” and produce as output a ditransitive verb with the
meaning “X causes Z to have Y.” Under this argument, the syntax of the ditransitive
alternant derives from quasi-universal linking rules that map thematic roles onto
grammatical functions. Central to Pinker’s account is the assumption that the surface
syntax is in one-to-one correspondence with the thematic properties of the verb. To
account for the fact that a given verb stem can appear in more than one argument
structure, different verb senses were posited; the different senses were associated
with different thematic structures and were related to one another via generative lexi-
cal rules. We refer to this account of argument structure as the multiple-sense view.
SUPPORT FROM LANGUAGE PROCESSING FOR A CONSTRUCTIONAL APPROACH TO GRAMMAR 5

The two senses for the verbal stem bring, as it occurs in the prepositional dative and
ditransitive frames, are represented in (6).
(6) Example Lexical rule:
bring-1 (sem: X causes Y to go to Z ) → bring-2 (sem: X causes Z to
have Y)
The constructional solution to argument structure shares with the lexicosemantic
accounts an emphasis on the meaning distinctions associated with different argument
structure patterns. As we will see, however, the constructional account has both theo-
retical and empirical advantages over lexical accounts that stipulate the existence of
different verb senses.
Instead of positing different verb senses without independent evidence for them,
the constructional approach assigns meaning directly to various abstract argument
structure types, thereby recognizing the argument structure patterns as linguistic
units in their own right (Goldberg 1995; Michaelis and Lambrecht 1996; Rappaport-
Hovav and Levin 1996). Examples of English argument structure constructions with
their forms and proposed meanings are shown in table 1.1.
On the constructional view, argument structure patterns contribute directly to the
overall meaning of a sentence, and a division of labor can be posited between the
meaning of the construction and the meaning of the verb in a sentence. Although the
constructional meaning may be redundant—perhaps prototypically—with that of the
main verb, the verb and construction may contribute distinct aspects of meaning to
the overall interpretation. For example, the ditransitive construction has been argued
to be associated with the meaning of transfer or “giving” (Goldberg 1995; Green
1974; Pinker 1989). When this construction is used with give, as in Kim gave Pat a
book, the contribution of the construction is wholly redundant with the meaning of
the verb. The same is true when the construction is used with send, mail, and hand.
As is clear from these latter verbs, lexical items typically have a richer core meaning
than the meanings of abstract constructions.
In many cases, however, the meaning of the construction contributes an aspect of
meaning to the overall interpretation that is not evident in the verb in isolation. For
example, the verb kick need not entail or imply transfer (cf. Kim kicked the wall). Yet
when kick appears in the ditransitive construction, the notion of transfer is entailed.
The ditransitive construction itself appears to contribute this aspect of meaning to the
Table 1.1.
English argument structure constructions
Construction Form Meaning Example
Transitive Subject Verb Object X acts on Y Pat opened the door.
Ditransitive Subject Verb Object1 X causes Y to Sue faxed her a letter.
Object2 receive Z
Resultative Subject Verb X causes Y to Kim talked himself silly.
Object RP become Z

sentence. That is, the sentence Kim kicked Pat the ball can be roughly paraphrased as
“Kim caused Pat to receive the ball by kicking it.” The construction contributes the
overall meaning of “X causes Y to receive Z” and the verb specifies the means by
which the transfer is achieved (i.e., the act of kicking).
The multiple-verb-sense approach to argument structure may run into problems
that the constructional approach avoids (Goldberg 1995). For example, consider the
sentences in (7):
(7) a. Pat sneezed.
b. She sneezed the foam off the cappuccino. (Ahrens 1995)
c. She sneezed a terrible sneeze.
d. She sneezed herself silly.
e. She sneezed onto the computer screen.
f. She sneezed her way to the doctor’s office.
To account for examples 7b–7f, the multiple-sense approach would require pos-
iting multiple special senses of the verb sneeze. For example, to license sentence 7b
the following implausible sense of sneeze would be required:
(8) Sneeze-2: “to cause something to move by sneezing”
Sneeze is not unusual in this respect. Additional examples of verbs for which the
multiple-verb sense account requires implausible senses are given in (9).
(9) a. The truck rumbled down the road.
b. She baked him a cake.
c. Pat eyebrow’d her surprise.
d. We will overnight you that package.
e. He kissed her unconscious.
f. They were unable to pray the little boys home.
g. The soldiers starved them out of their hiding place.
h. They wined, dined, and golfed their way into millions of yen.
In a constructional approach, stipulation of these implausible verb senses is
avoided by recognizing that phrasal patterns themselves can be associated with
meaning. The constructional meaning integrates with the more specific verb mean-
ing in particular ways. For example, in 7b sneezing causes the transfer; in 9a rum-
bling is an effect of the motion, and in 9b baking is a precondition of transfer.
In light of such cases, even some early proponents of the multiple-verb-sense ap-
proach have recognized certain instances in which it seems preferable to view verb
meaning as entering into composition with an independently existing construction or
template (Pinker 1994; Rappaport-Hovav and Levin 1998). Additional arguments ap-
pear in Goldberg (1995). In the following section we provide arguments for an ad-
vantage of a constructional account over a multiple-sense account from language
comprehension.

Evidence for Constructional Meaning from Comprehension
What types of linguistic information do people use to construct the meaning of a sen-
tence? Most psycholinguistic models of sentence comprehension follow linguistic
theory and assume that the main determinant of sentence meaning is the verb. It does
seem to be true that of all the words in a sentence, verbs carry the most information
about the syntax and semantics of the sentence. Because of the high predictive value
of verbs, it is reasonable to assume that people use this information during compre-
hension to predict other lexical items in the sentence and the overall meaning of the
sentence. Experimental evidence has demonstrated that in fact the main verb is a crit-
ical factor in sentence comprehension (e.g., Ahrens 2003; Garnsey et al. 1997).
In this section, however, we review studies that provide evidence for the exis-
tence of sentence-level generalizations that are used in language comprehension to
construct an interpretation of an unknown predicate as well as the overall meaning of
the sentence. Studies by Ahrens (1995) and Kaschak and Glenberg (2000) show that
the way comprehenders interpret novel verbs depends on the sentence patterns in
which the verbs occur.
Ahrens (1995) conducted an experiment with a novel verb form. She asked 100
native English speakers to decide what moop meant in the sentence She mooped him
something. Sixty percent of subjects responded by saying that moop meant “give,”
despite the fact that several verbs have higher overall frequency than give and could
be used in that frame, including take and tell.
Kaschak and Glenberg (2000) show that subjects rely on constructional meaning
when they encounter nouns used as verbs in novel ways (e.g., to crutch). In particular,
they show that different constructions differentially influence the interpretations of
the novel verbs. For example, She crutched him the ball (ditransitive) is interpreted to
mean that she used the crutch to transfer the ball to him, perhaps using the crutch as
one would a hockey stick. On the other hand, She crutched him (transitive) might be
interpreted to mean that she hit him over the head with the crutch. They suggest that
the constructional pattern specifies a general scene and that the “affordances” of par-
ticular objects are used to specify the scene in detail. It cannot be the semantics of the
verb that is used in comprehension because the word form is not stored as a verb but
as a noun.
Bencini and Goldberg (2000) conducted an experiment with the aim of directly
comparing the semantic contribution of the construction with that of the morphologi-
cal form of the verb. As shown in table 1.2, the stimuli were sixteen sentences created
by crossing four verbs with four different constructions.2
Seventeen University of Illinois undergraduate students were asked to sort the six-
teen sentences, which were provided in random order, into four piles on the basis of
“overall sentence meaning.” Subjects could sort equally well by verb: For example, all
instances of throw (1a–d) could have been grouped together, regardless of construc-
tion. Subjects also could sort by construction: For example, all instances of the VOO or
ditransitive construction (1a, 2a, 3a, and 4a) could have been grouped together.
It would be possible, of course, to design stimuli with a great deal of overlapping
propositional content so that we could a priori predict either a verb or constructional
sort. For example, the sentences Pat shot the duck and Pat shot the duck dead would

very likely be grouped together on the basis of overall meaning despite the fact that
the argument structure patterns are distinct. Conversely, Pat shot the elephant and Pa-
tricia stabbed a pachyderm probably would be grouped together despite the fact that
no exact words were shared. The stimuli were designed to minimize such contentful
overlap contributed by anything other than the lexical verb. No other lexical items in
the stimuli were identical or near-synonyms.
Use of the sorting paradigm is a particularly stringent test to demonstrate the
role of constructions. Medin, Wattenmaker, and Hampson (1987) have shown that
there is a strong, domain-independent bias toward sorting on the basis of a single di-
mension, even with categories that are designed to resist such one-dimensional sorts
in favor of a sort based on a family resemblance structure (Rosch and Mervis 1975).
One-dimensional sorting has been found even with large numbers of dimensions
(Smith 1981), ternary values on each dimension (Anh and Medin 1992), holistic
stimuli, and stimuli for which an obvious multidimensional descriptor was available
(Regehr and Brooks 1995).
The stimuli used in Bencini and Goldberg (2000) presented subjects with an op-
portunity to sort according to a single dimension: the verb. Constructional sorts
Table 1.2.
Stimuli for sorting experiment
Example sentences Construction types
1a. Pat threw the hammer. (VO) Transitive
b. Chris threw Linda the pencil. (VOO) Ditransitive
c. Pat threw the key onto the roof. (VOL) Caused Motion
d. Lyn threw the box apart. (VOR) Resultative
2a. Michelle got the book. (VO) Transitive
b. Beth got Liz an invitation. (VOO) Ditransitive
c. Laura got the ball into the net. (VOL) Caused Motion
d. Dana got the mattress inflated. (VOR) Resultative
3a. Barbara sliced the bread. (VO) Transitive
b. Jennifer sliced Terry an apple. (VOO) Ditransitive
c. Meg sliced the ham onto the plate. (VOL) Caused Motion
d. Nancy sliced the tire open. (VOR) Resultative
4a. Audrey took the watch. (VO) Transitive
b. Paula took Sue a message. (VOO) Ditransitive
c. Kim took the rose into the house. (VOL) Caused Motion
d. Rachel took the wall down. (VOR) Resultative

required subjects to note an abstract relational similarity that required recognition
that several grammatical functions co-occur. Thus, we would expect verb sorts to
have an inherent advantage over constructional sorts.
Nonetheless, six subjects produced entirely construction sorts. Seven other sub-
jects produced entirely verb sorts, and four subjects provided mixed sorts. To include
the mixed sorts in the analysis, the results were analyzed according to how many
changes would be required from the subject’s sort to produce a sort entirely by verb
(VS) or a sort entirely by construction (CS). The average number of changes required
for the sort to be entirely by verb was 5.5; the average number of changes required for
the sort to be entirely by construction was 5.7. The difference between these scores
does not approach significance. That is, subjects were just as likely to sort by construc-
tion as they were to sort according to the single dimension of the morphological form
of the verb. If verbs provided equally good cues to overall sentence meaning, there
would be no motivation to overcome the well-documented preference for one-dimen-
sional sorts: Subjects would have no motivation to sort by construction instead of by
verb. Bencini and Goldberg (2000) hypothesize that constructional sorts were able to
overcome the one-dimensional sorting bias to this extent because constructions are
betterpredictorsofoverallsentencemeaningthanthemorphologicalformoftheverb.
A question arises about why constructions should perform at least as well as pre-
dictors of overall sentence meaning as verbs. The answer, we believe, stems from the
fact that in context, knowing the number and type of arguments conveys a great deal
about the scene being conveyed. To the extent that verbs encode rich semantic frames
that can be related to several different basic scenes, the complement configuration or
construction will be as good a predictor of sentence meaning as the semantically
richer but more flexible verb.
On the multiple-sense view, the reason instances of throw, for example, were put
into separate piles was that each instance represented a distinct sense that was more
similar in meaning to one of the senses of another verb than to the other senses of
throw. The only way for subjects to discern which verb sense was involved, however,
was to recognize the argument structure pattern and its associated meaning. That is,
the proposed different verb senses all look the same; the only way to determine that a
particular sense is involved is to note the particular argument structure pattern that is
expressed and infer which verb sense must have produced such a pattern. Therefore,
at least from a comprehension point of view, the pairing of argument structure pattern
with meaning must be primary.
The most important contribution of the studies cited in this section are that they
provide a sufficiency proof that types of complement configurations play a crucial
role in sentence interpretation, independent of the contribution of the main verb. The
results suggest that constructions are psychologically real linguistic categories that
speakers use in comprehension.3
Evidence for Constructions in Language Production
In this section we provide experimental evidence that units of the type and kind of
constructions are activated during the processes of language production. We present
three studies. These three studies build on a large body of work conducted primarily
by Bock and colleagues that employ a structural priming methodology (Bock 1986).

In structural priming, the basic finding is that people tend to reuse the syntactic struc-
ture of a sentence (the prime) they have previously produced (Bock 1986) or heard
(Branigan et al. 2000). Participants are exposed to a long sequence of pictures and
auditorily presented sentences. On each priming trial, participants first hear a prim-
ing sentence such as “The new graduate was hired by the software company.” They
repeat the sentence out loud and decide whether they have said this sentence before.
They then see and describe a pictured event that can be described with either of the
targeted structural alternatives. Their description might be something like “The mail-
man is being chased by an angry poodle.” On the same priming trial, another group
of participants hears and repeats the sentence “The new graduate left the software
company.” Structural priming occurs when participants match the overall structure of
the sentence prime in their subsequent picture description.
Structural priming provides a powerful tool to investigate the mental representa-
tion of linguistic units at the level of the sentence (Branigan et al. 1995; Bencini
2002). The ecological validity of the structural priming paradigm is supported by the
finding that a tendency toward structural repetition occurs in more naturalistic set-
tings, such as written and spoken corpora (e.g., Tannen 1989; Weiner and Labov
1983) and dialogue (Levelt and Kelter 1982), suggesting that it is not a laboratory-in-
duced phenomenon.
Bock and colleagues (e.g., Bock and Griffin 2000; Bock and Loebell 1990;
Bock, Loebell, and Morey 1992) have shown in several studies that passives prime
passives, ditransitives prime ditransitives, and datives prime datives (cf. also
Abbott-Smith, Lieven, and Tomasello 2004; Branigan et al. 1995; Chang et al. 2000;
Hare and Goldberg 1999; Potter and Lombardi 1998; Smith and Wheeldon 2001;
Tomasello 2003; Yamashita, Chang, and Hirose 2003). Bock’s original claim was
that syntactic tree structures, not constructions with associated meanings, were in-
volved in priming.
More recent work has investigated the question of whether constructional prim-
ing exists. That is, can abstract pairings of form with meaning be primed? Chang,
Bock, and Goldberg (2003) conducted an experiment in which syntactic structure
was controlled for, while two different constructions were used as primes. Sample
prime and target sentences are given below:
Sample Primes
load with She loaded the wagon with hay.
load onto She loaded hay onto the wagon.
Sample Targets
He embroidered the shirt with flowers.
He embroidered flowers onto the shirt.
Subjects were asked to recall a sentence as it was presented after a short
distractor task. Such rapid serial visual presentation (RSVP) tasks have been shown
to allow priming effects (Potter and Lombardi 1998). If semantics matters in priming,
we should see “load-with” structures priming other “load-with” structures more than
“load-onto” structures. In fact, this is exactly what Chang, Bock, and Goldberg
(2003) found. In our experiment with 80 subjects, “load-with” structures primed

other “load-with” structures in 99 percent of cases, whereas “load-onto” structures
primed “load-with” structures in 95 percent of the cases.
Likewise, Griffin and Weinstein-Tull (2003) have shown that object-raising sen-
tences prime object-raising sentences more than object-control sentences. This find-
ing also suggests constructional priming; the results would be unexpected if only sur-
face form were taken into account because the two constructions arguably have the
same surface form.
Given these results, it is worth returning to the original motivation for earlier
claims that syntactic constituent structure, not constructions (form-meaning pair-
ings), are primed. Bock and Loebell (1990) made perhaps the strongest case for this
claim with a series of experiments. In one experiment, they showed that both datives
and locatives primed dative descriptions of (unrelated) pictures equally well. Exam-
ple primes are given below:
Primes:
A. The wealthy widow gave her Mercedes to the church. (dative)
B. The wealthy widow drove her Mercedes to the church. (locative)
The constructional interpretation of this result stems from the idea that “dative”
and “locative” expressions are actually both instances of the same Caused-Motion
construction (Goldberg 1995, 2002).
Caused-Motion Construction:
X causes Y to move Z
Subj V Obj Obl
Examples:
She drove the box to Missouri.
She drove the box to Mary.
She threw the box to Mary.
She gave the box to Mary.
Therefore, the findings are that caused-motion expressions prime caused-motion
expressions—a result that is predicted by a constructionist account of priming. In fact,
Bock and Loebell (1990) also acknowledge that locative and dative expressions are se-
mantically similar. They therefore performed a second experiment in which they in-
vestigated whether intransitive locative expressions primed passives—a construction
with the same syntax as locatives but with clearly distinct meaning (table 1.3).
Table 1.3.
Example of priming sentences in Bock and Loebell (1990)
Prime Type Examples
Passive The 747 was radioed by the airport control tower.
Intransitive locative The 747 was landing by the airport control tower.
Active (control) The 747 radioed the airport control tower.

Bock and Loebell (1990) found that, in fact, intransitive locatives did prime
passives. This finding was the strongest evidence for purely syntactic,
nonconstructional priming. Yet a close look at the stimuli used revealed that the prep-
osition by and the auxiliary be appeared in every intransitive locative prime. A ques-
tion naturally arises: Was it the shared morphemes, rather than the shared syntactic
structure, that produced the priming? To address this question, Bencini, Bock, and
Goldberg (n.d.) attempted to replicate the Bock and Loebell (1990) findings while
adding a fourth condition in which intransitive locatives without shared morphology
were used as primes. The four conditions are shown in table 1.4.
Bencini, Bock, and Goldberg (n.d.) replicated Bock and Loebell’s (1990) results,
demonstrating that locatives with shared morphology primed passives and, also as
expected, passives prime passives. Bencini, Bock, and Goldberg (n.d.) also found a
significant difference, however, between the passive condition and the locative con-
dition without shared morphology; in contrast, the locative without shared morphol-
ogy condition did not prime passives significantly more than the control group. This
finding is intriguing because it may indicate that shared syntactic structure is not suf-
ficient to induce priming.4
Hare and Goldberg (1999) designed a different test of the idea that pure syntactic
tree structure rather than some sort of form-meaning pairing was involved in prim-
ing. Recall that it has been well established that ditransitives prime ditransitives and
that instances of the caused-motion construction prime other instances of the
caused-motion construction. Hare and Goldberg (1999) attempted to determine
whether a third sort of prime—“provide-with” primes—would differentially prime
either caused-motion expressions (“datives”) or ditransitive descriptions of scenes of
transfer. Examples of the “provide-with” sort of primes are given in table 1.5.
“Provide with” sentences arguably have the same syntactic form as caused-mo-
tion expressions—NP [V NP PP]—yet the order of rough semantic roles involved
parallels the ditransitive: Agent Recipient Theme. Results demonstrated that “pro-
vide-with” expressions prime ditransitive descriptions of (unrelated) pictures as
much as ditransitives do. There was no evidence of priming of caused-motion expres-
sions, despite the shared syntactic form. Thus, when order of semantic roles is con-
trasted with constituent structure, only the order of semantic roles shows priming,
with no apparent interaction with constituent structure.5
What do the structural priming facts mean? First, constructions can be primed,
which means that the level of generalization involved in argument structure construc-
tions is psychologically real. Furthermore, it is possible that priming of structure may
Table 1.4.
Example of priming sentences in Bencini, Bock, and Goldberg (n.d.)
Prime Type Examples
Passives The 747 was landed by the pilot.
Locatives w/ shared morphology The 747 was landing by the airport control tower.
Locatives w/o shared morphology The 747 might land near the airport control tower.
Active (control) The 747 radioed the airport control tower.

not be independent of meaning; thus, the priming mechanism may encourage speak-
ers to categorize on the basis of form and meaning—exactly the sort of categorization
that is required by construction-based approaches. We briefly discuss a very general
processing advantage of CxG and other theories that assume a parallel representation
of linguistic knowledge (e.g., Jackendoff 2002; Sag and Wasow 1999).
Recognition of the Incrementality of Language Processing
Like other parallel constraint-based models of grammar (e.g., head-driven phrase
structure grammar), CxG assumes that syntactic, semantic, and phonological infor-
mation is represented in parallel. Parallelism has some attractive features from the
perspective of processing (Bencini 2002; Sag and Wasow 1999; Jackendoff 2002).
These features are evident when one contrasts parallel grammars with the main-
stream generative grammar architecture, in which syntax is the “engine” and seman-
tics and phonology are interpretative levels (e.g., Chomsky 1981). Although main-
stream generative grammar is a model of competence, it is possible to derive a certain
general processing implication from its main organizational features (see Bock,
Loebell, and Morey 1992 for a good example): Namely, it is presented as inherently
directional. Derivations are always computed from the syntax (through various inter-
mediate derivational steps) to meaning and sound.
This type of grammar clearly requires some sort of adjustment to be compatible
with the disparate processes of language comprehension and production. In language
production, the starting point is a meaning to be conveyed, and the goal is to convert
this meaning into a motor program for speech. Along the way, syntactic information
is used to produce a well-formed utterance (Bock 1995; Bock and Levelt 1994;
Garrett 1980). In comprehension, the starting point is a phonetic representation de-
rived from auditory or visual input, and the goal is to compute a meaning.
The parallel architecture, on the other hand, is not inherently directional and
therefore is compatible in a transparent way with both comprehension and produc-
tion. Parallelism allows for differential weighting and ordering of linguistic informa-
tion in processing (e.g., Bencini 2002). For example, in comprehension one need not
assume that a syntactic structure must be computed “first.” A more rapid and incre-
mental integration of different sources of information can be assumed to constrain
the interpretation as information becomes available. This analysis is consistent with
many findings from online comprehension studies (e.g., Just and Carpenter 1982;
Garnsey et al. 1997).
Table 1.5.
Example of priming sentences in Hare and Goldberg (1999)
Sample primes in “provide-with” condition
The government provided the troops with arms.
His editor credited Bob with the hot story.
The father entrusted his daughter with the keys.

Summary and Conclusions
In this essay we have reviewed several pieces of experimental evidence for argument
structure constructions from adult language processing. Section 2 provides evidence
from comprehension that sentence patterns such as the ditransitive, caused-motion,
and resultative contribute to overall sentence interpretation independently of the spe-
cific contribution of the main verb. Section 3 shows that linguistic units of the kind
posited by construction-based grammars are accessed in the normal processes of lan-
guage production. Section 4 observes a basic advantage of parallel approaches such
as CxG: They allow for a fairly transparent interface simultaneously with both com-
prehension and production.
Although there has been a large amount of fruitful research in language acquisi-
tion from a constructional perspective, research in language processing from this per-
spective is only beginning. We believe that continued research in this area may well
aid in discovering new mechanisms at work in processing and ultimately may lead to
our grounding of linguistic theory in a theory of performance.
NOTES
1. See Jackendoff (1997, 2002) or Goldberg (2003) for more comprehensive summaries and discus-
sions of the differences between parallel constraint-based approaches to grammar (of which CxG is
an exemplar) and classical generative accounts. See Bencini (2002) for a discussion of the advan-
tages of parallel constraint-based grammars from a processing point of view.
2. The study by Bencini and Goldberg (2000) was inspired by a Healy and Miller (1970) sorting experi-
ment, which had been titled “The Verb as the Main Determinant of Sentence Meaning.” Healy and
Miller (1970) created stimuli by crossing subject arguments with verbs because they assumed that
the two best candidates for determining what the sentence was about were the verb and the subject
argument.
3. See also work by McRae, Ferretti, and Amyote (1997) and Ahrens (2003), who demonstrate that the
sorts of verb-specific participant roles posited by CxG (e.g., Goldberg 1995) are required to account
for on-line sentence processing.
4. At the same time, the data are a bit ambiguous because there is a stepward trend such that passives
prime passives and by-locatives prime passives (significantly), but more passives also were pro-
duced after non–by-locatives than after controls; the difference was nonsignificant, however, despite
the running of 130 subjects.
5. One interpretation of the Hare and Goldberg (1999) findings is that order of animate participants,
not the order of semantic roles, affected priming. This possibility is strengthened by the fact that
animacy has been shown to induce priming, even when the overall construction is held constant
(Bock, Loebell, and Morey 1992). In fact, it remains to be shown that ditransitive versus dative prim-
ing is not induced by differing order of animate participants as well (see Bencini 2002 for
discussion).
Yamashita, Chang, and Hirose (2003) have shown that dative sentences with the order AGENT-wa
RECIPIENT-ni PATIENT-o (wa-ni-o) prime other wa-ni-o ordered productions, even though the
animacy of recipients and patients was controlled for. Although these results suggest that structural
priming can be sensitive to the order of syntactic functions, thematic roles, or morphology independ-
ently of animacy, the experiment does not tease apart these possible interpretations.
REFERENCES
Abbott-Smith, Kirsten, Elena Lieven, and Michael Tomasello. 2004. Training 2;6-year-olds to produce the
transitive construction: The role of frequency, semantic similarity and shared syntactic distribution.
Developmental Science 7, no. 1:48–55.

Ahrens, Kathleen. 1995. The mental representation of verbs. Ph.D. diss., University of California at San
Diego.
———. 2003. Verbal Integration: The interaction of participant roles and sentential argument structure.
Journal of Psycholinguistic Research 22:497–516.
Anderson, Steven. R. 1971. On the role of deep structure in semantic interpretation. Foundations of Lan-
guage 6:197–219.
Anh, Woo-Kyong, and Douglas L. Medin. 1992. A two-stage model of category construction. Cognitive
Science 16:81–121.
Bates, Elizabeth, and Judith Goodman. 1997. On the inseparability of grammar and the lexicon: Evidence
from acquisition, aphasia, and real time processing. Language and Cognitive Processes 12, nos. 5–
6:507–84.
Bencini, Giulia M. L. 2002. The representation and processing of argument structure constructions. Ph.D.
diss., University of Illinois.
Bencini, Giulia, and Adele Goldberg. 2000. The contribution of argument structure constructions to sen-
tence meaning. Journal of Memory and Language 43:640–51.
Bencini, Giulia, Katheryn J. Bock, and Adele E. Goldberg. n.d. How abstract is grammar? Unpublished
manuscript.
Bock, Kathryn. 1986. Syntactic persistence in language production. Cognitive Psychology 18:355–87.
———. 1995. Sentence production: From mind to mouth. In Handbook of perception and cognition:
Speech, language and communication, vol. 11, ed. Joanne L. Miller and Peter D. Eimas, 181–216.
Orlando, Fla.: Academic Press.
Bock, Kathryn, and Zenzi M. Griffin. 2000. The persistence of structural priming: Transient activation or
implicit learning? Journal of Experimental Psychology-General 129:177–92.
Bock, Kathryn, and Wilhelm Levelt. 1994. Language production: Grammatical encoding. In Handbook of
psycholinguistics, ed. Morton Ann Gernsbacher, 945–84. San Diego: Academic Press.
Bock, Kathryn, and Helga Loebell. 1990. Framing sentences. Cognition 35:1–39.
Bock, Kathryn, Helga Loebell, and Randal Morey. 1992. From conceptual roles to structural relations:
Bridging the syntactic cleft. Psychological Review 99:150–71.
Branigan, Holly P., Martin J. Pickering, Simon P. Liversedge, Andrew J. Stewart, and Thomas P. Urbach.
1995. Syntactic priming: Investigating the mental representation of language. Journal of
Psycholinguistic Research 24:489–506.
Branigan, Holly P., Martin J. Pickering, Andrew J. Stewart, and Janet F. McLean. 2000. Syntactic priming
in spoken production: Linguistic and temporal interference. Memory and Cognition 28, no. 8:1297–
1302.
Chang, Franklin, Kathryn Bock, and Adele E. Goldberg. 2003. Do thematic roles leave traces in their
places? Cognition 90, no. 1:29–49.
Chang, Franklin, Garry S. Dell, Kathryn Bock, and Zenzi M. Griffin. 2000. Structural priming as implicit
learning: A comparison of models of sentence production. Journal of Psycholinguistic Research
29:217–29.
Chang, Nancy C., and Tiago V
. Maia. 2001. Learning grammatical constructions. Proceedings of the 24th
Annual Cognitive Science Society. Mahwah, N.J.: Lawrence Erlbaum Associates.
Childers, Jane B., and Michael Tomasello. 2001. The role of pronouns in young children’s acquisition of
the English transitive construction. Developmental Psychology 37:739–48.
Chomsky, Noam. 1981. Lectures on government and binding. Dordrecht, The Netherlands: Foris
Publications.
Croft, William. 2001. Radical construction grammar. Oxford: Oxford University Press.
Diessel, Holger, and Michael Tomasello. 2001. The acquisition of finite complement clauses in English: A
usage based approach to the development of grammatical constructions. Cognitive Linguistics
12:97–141.
Fillmore, Charles J. 1968. The case for case. In Universals in linguistic theory, ed. Emmon Bach and Rob-
ert T. Harms, 1–88. New York: Holt, Rinehart and Winston.
Garnsey, Susan, Neal J. Pearlmutter, Elizabeth Myers, and Melanie A. Lotocky. 1997. The contributions of
verb bias and plausibility to the comprehension of temporarily ambiguous sentences. Journal of
Memory and Language 37:58–93.

Garrett, Merril. F. 1980. Levels of processing in sentence production. In Language production, vol. 1, ed.
Brian Butterworth. London: Academic Press.
Gleitman, Lila, Henry Gleitman, Carol Miller, and Ruth Ostrin. 1996. Similar and similar concepts. Cog-
nition 3:21–76.
Goldberg, Adele. E. 1995. Constructions: A construction grammar approach to argument structure. Chi-
cago: Chicago University Press.
———. 2002. Surface generalizations: An alternative to alternations. Cognitive Linguistics 13:327–56.
———. 2003. Constructions: A new theoretical approach to language. Trends in Cognitive Sciences
7:219–24.
———. In press. Constructions in context. Oxford: Oxford University Press.
Green, Georgia. 1974. Semantics and syntactic regularity. Bloomington: Indiana University Press.
Griffin, Zenzi M., and Justin Weinstein-Tull. 2003. Conceptual structure modulates structural priming in
the production of complex sentences. Journal of Memory and Language 49:537–55.
Hare, Mary, and Adele E. Goldberg. 1999. Structural priming: Purely syntactic? Proceedings of the 22nd
Annual Cognitive Science Society. Mahwah, N.J.: Lawrence Erlbaum Associates.
Healy, Alice F., and George A. Miller. 1970. The verb as the main determinant of sentence meaning.
Psychonomic Science 20:372.
Jackendoff, Ray. 1997. Twistin’ the night away. Language 73, no. 3:534–59.
———. 2002. Foundations of language. Oxford: Oxford University Press.
Just, Marcel, and Patricia Carpenter. 1982. A model of the time course and content of reading. Cognitive
Science 6:157–203.
Kaschak, Michael P., and Arthur M. Glenberg. 2000. Constructing meaning: The role of affordances and
grammatical constructions in sentence comprehension. Journal of Memory and Language 43:508–
29.
Kay, Paul, and Fillmore, Charles J. 1999. Grammatical constructions and linguistic generalizations: The
what’s X doing Y? construction. Language 75:1–34.
Lakoff, George. 1970. Irregularity in syntax. New York: Holt Rinehart and Winston.
Lambrecht, Knud. 2001. A framework for the analysis of cleft constructions. Linguistics 39:463–516.
Langacker, Ronald W. 1988a. An overview of cognitive grammar. In Topics in cognitive linguistics, ed.
Brygida Rudzka-Ostyn. Philadelphia: John Benjamins Publishing Co.
———. 1988b. A view of linguistic semantics. In Topics in cognitive linguistics, ed. Brygida
Rudzka-Ostyn. Philadelphia: John Benjamins Publishing Co.
Levelt, Willem, and Kelter, Stephanie. 1982. Surface form and memory in question answering. Cognitive
Psychology 14:78–106.
Levin, Beth. 1993. English verb classes and alternations. Chicago: University of Chicago Press.
McRae, Ken, Todd Ferretti, and Liane Amyote. 1997. Thematic roles as verb specific concepts. Language
and Cognitive Processes 12:137–76.
Medin, Doug L., William D. Wattenmaker, and Sarah E. Hampson. 1987. Family resemblance conceptual
cohesiveness, and category construction. Cognitive Psychology12:242–79.
Michaelis, Laura A. 2001. Exclamative constructions. In Language universals and language typology: An
international handbook, ed. Martin Haspelmath, Ekkehard König, Wulf Österreicher, and Wolfgang
Raible, 1038–50. Berlin: Walter de Gruyter.
Michaelis, Laura A., and Lambrecht, Knud. 1996. Toward a construction-based model of language func-
tion: The case of nominal extraposition. Language 72:215–47.
Partee, Barbara H. 1965. Subject and object in modern English. New York: Garland.
———. 1971. On the requirement that transformations preserve meaning. In Studies in linguistic seman-
tics, ed. Charles J. Fillmore and D. Terence Langedoen, 1–21. New York: Holt, Rinehart and
Winston.
Pinker, Stephen.1989. Learnability and cognition: The acquisition of argument structure. Cambridge,
Mass.: MIT Press/Bradford Books.
———. 1994. How could a child use verb syntax to learn verb semantics? In The acquisition of the lexi-
con, ed. Lila Gleitman and Barbara Landau. Cambridge, Mass.: MIT Press.
Pollard, Carl, and Ivan Sag. 1994. Head-driven phrase structure grammar. Stanford, Calif., and Chicago:
CSLI and University of Chicago Press.

Potter, Mary C., and Linda Lombardi. 1998. Syntactic priming in immediate recall of sentences. Journal
of Memory and Language 38:265–82.
Rappaport-Hovav, Malka, and Beth Levin. 1996. Building verb meanings. In The projection of arguments,
ed. M. Butt and W. Gerder. Stanford, Calif.: Center for the Study of Language and Information.
Regehr, Glenn and Lee R. Brooks. 1995. Category organization in free classification: The organizing ef-
fect of an array of stimuli. Journal of Experimental Psychology: Learning, Memory and Cognition
21:347–63.
Rosch, Eleanor, and Carolyn B Mervis. 1975. Family resemblances: Studies in the internal structure of
categories. Cognitive Psychology 7:573–605.
Sag, Ivan, and Thomas Wasow. 1999. Syntactic theory. Stanford, Calif.: Center for the Study of Language
and Information.
Smith, Linda B. 1981. Importance of overall similarity of objects for adults’ and children’s classifications.
Journal of Experimental Psychology: Human Perception and Performance 7:811–24.
Smith, Mark, and Linda Wheeldon. 2001. Syntactic priming in spoken sentence production—An online
study. Cognition 78:123–64.
Tannen, Deborah. 1989. Talking voices: Repetition, dialogue and imagery in conversational discourse.
Cambridge: Cambridge University Press.
Tomasello, Michael. 2003. Constructing a language: A usage-based theory of language acquisition. Cam-
bridge, Mass.: Harvard University Press.
Weiner, Judith, and William Labov. 1983. Constraints on the agentless passive. Journal of Linguistics
19:29–58.
Wierzbicka, Anna. 1988. The semantics of grammar. Amsterdam: John Benjamins.
Yamashita, Hiroko, Franklin Chang, and Yuki Hirose. 2002. Separating functions and positions: Evidence
from structural priming in Japanese. Paper presented at 15th Annual Human Sentence Processing
Conference, City University of New York (March).
Zwicky, Arnold. 1994. Dealing out meaning: Fundamentals of syntactic constructions. Berkeley Linguis-
tics Society 20:611–25.

2
Homonyms and Functional Mappings in
Language Acquisition
DEVIN M. CASENHISER
Princeton University
Mapping Form and Meaning
Because language is used for communication, we might expect it to exist in and
evolve toward a state in which communication is optimally facilitated. Yet there are
some aspects of language whose existence contradicts the communicative purpose of
language. The case under consideration in this chapter involves mappings between
form and meaning. One might expect that sounds and meanings should be mapped to
each other with a one-to-one correspondence as in bi-uniqueness—each sound (or
string of sounds) having a single meaning associated with it and vice versa. This sort
of language would both minimize ambiguity and maximize the language’s communi-
cative potential. In fact, languages are not so perfectly composed. For example, there
are cases in which the mapping between sounds (or strings of sounds) and meanings
are many-to-one or one-to-many, as represented by examples B and C in figure 2.1.
The non–bi-unique mappings B and C create a potential problem for language
and communication. Synonymy (represented by B in figure 2.1) should be less pre-
ferred for reasons of economy. In principle, a language that reserves more than a sin-
gle word for a particular meaning creates the potential for redundancies that may
overwhelm a speaker’s memory. If two-to-one mappings are not dispreferred on
grounds of economy, then three-, four-, or five-to-one mapping might also be
created.
Where homophony (represented by example C in figure 2.1) is concerned, it is
not economy that degrades the communicative potential of language but ambiguity
resulting from the fact that there are too few forms to associate with each meaning.
When a speaker is confronted with a homophonous word, the speaker cannot rely on
the form of the word alone to determine its meaning. Instead, the speaker must inter-
pret its meaning from the context in which the word is encountered. If the context
does not effectively disambiguate the meaning of the homonym, the potential for
misunderstanding arises, which may make it more difficult for speakers to communi-
cate successfully.
If synonymy and homophony do degrade the communicative potential of lan-
guage, and language exists for communication, then both synonymy and homophony
should be dispreferred ways of mapping form to meaning. Scholars have long noted a
19

tendency for languages to avoid synonymy. Bréal (1897), for example, proposes a
law of differentiation in which several forms that are associated with a single mean-
ing come to distinguish themselves so that their meanings are no longer synonymous.
The law predicts, in other words, that however two words come to mean the same
thing, the language will evolve so that eventually the two words will no longer be
synonyms—thus preserving bi-uniqueness.
Once we have noted that languages seem to change to avoid synonymy, the next
step is to attempt to determine if there is a psychological or cognitive explanation for
this tendency. In the case of synonymy, language researchers have indeed discovered
evidence for a psychological tendency that may account for the rarity of synonymy in
language. The “disambiguation effect” (Merriman and Bowman 1989) occurs when
children are asked to find the referent of an unfamiliar object in a set of objects that
contains one familiar and one unfamiliar object. In such a situation, children nor-
mally choose an object whose name they do not know. To account for this effect (at
least in part), researchers have proposed several principles—including the Mutual
Exclusivity Principle (Markman 1992), Novel Name for Nameless Categories
(Golinkoff et al. 1992), the principle of contrast (Clark 1983), and Grice’s (1975) Co-
operation Principle (Gathercole 1989). Although there is disagreement about the ex-
act nature of the cognitive principle that is responsible for the disambiguation effect,
what is important for the present discussion is that children assume by default that a
many-to-one mapping between form and meaning is less likely than a one-to-one
mapping. Given that children have such an assumption, to possibly learn a synonym
they must first overcome their bias against many-to-one mappings.
Where a single form is mapped to many meanings, the case is not so clear. For
one thing, homophony is not uncommon in languages—especially if one considers
grammatical morphemes as well as words. Along these lines, Clark (1983, 70) points
out that the English –s morpheme is used for possessives, marking plural forms on
nouns and marking the third person singular form on present-tense verbs. Moreover,
polysemy, which involves different but related senses, seems to be the norm for both
lexical items and grammatical morphemes (Lakoff 1987; Taylor 1995). Nevertheless,
there is some research that suggests that at least homonyms, though not impossible,
may be dispreferred in lexical acquisition. Research on near-mergers, for example,
indicates that words that seem to be converging phonetically over time often will
change (or be changed) to make their differences more acoustically salient, or one of
the pair will be lost entirely (Labov 1994, esp. Part C). Research into processes of
phonological change has yielded similar conclusions. For example, Gurevich (2004)
found that among nearly 300 processes of lenition in 170 different languages, only
Figure 2.1. Three types of form-to-meaning mappings

about 15 percent resulted in a change that might potentially cause a form in the lan-
guage to become homophonous.1 Accordingly, one might expect that there is a corre-
sponding psychological constraint to account for this diachronic change, just as the
disambiguation effect accounts for Bréal’s Law of Differentiation. In other words, is
there evidence that, as Slobin (1973, 1977, 1985) argues, children do have a
one-to-one form-to-meaning assumption that extends to homophony as well as
synonymy?
In this chapter I raise several methodological issues that indicate that previous
research has not demonstrated this finding conclusively. I then describe an investiga-
tion that offers an empirical test that aims to eliminate previous confounds. Evidence
from this study suggests that in fact homonyms are dispreferred in lexical
acquisition.
Previous Research on Homophony
Campbell and Bowe-Macdonald (1977, 1983) conducted an experiment in which
they read stories containing the less familiar meanings of homonyms to children
(three- to five-year-olds). The experimenters then questioned the children about their
understanding of the homonyms and asked them to illustrate their referents. For ex-
ample, one story told about a wing of a castle:
At the far side of the wood was a castle. “Look at this castle,” said Jane’s
Daddy. “The oldest wing is over 500 years old.”
The children’s illustrations of this passage often depicted the castle’s wing like that of
a bird or airplane sticking out of a building. The results of the experiment indicated
that 31 percent of the children in all age groups responded by giving the primary,
more familiar meaning of the homonym in spite of the fact that clinging to the pri-
mary meaning of the homonym caused the story to seem bizarre or fanciful.2
The design of the experiment does not allow us to draw the conclusions we need,
however. First, the experiment did not take into account children’s previous experi-
ence with the homonyms used in the experiment. The number of children who cor-
rectly interpreted the homonyms may be related to individual children’s familiarity
with a particular homonym; on the other hand, children’s failure to interpret hom-
onyms correctly may result from a lack of previous experience with the secondary
meaning of the word. Moreover, the children who failed to interpret the homonym
correctly might not have been provided with enough context to fully disambiguate
the two meanings (Donaldson 1978).
Furthermore, where homonym acquisition is concerned, the task requires that
participants who are unfamiliar with the secondary meaning of the homonym be
willing to override their present understanding of a word’s meaning to accept the ad-
dition of some new understanding. Children simply may require more evidence to
change the established meaning of a word than they need to map an entirely new
word and its meaning (Keil 1991). In this case, children might not be receiving
enough evidence (linguistic and/or extralinguistic) to necessitate a modification of
their established mapping. Moreover, the fact that homonyms are used in the study
without a baseline for lexical acquisition prevents us from determining if there was
HOMONYMS AND FUNCTIONAL MAPPINGS IN LANGUAGE ACQUISITION 21

enough context for a child to map even a nonhomophonous word. The experiment
demonstrates only that young children will sometimes make fanciful or bizarre inter-
pretations of stories to try to make the context fit with what they already know.
To address these problems, Mazzocco (1997) conducted an experiment with
“pseudohomonyms.” A pseudohomonym is a word that is applied to an object other
than the one commonly associated with the word. For example, labeling a clown with
the word “door” makes “door” a pseudohomonym. In the study, participants (thirty-
two preschoolers, thirty-two second graders, and sixteen college students) were read
a story and then shown a page containing six illustrations. The story incorporated one
of three word types: pseudohomonyms, nonsense words, and familiar words (used
correctly). The context of the story was crafted to indicate a plausible meaning for
the target word. One story, for example, takes place at a birthday party: “then James
saw that a door was standing there making funny faces and doing tricks. James
laughed, because the door looked so funny.” After the participants listened to the
story, each was shown a page containing six illustrations: the keyword’s familiar
meaning (i.e., a door); the keyword’s meaning in the context (clown); something re-
lated to the context (e.g., cake); and three unrelated objects. The participant was then
asked to “look at all of the pictures on this page, then show me the picture of the key-
word in the story.” The results are reproduced in table 2.1.
Because Mazzocco’s (1997) study used nonsense words as well as pseudo-
homonyms, the results allow a point of comparison between the acquisition of
homophonous and nonhomophonous words. The relatively high success with which
preschool children were able to interpret the correct meaning of nonsense words
from the context compared to the much lower rate of success for the same task with
pseudohomonyms seems to indicate that a one-to-one mapping assumption may be at
work in language acquisition with young children. Preschoolers in particular seem
likely to base their interpretations of a homonym on their preexisting notions of what
that word means rather than on the context in which the word is heard.
The design of the study, however, introduces the problem of synonymy into the
acquisition task. The confusion lies in the fact that both the pseudohomonyms and
their referents are words the children already know. Consider the situation of the task:
The child hears a story in which what must be a clown, according to the context, is
discussed, but the story is labeling the clown with a different name—in this case
door. The child knows the meaning of both clown and door but is being asked to use
both meanings interchangeably: clown means CLOWN; door means DOOR and
CLOWN.
Table 2.1.
Mean number of correct responses from Mazzocco (1997)
Type of key word
Age Group Accurately used Nonsense Pseudohomonym
Preschool 5.81 (0.40) 4.16 (1.61) 0.75 (0.98)
Note: N = 32; maximum correct = 6; standard deviation appears in parentheses.

It is true, however, that children do not seem to have the same difficulties when a
nonsense word is used to refer to the object discussed in the story (twenty-three pre-
schoolers got four or more nonsense word interpretations correct, compared with one
preschooler who got four or more homonym interpretations correct). The existence
of synonymy in the task leaves the interpretation of the processes involved somewhat
uncertain, however. Consider that the nonsense word condition could be regarded as
differing from the homonym condition in at least two ways that are relevant to lexical
acquisition. First, although subjects know that the object being discussed is a clown,
it is being given a new label in the nonsense condition; hence, the new label functions
as a synonym for “clown.” In the pseudohomonym condition, children must not only
accept two labels for the same object (synonymy); they must accept a label that al-
ready refers to another object (homophony) as well. If we can simply subtract the ef-
fect of synonymy from both tasks, we are left with the conclusion that the difference
in participants’ performance is related to homophony alone. This conclusion, how-
ever, requires one to assume that the effect of violating these language constraints is
additive. It may be, however, that concurrent violation of two constraints on language
acquisition has an effect that is more or less than additive. This nonadditive effect
may obtain especially if the difference between participants’ performance on the ac-
curately used words and the nonsense words did not differ significantly—an infer-
ence the reader is left to make because the comparison is not specifically reported by
Mazzocco (1997). Moreover, the study also does not provide an accurate baseline we
might use to quantify the relative difficulty in acquiring a novel word and meaning
pair compared with the difficulty in mapping a familiar word to a second (i.e.,
homophonous) meaning.
The study I describe below seeks to address the foregoing questions and prob-
lems. An experiment was prepared to make a direct comparison between the acquisi-
tion of nonsense and homophonous words. The experiment also was designed to
determine the extent to which the linguistic context and the visual context of a scene
affect the acquisition of homonyms in comparison to nonhomonyms. The expected
outcome is that children will be at least partially successful in mapping homonyms to
new objects but that they will be significantly more successful in mapping nonsense
words to new objects. The hypothesis is that differentiation provided by linguistic
and visual context will aid children’s acquisition of homophonous words.
Figure 2.2. Comparison of many-to-one mapping with one-to-many mapping

Experiment One
Method
Participants The participants were sixteen four-year-olds (mean age ⫽ 4;6) from the
Champaign-Urbana area of Illinois. All children were monolingual native speakers
of English.
Materials A children’s story was written in which the main character, Tommy, was
searching for his baseball in each of five different scenes. Each of the five scenes had
an accompanying illustration: a park, a pond, the front yard of a house, a kitchen, and a
bathroom. Each scene included four landmarks (e.g., a pink tree, a boat, a bathtub, a
stove); two novel creatures; two novel objects; and three known objects or animals, two
of which were the same kind of object (e.g., two different kinds of dogs). All illustra-
tions were in color and were similar in appearance to those in a children’s picture book.
An “answer” page reproduced the objects from the illustration out of context; children
pointed to these illustrations to answer the experimenter’s questions. A sketch of the il-
lustrations used for Scene One of the story is included in figure 2.3.
The Stories Children heard a story about a little boy who was searching for his lost
baseball. The landmarks in the scenes corresponded to each of the specific places
where he looked for the baseball. In the course of his searching, the little boy encoun-
tered each of the target objects in the illustration. The target object is named in refer-
ence to the landmark with which it appears (e.g., “Tommy look under the slide, but
all he saw there was a snake.”), and the little boy engages in some activity specifically
associated with the landmark so that the children’s attention is unambiguously drawn
to that place (e.g., “Tommy thought the slide looked like fun, so he climbed up and
slid down it. ‘Whee!’ Tommy giggled.”). The portion of the story immediately fol-
lowing the target word was approximately twenty-four words long (plus or minus
two). The illustration of the story was then covered up and the answer page shown to
the child. The child is reminded that “Tommy saw a snake under the slide, right?” and
then asked, “Can you point to it?”
Target Words and Objects There were a total of twenty-five high-frequency, monosyllabic
target words and twenty-five target objects in the experiment. The words and targets
are divisible into five categories each, according to table 2.2.
Each of the seven objects noted in the last column of table 2.2 appear in the illus-
tration of the story (see figure 2.3). These same seven objects, plus the object named
by the noun-absent condition, appears on the answer sheet of illustrations—making a
total of eight illustrations, arranged in a 2-by-4 configuration. With the exception of
the vague condition, the objects to which the words referred were counterbalanced
across participants so that a purple creature, for example, was called a fidge in the
story for the first participant, a pie for the second participant, a snake for the third
participant, and so on. The order of conditions for each scene was counterbalanced
across trials.

Figure 2.3. Sketches of the illustrations used in experiment one

Explanation of Conditions Two baseline conditions were included in the experiment: the
nonsense and vague conditions. The nonsense condition provides a baseline for com-
parison between acquiring homophonous and nonhomophonous words. The vague
condition provides a way to gauge how well children are paying attention to the object
named in the specific location. Failure in the vague condition might indicate that chil-
dren were looking for any such object named by the experimenter—not necessarily the
particular object associated with the location in the story. The vague condition is an es-
pecially appropriate comparison for the two noun homonym conditions. Both condi-
tions require that the child attend to and remember the object referred to in the story
and in the question, but the noun-present condition requires the additional task of cre-
ating a new form-to-meaning mapping. Differences between performance on these
two tasks can be attributed to the additional task in the noun-present condition. A com-
parison of the noun-present and noun-absent conditions will illustrate the effect of
context (in this case, visual context) on children’s acquisition of homonyms. Finally,
the verb condition was included to determine if linguistic context might play a role in
homonym acquisition. The main comparison for the verb condition is the noun-absent
condition because neither condition has a corresponding object that appears in the il-
lustration. Children are expected to perform best on the vague and nonsense word con-
ditions. The verb condition should be the next easiest task, followed by the noun-ab-
sent condition. The most difficult task is expected to be the noun-present condition.
Table 2.3 lists words and landmarks in the five conditions in each of the five senses.
Procedure Children were tested individually and told that the experimenter was going
to read a story to them while they looked at some pictures of the story. The instruc-
tions were as follows: “I’ve got some pictures for you to look at, and I’m going to
read a story about the pictures. Once in a while I’m going to stop and ask you a
Table 2.2.
Target words and corresponding objects
Word
Condition Target word type example Target object type Object example
Vague Linguistically dog Pair of objects of Two different
vague word the same category dogs
Nonsense Nonsense word fidge Unknown object Novel object or
or creature creature
Noun-Present Noun homonym pie Novel object or creature Novel object or
present in the plus an object that creature plus
scene corresponds to the an illustration
meaning of the of a pie
pseudohomonym
Noun-Absent Noun homonym snake Unknown object or Novel object or
absent from scene creature creature
Verb Verb homonym give Unknown object or Novel object or
creature creature

Table 2.3.
List of words and landmarks
Nonsense Nouns- Nouns- Vague
Landmarks words Present Absent Verbs words
Scene One: fidge pie snake give dog
fountain
slide (under)
slide (on)
(pink) tree
flowers
Scene Two: goot cake owl have cat
bridge
boat
island
flowers
water
Scene Three: snock horse bone put pig
wall
roof
car
tree
front door
Scene Four: sarn boat leaf get fish
refrigerator
sink
oven
table
chair
Scene Five: mish sheep kite sit duck
bathtub
stool
rug
door
toilet

question that only kids know the answer to—that’s why I need your help—grown-ups
don’t know the answers, but I’ll bet you do! Are you ready?” Children were shown a
warm-up scene and read a short story in which they were asked to identify several ob-
jects in the picture that the character, Jan, saw in the park. Half of the word-object
pairs were of the vague type, to reinforce the idea that children needed to pay atten-
tion to the location in which an object appeared.
After the warm-up scene, children listened to the story until the point at which
the first landmark and target object were named:
Next to the slide, Tommy saw a pink tree. Way up in the tree, Tommy saw a
[keyword]. “Maybe my baseball is stuck up in the tree too,” Tommy thought.
Tommy climbed the tree and looked all around, but still didn’t find his
baseball.
At this point, the scene was covered up and the answer page was presented to the
child. The experimenter then instructed the child as follows: “Tommy saw a [key-
word] in the tree, right? Can you point to it?” The experimenter noted the child’s an-
swer, removed the answer page, uncovered the scene, and continued the story. A fi-
nal, sixth scene was included at the end so that Tommy could finally find his
baseball, but no questions were asked for this last scene.
Scoring The experimenter recorded participants’ answers on a prepared answer form.
Answers were scored as ‘correct’ if the participant indicated the object to which the
story referred. ‘Incorrect’answers were those in which the participant indicatedeither
the familiar meaning of the homonym (in the case of the Noun-Present and
Noun-Absent conditions) or the referent which appeared in a place in the scene other
than where the story indicated (in the case of the Vague condition). Answers were
scored as ‘other’when participants indicated an object that did not fit either category.
Results
A repeated measures analysis of variance (ANOVA), with the dependent variable as
the number of correctly mapped words, was performed for Experiment One. Signifi-
cant differences were found between the two noun conditions and the other three con-
ditions (F(4, 15) ⫽ 33.32, p ⬍ .001), but there were no significant difference be-
tween the noun-present and noun-absent conditions, nor among the three other
conditions (see figure 2.4).
Scores support the main effect of homonyms anticipated by this study. Overall,
participants mapped 77.5 percent of the vague words and 72.5 percent of the non-
sense words to the correct referent. By comparison, only 11.25 percent and 17.5
percent of the homonyms in the noun-present and noun-absent conditions were
mapped correctly. In the case of the noun homonym conditions, the majority of the
incorrect responses (75 percent) indicated the familiar meaning of the homonym.
The scores also support the assertion that verb homonyms are easier to map to object
referents than are noun homonyms. Participants correctly mapped 76.25 percent of
the verb homonyms, in contrast to the much lower scores on the noun homonym

conditions. The difference between the verb and nonsense conditions is not sig-
nificant, however. Likewise, there is only a slight, nonsignificant difference between
the noun-present and noun-absent conditions. Complete numbers are reported in
table 2.4.
Comparison to Chance Performance If performance were random, the probability that a par-
ticipant would respond correctly to any single item is .125 because there are eight
possibilities and only one correct answer. Because each participant was asked to an-
swer five questions of each condition type, a score of three or better would indicate
that the participant was performing significantly above chance (binomial p ⫽ 0.015).
Results are reported in table 2.5.
Figure 2.4. Mean number of correct responses
Table 2.4.
Participant responses as a function of condition and response typea
Condition Correct (%) Distractor (%) Other (%)
Nonsense 72.5 (58) 27.5 (22)
Vague 77.5 (62) 21.25 (17) 1.25 (1)
Verb 76.25 (61) 23.75 (19)
Noun-Present 17.5 (14) 81.25 (65) 1.25 (1)
Noun-Absent 11.25 (9) 86.25 (69) 2.5 (2)
aN = 16. There were five items for each condition. Raw numbers are given in parentheses;
the maximum number for each cell is 80.

The scores from table 2.5 demonstrate that participants consistently failed to
map noun homonyms to the novel objects they encountered, despite the fact that they
were able to map nonsense words as well as verb homonyms to these same objects.
Discussion
The results concerning the learning of noun homonyms from Experiment One sug-
gest that children disprefer homophony during lexical acquisition even when they are
faced with evidence that a new meaning is appropriate. There also is evidence that
the linguistic context—particularly part of speech—may play a prominent role in dis-
ambiguating the context for the child during acquisition. This conclusion is sup-
ported by the fact that children mapped verb homonyms as easily as they mapped
nonsense words.
The design of Experiment One is biased, however, in favor of the verb homonym
condition compared with the noun homonym conditions because a referent for the pri-
mary meaning of each noun homonym appeared on the answer sheet, whereas the verb
homonyms did not have corresponding referents for the children to point to (e.g., there
was no illustration for the verb sit). To attempt to correct this bias, a second study was
designed in which an initial “go/no-go” decision was added to the existing story task.
That is, children were first asked to determine if the object Tommy saw in the story ap-
peared in the illustration they were looking at. If they saw the keyword in the illustra-
tion, they were to point it out on the answer sheet. Otherwise, they were to indicate that
the keyword was not in the illustration. If the difference between the noun and verb
homonym conditions was a result of the fact that an illustration for the verb referent
was missing on the answer sheet (as in Experiment One), the number of no-go re-
sponses in the noun homonym and verb homonym conditions should be roughly the
same. On the other hand, if the result was caused by part-of-speech differences, Exper-
iment Two should show a significantly greater number of no-go responses for the noun
homonyms than for the verb homonyms. Because the illustrations of the scenes con-
tain neither pictures of the noun homonyms nor pictures of the verb homonyms, any
difference that does exist between the number of no-go responses for the noun-hom-
onym and verb-homonym conditions might be accounted for by part of speech.
Experiment Two
Method
Participants Participants were eighteen four-year-old subjects (mean age 4;5) from the
Champaign-Urbana area. All subjects were native speakers of English.
Table 2.5.
Participants performing significantly above chance (N = 16)
Nonsense Vague Verb Noun-Absent Noun-Present
Number 15/16 15/16 15/16 1/16 1/16
% of Total 93.7 93.7 93.7 6.25 6.25

Materials The illustrations and story from Experiment One were modified so that only
four objects appeared in each illustration, and all four objects were named in the cor-
responding story. The answer sheets included eight objects, as in Experiment One.
Explanation of Conditions Experiment Two included noun, verb, and nonsense conditions
that were identical to the noun-absent, verb, and nonsense conditions in Experiment
One. In addition, there was a foil condition in which a noun was named that did not
appear in the illustration of the story. Order of word type was counterbalanced across
trials. The answer sheet included the three novel objects or creatures that were re-
ferred to in the first three conditions, an object representing the primary meaning of
the noun homonym, the object named in the foil condition, and three novel objects
that did not appear in the illustration.
Procedure The procedure was the same as with Experiment One except that children
were instructed to first determine whether the object Tommy saw in the story ap-
peared in the illustration. For example, children heard the following portion of the
story:
Next to the slide, Tommy saw a pink tree. Way up in the tree, Tommy saw a
[keyword]. “Maybe my baseball is stuck up in the tree too,” Tommy thought.
Tommy climbed the tree and looked all around, but still didn’t find his
baseball.
The experimenter then covered the illustration of the story and said, “Now, Tommy
saw a [keyword] in the story. Did you see one too?” If the child answered yes, the
child was presented with the answer sheet and asked to point to it. If the child re-
sponded no, the experimenter marked the answer as a no-go and proceeded to the
next part of the story.
Results
Statistics were first run to compare the number of no-go answers in each of the four
conditions. Repeated measures ANOVA with post hoc comparisons were performed
using word type (foil ⫻ noun ⫻ verb ⫻ nonsense) as the within-subjects factor. Tests
show that all conditions differ significantly at or below the .01 level. In particular, a
comparison between the noun (M ⫽ 2.89, SD ⫽ 1.23) and the verb (M ⫽ 1; SD ⫽
0.69) conditions reveals a significant difference (F(1,17) ⫽ 32.32; p ⬍ .001), indicat-
ing that children rejected noun homonyms more readily than verb homonyms. The
mean number of no-go answers for each condition is reported in figure 2.5.
Statistics also were run to compare participants’ success at mapping each of the
word types to their object referents. For the purposes of scoring, no-go responses as
well as responses that indicated an object other than the intended referent of the story
were considered incorrect. Again, repeated measures ANOVA with post hoc compar-
isons were performed, using word type (noun ⫻ verb ⫻ nonsense) as the within-sub-
jects factor. The tests found that all conditions differed significantly at or below the p
⬍ .001 level. Results are reported in figure 2.6.

The line graph in figure 2.6 shows the trend of increased difficulty in mapping
each word type. As with Experiment One, children are significantly less willing to
map a new meaning to a noun homonym than to a verb homonym. Unlike Experi-
ment One, however, this experiment suggests that children do experience more diffi-
culty mapping verb homonyms than nonsense words (F(1,17) ⫽ 17, p ⬍ .001).
Discussion
The primary goal of this investigation was to determine if children disprefer
homophony during lexical acquisition. The results of Experiment Two indicate that
when children encounter a lexical item for which they have an established meaning,
they will prefer that meaning even when they are faced with evidence indicating that
a different meaning is appropriate. The point is most obvious when one considers the
data concerning the mapping of noun homonyms to new meanings, but Experiment
Two indicates that even verb homonyms are significantly more difficult to map to
new meanings compared with mapping nonsense words to those same meanings.
This conclusion is further supported by children’s comments during the experiment
that indicated that they were resistant to accepting a second meaning for a word. To
cite one of several examples: After hearing a sentence that said a snake was under the
slide, B.S. remarked, “Where? I don’t see a snake! [looks closely at the picture] Oh, I
know! Maybe the snake is behind this thing! [points to the novel object under the
slide].”
The results also support the hypothesis that it is easier to map new meanings
onto verb homonyms than onto noun homonyms. Children’s comments during the
experiment indicate that they treated verb homonyms as novel words rather than as
Figure 2.5. Mean number of no-go responses

homonyms, sometimes asking questions such as, “Sit? What’s a sit?” (O.J.). Because
children’s performance on the verb homonyms also is significantly better than their
performance on noun homonyms, it may be that the linguistic context—specifically
part of speech—plays a prominent role in disambiguating the context for the child
during acquisition. Exactly what portion of this effect is related to the syntactic or se-
mantic differences associated with different parts of speech must be determined by
further study.
The third hypothesis in this investigation was that the presence or absence of a
referent for the primary meaning of the homonym in the illustration accompanying
the story would influence children’s willingness to map a new meaning to the hom-
onym. The results did not support this hypothesis. There was a slight but non-signifi-
cant difference, however, between the noun-present and noun-absent conditions,
which might indicate that the design of this experiment was not sensitive enough to
illuminate the differing effects of these two conditions.
Finally, it should be noted that this experiment investigates only one type of
one-to-many mapping—one in which meanings are unrelated. Another type of
one-to-many mapping is polysemy, which involves two or more different but related
meanings. This experiment does not manipulate the relatedness of the primary refer-
ents of the homonyms and the novel objects introduced because the experimenters
consistently used two unrelated meanings. Therefore it is impossible to draw any
conclusions about the acquisition of polysemes per se. Furthermore, polysemes rep-
resent a different situation, in which the primary meaning of the word may actually
aid in learning extended meanings for the same word. This situation would obtain if
theorists who believe that there are definite patterns for defining extended meanings
Figure 2.6. Mean number of incorrect responses

of words are correct (Langacker 1987, 1991; Lakoff 1987). If such patterns exist and
children do learn them, it may be possible to learn new meanings for words that have
been defined by one of the learned patterns. Precious little research (Gropen, Ep-
stein, and Schumacher 1997) has been done on the acquisition of polysemes, so any
proposal about how polysemes are learned remains speculative. Nonetheless, this
topic may shed a good deal of light on the way we learn the meanings of words and
therefore is worthy of investigation.
Conclusion
The research described in this chapter supports the idea that communicative goals of
language are reflected in learning biases. I have demonstrated that children disprefer
learning a different, unrelated meaning for a known word when that word is used in a
linguistic context that fails to bias strongly for a new meaning. Children appear to
have much less difficulty in learning homonyms when the syntactic context clearly
indicates that a new meaning is required (cf. the verb homonym condition verses the
two noun homonym conditions). I do not suggest that children are incapable of learn-
ing homonyms; clearly that is not the case. Nonetheless, the experiment presented in
this chapter does demonstrate a bias against homophony in the acquisition process.
Acknowledgments
I am very grateful for the advice and guidance of Adele Goldberg in the planning
stages of this study as well as in the preparation of this chapter. I am also indebted to
Dan Silverman, whose thought-provoking phonology courses planted the initial
seeds for this research. Finally, I thank the teachers, students, and parents at Country-
side and Next Generation schools for assisting with and participating in this study.
This chapter is an abbreviated version of “Children’s resistance to homonymy: An
experimental study of pseudohomonyms,” which appears in the Journal of Child
Language, 32, no. 1 (2005), © Cambridge University Press.
NOTES
1. Gurevich (2004) did not verify that the changes did result in homophony of any kind.
2. Compare this result with the 25 percent of all participants who responded with the correct meaning
of the homonym (the remaining 44 percent of participants invented answers, could not respond, or
gave answers the experiments could not classify).
REFERENCES
Bréal, Michel. 1897. Essai de sémantique. Paris: Hachette.
Campbell, Robin N., and Theresa Bowe. 1977. Functional asymmetry in early language understanding. In
Salzberger Beitrage für Linguistik, ed. Gaberell Drachman. Tubingen, Germany: Gunter Narr.
Campbell, Robin N., and Theresa Bowe-Macdonald. 1983. Text and context in early language comprehen-
sion. In Early child development and education, ed. Margaret Donaldson, Robert Grieve, and Chris
Pratt, 115–26. Oxford: Blackwell.
Clark, Eve. 1983. Meanings and concepts. In Handbook of child psychology, 4th ed., vol. 3: Cognitive de-
velopment, ed. John H. Flavell and Ellen M. Markman, 787–840. New York: John Wiley and Sons.
Donaldson, Margaret C. 1978. Children’s minds. London: Fontana.
Gathercole, Virginia C. 1989. Contrast: A semantic constraint? Journal of Child Language 16:685–702.

Golinkoff, Roberta M., Kathy Hirsch-Pasek, Leslie M. Bailey, and Neill R. Wenger. 1992. Young children
and adults use lexical principles to learn new nouns. Developmental Psychology 28:99–108.
Grice, H. Paul. 1975. Logic and Conversation. In Syntax and Semantics 3: Speech Acts, ed. Peter Cole and
Jerry L. Morgan, 41–58. New York: Academic Press.
Gropen, Jess, Trina Epstein, and Lisa Schumacher. 1997. Context-sensitive verb learning: Children’s abil-
ity to associate contextual information with the argument of a verb. Cognitive Linguistics 8, no.
2:137–82.
Gurevich, Naomi. 2004. Lenition and contrast: The functional consequences of certain phonetic condi-
tioned sound changes. New York: Routledge.
Keil, Frank C. 1991. Theories, concepts, and the acquisition of word meaning. In Perspectives on lan-
guage and thought: Interrelations in development, ed. Susan Gelman and James P. Byrnes. New
York: Cambridge University Press.
Labov, William. 1994. Principles of linguistic change: Internal factors. Cambridge, Mass.: Blackwell.
Lakoff, George. 1987. Women, fire, and dangerous things. Chicago: University of Chicago Press.
Langacker, Ronald. 1987. Foundations of cognitive grammar, vol. 1. Stanford, Calif.: Stanford University
Press.
———. 1991. Foundations of cognitive grammar, vol. 2. Stanford, Calif.: Stanford University Press.
Markman, Ellen. 1992. Constraints on word learning: Speculations about their nature, origins and domain
specificity. In Modularity and constraints in language and cognition. The Minnesota symposia on
child psychology, ed. Megan R. Gunnar and Michael P. Maratsos. Hillsdale, N.J.: Lawrence Erlbaum
Associates.
Mazzocco, Michèle M. M. 1997. Children’s interpretations of homonyms: A developmental study. Journal
of Child Language 24:441–67.
Merriman, William E., and Laura L. Bowman. 1989. The mutual exclusivity bias in children’s word learn-
ing. Monographs of the Society of Research in Child Development 54, nos. 3–4 (serial no. 220): 1–
123.
Slobin, Dan. I. 1973. Cognitive prerequisites for the development of grammar. In Studies of child language
development, ed. Dan. I. Slobin. New York: Springer.
———. 1977. Language change in childhood and in history. In Language learning and thought, ed. John
T. Macnamara, 185–214. New York: Academic Press.
———. 1985. Crosslinguistic evidence for the language-making capacity. In A Crosslinguistic study of
language acquisition, ed. D. I. Slobin, 1157–1256. Hillsdale, N.J.: Lawrence Erlbaum Associates.
Taylor, John R. 1995. Linguistic categorization. Oxford: Oxford University Press.

Other documents randomly have
different content

Mey and Jorisz in 1623, who built the Dutch Fort Nassau a short
distance below Philadelphia; but it did not last.
Delaware Bay is an expansive inland sea, subject to fierce storms,
and broadening on its eastern side into Maurice River Cove, noted for
its oysters. A deep ship channel conducts commerce through the
centre of the bay, marked by lighthouses built out on mid-bay shoals,
and, as the shores approach, by range lights on the banks, the
Delaware Bay and River being regarded as the best marked and
lighted stream in the country. Up at the head of the bay, years ago, a
ship loaded with peas and beans sank, and this in time made at first
a shoal, and afterwards an island, since known as the "Pea Patch."
Here and on the adjacent shores the Government has lately erected
formidable forts, which make, with their torpedo stations in the
channel, a complete system of defensive works in the Delaware, first
put into active occupation during the Spanish War of 1898, as a
protection against a hostile fleet entering the river. Over in the
"Diamond State" of Delaware, near here, on the river shore, is the
aged town of Newcastle, quiet and yet attractive, having in operation,
and evidently to the popular satisfaction, the whipping-post and
stocks, a method of punishment which is a terror to all evil-doers,
and is said to be most successful in preventing crime, as thieves and
marauders give Newcastle a wide berth. This was originally a Swedish
settlement, the standard of the great Gustavus Adolphus being
unfurled there in 1640, when it was called Sandhuken, or Sandy
Hook, it being a point of land jutting out between two little creeks.
The Dutch soon captured it, changing the name to New Amstel; and
about 1670 the settlement, then containing nearly a hundred houses,
became New Castle, under English auspices. The northern boundary
of the State of Delaware, dividing it from Pennsylvania, is an arc of a
circle, made by a radius of twelve miles described around the old
Court House at Newcastle, which still has in its tower the bell
presented by Queen Anne.
MASON AND DIXON'S LINE.

In coming over by railroad from the Chesapeake to the Delaware, the
train, after crossing the broad Susquehanna and the head of Elk, and
rounding in Maryland the Northeast Arm of Chesapeake Bay, soon
enters the State of Delaware near the northeastern corner of the
former State. This corner is at the termination of the crescent-shaped
northern boundary of Delaware. The northern boundary of Maryland
here beginning and laid down due west, to separate it from
Pennsylvania, is the famous "Mason and Dixon's Line," surveyed by
Charles Mason and Jeremiah Dixon, two noted English
mathematicians and astronomers in the eighteenth century. This
boundary gained great notoriety because it so long marked the
northern limit of slavery in the United States. For almost a century
there were conflicts about their respective limits between the rival
proprietaries of the two States, producing sometimes riot and
bloodshed, until, in 1763, these men were brought over from
England, and in December began laying out the line on the parallel of
latitude 39° 43' 26.3" North. They were at the work several years,
surveying the line two hundred and forty-four miles west from the
Delaware River, and within thirty-six miles of the entire distance to be
run, when the French and Indian troubles began, and they were
attacked and driven off, returning to Philadelphia in December, 1767.
At the end of every fifth mile a stone was planted, graven with the
arms of the Penn family on one side and of Lord Baltimore on the
other. The intermediate miles were marked by smaller stones, having
a P on one side and an M on the other, all the stones thus used for
monuments being sent out from England. After the Revolution, in
1782, the remainder of the line was laid down, and in 1849 the
original surveys were revised and found substantially correct.
When the little colony of Swedes and Finns under Peter Minuet came
into Christiana Creek in April, 1638, and established their fort, they
began the first permanent settlement in the valley of the Delaware. It
was built upon a small rocky promontory, and they named it
Christina, in honor of the daughter of Gustavus Adolphus. The Dutch
afterwards captured it and called it Fort Altena; but the town retained
part of the original name in Christinaham, and the creek also retained

the name, the English taking possession in 1664. The Swedes,
however, regardless of the flag that might wave over them, still
remained; and their old stone church, built in 1698, still stands, down
near the promontory by the waterside, in a yard filled with time-worn
gravestones. This old Swedes' Church of the Holy Trinity, the oldest
now on the Delaware, was dedicated on Trinity Sunday, 1699, and
Rev. Ericus Tobias Biorck came out from Sweden to take charge as
rector. It was sixty by thirty feet and twenty feet high, and a little bell
tower was afterwards added. The ancient church was recently
thoroughly restored to its original condition, with brick floor, oaken
benches, and stout rafters supporting the roof. This interesting
church building is in a factory district which is now part of
Wilmington, the chief city of Delaware, a busy manufacturing
community of sixty-five thousand people, built on the Christiana and
Brandywine Creeks, which unite about a mile from the Delaware. This
active city was laid out above the old settlement, in 1731, by William
Shipley, who came from Leicestershire, England. Ships, railway cars
and gunpowder are the chief manufactures of Wilmington. The
Brandywine Creek, in a distance of four miles, terminating in the city,
falls one hundred and twenty feet, providing a great water power. Up
this stream are the extensive Dupont powder-mills, among the
largest in the world, founded by the French statesman and
economist, Pierre Samuel Du Pont De Nemours, who, after the
vicissitudes of the French Revolution, migrated with his family to the
United States in 1799, and was received with distinguished
consideration. He afterwards was instrumental in securing the treaty
of 1803 by which France ceded Louisiana, and was in the service of
Napoleon, but finally returned to America, where his sons were
conducting the powder-works, and he died near Wilmington in 1817.
Admiral Samuel Francis Dupont, of the American Navy, was his
grandson. Farther up the Brandywine Creek, at Chadd's Ford and
vicinity, was fought, in September, 1777, the battle of the
Brandywine, where the English victory enabled them to subsequently
take possession of Philadelphia.

WILLIAM PENN.
Above Wilmington, the Delaware River is a noble tidal stream of
about a mile wide, flowing between gently sloping shores, and
carrying an extensive commerce. The great river soon brings us to
the famous Quaker settlements of Pennsylvania. William Penn, who
had become a member of the Society of Friends, was bequeathed by
his father, Admiral Sir William Penn, an estate of £1500 a year and
large claims against the British Government. Fenwick and Byllinge,
both Quakers, who had proprietary rights in New Jersey, disputed in
1674, and submitted their difference to Penn's arbitration. He decided
in favor of Byllinge, who subsequently became embarrassed, and
made over his property to Penn and two creditors as trustees. This
seems to have turned Penn's attention to America as a place of
settlement for the persecuted Quakers, and he engaged with zeal in
the work of colonization, and in 1681 obtained from the king, for
himself and heirs, in payment of a debt of £16,000 due his father, a
patent for the territory now forming Pennsylvania, on the fealty of the
annual payment of two beaver skins. He wanted to call his territory
New Wales, as many of the colonists came from there, and
afterwards suggested Sylvania as specially applicable to a land
covered with forests; but the king ordered the name Pennsylvania
inserted in the grant, in honor, as he said, of his late friend the
Admiral. In February, 1682, Penn, with eleven others, purchased
West Jersey, already colonized to some extent. Tradition says that
some of these West Jersey colonists sent Penn a sod in which was
planted a green twig, to show that he owned the land and all that
grew upon it. Next they presented him with a dish full of water,
because he was master of the seas and rivers; and then they gave
him the keys, to show he was in command and had all the power.

Penn's Letitia Street House, Removed to
Fairmount Park
When William Penn was granted his province, he wrote that "after
many waitings, watchings, solicitings and disputes in council, this day
my country was confirmed to me under the great seal of England."
He had great hopes for its future, for he subsequently wrote: "God
will bless and make it the seed of a nation; I shall have a tender care
of the government that it will be well laid at first." Some of the
Swedes from Christina had come up the river in 1643 and settled at
the mouth of Chester Creek, at a place called Upland. The site was
an eligible one, and the first parties of Quakers, coming out in three
ships, settled there, living in caves which they dug in the river bank,
these caves remaining for many years after they had built houses.
Penn drew up a liberal scheme of government and laws for his
colony, in which he is said to have had the aid of Henry, the brother
of Algernon Sidney, and of Sir William Jones. He was not satisfied
with Upland, however, as his permanent place of settlement, but
directed that another site be chosen higher up the Delaware, at some

point where "it is most navigable, high, dry and healthy; that is,
where most ships can best ride, of deepest draught of water, if
possible, to load or unload, at the bank or key-side, without boating
or lightening of it." This site being selected between the Delaware
and Schuylkill Rivers, and the city laid out, Penn, with about a
hundred companions, mostly Welsh Quakers, in September, 1682,
embarked for the Delaware on the ship "Welcome," arriving at Upland
after a six weeks' voyage, and then going up to his city site, which he
named Philadelphia, the "City of Brotherly Love."
The first explorers of the Delaware River found located upon the site
of Philadelphia the Indian settlement of Coquanock, or "the grove of
long pine trees," a sort of capital city for the Lenni Lenapes. Their
great chief was Tamanend, and the primeval forest, largely composed
of noble pine trees, then covered all the shores of the river. The ship
"Shield," from England, with Quaker colonists for Burlington, in West
Jersey, higher up the river, sailed past Coquanock in 1679, and a note
was made that "part of the tackling struck the trees, whereupon
some on board remarked that 'it was a fine spot for a town.'" When
Penn sent out his advance agent and Deputy Governor, Captain
William Markham, of the British army, in his scarlet uniform, to lay
out the plan of his projected city, he wrote him to "be tender of
offending the Indians," and gave instructions that the houses should
have open grounds around them, as he wished the new settlement to
be "a green country town," and at the same time to be healthy, and
free from the danger of extensive conflagrations. Penn bought the
land farther down the Delaware from the Swedes, who had originally
bought it from the Indians, and the site for his city he bought from
the Indians direct. They called him Mignon, and the Iroquois, who
subsequently made treaties with him, called him Onas, both words
signifying a quill pen, as they recognized the meaning of his name.
Out on the Delaware, in what is now the Kensington shipbuilding
district, is the "neutral land of Shackamaxon." This word means, in
the Indian dialect, the "place of eels." Here, for centuries before
Penn's arrival, the Indian tribes from all the region east of the
Alleghenies, between the Great Lakes, the Hudson River and the

Potomac, had been accustomed to kindle their council fires, smoke
the pipe of deliberation, exchange the wampum belts of explanation
and treaty, and make bargains. Some came by long trails hundreds of
miles overland through the woods, and some in their birch canoes by
water and portage. It was on this "neutral ground" by the riverside
that Penn, soon after his arrival, held his solemn council with the
Indians, sealing mutual faith and securing their lifelong friendship for
his infant colony. This treaty, embalmed in history and on canvas,
was probably made in November, 1682, under the "Treaty Elm" at
Shackamaxon, which was blown down in 1810, the place where it
stood by the river being now preserved as a park. Its location is
marked by a monument bearing the significant inscription: "Treaty
Ground of William Penn and the Indian Nation, 1682—Unbroken
Faith." Thus began Penn's City of Brotherly Love, based on a compact
which, in the words of Voltaire, was "never sworn to and never
broken." It is no wonder that Penn, after he had seen his city site,
and had made his treaty, was so abundantly pleased that he wrote:
"As to outward things, we are satisfied, the land good, the air clear
and sweet, the springs plentiful, and provision good and easy to
come at, an innumerable quantity of wild fowl and fish; in fine, here
is what an Abraham, Isaac and Jacob would be well contented with,
and service enough for God, for the fields here are white for harvest.
O, how sweet is the quiet of these parts, freed from the anxious and
troublesome solicitations, harries and perplexities of woeful Europe."
The Lenni Lenapes, it is stated, told Penn and his people that they
often spoke of themselves as the Wapanachki, or the "men of the
morning," in allusion to their supposed origin in the lands to the
eastward, towards the rising sun. Their tradition was that at the time
America was discovered, their nation lived on the island of New York.
They called it Manahatouh, "the place where timber is got for bows
and arrows." At the lower end of the island was a grove of hickory
trees of peculiar strength and toughness. This timber was highly
esteemed for constructing bows, arrows, war-clubs, etc. When they
migrated westward they divided into two bands. One, going to the
upper Delaware, among the mountains, was termed Minsi, or "the

great stone;" and the other band, seeking the bay and lower river,
was called Wenawmien, or "down the river." These Indians originated
the name of the Allegheny Mountains, which they called the
Allickewany, the word meaning "He leaves us and may never
return"—it is supposed in reference to departing hunters or warriors
who went into the mountain passes.
THE QUAKER CITY.
The great city thus founded by William Penn is built chiefly upon a
broad plain between the Delaware and Schuylkill Rivers, about one
hundred miles from the sea, and upon the undulating surface to the
north and west. The shape of the city is much like an hour-glass,
between the rivers, although it spreads far west of the Schuylkill. The
Delaware River, in front of the built-up portion, sweeps around a
grand curve from northeast to south, and then, reversing the
movement, flows around the Horseshoe Bend below the city, from
south to west, to meet the Schuylkill. The railway and commercial
facilities, the proximity to the coal-fields, and the ample room to
spread in all directions, added to the cheapness of living, have made
Philadelphia the greatest manufacturing city in the world, and
attracted to it 1,300,000 inhabitants. The alluvial character of the
shores of the two rivers surrounds the city with a region of the
richest market gardens, and the adjoining counties are a wealthy
agricultural and dairy section. Clay, underlying a large part of the
surface, has furnished the bricks to build much of the town. Most of
the people own their homes; there are over two hundred and fifty
thousand dwellings and a thousand miles of paved streets, and new
houses are put up by the thousands every year as additional territory
is absorbed. When Penn laid out his town-plan, he made two broad
highways pointing towards the cardinal points of the compass and
crossing at right angles in the centre, where he located a public
square of ten acres. The east and west street, one hundred feet
wide, he placed at the narrowest part of the hour-glass, where the
rivers approached within two miles of each other. This he called the
High Street, but the public persisted in calling it Market Street. The

north and south street, laid out in the centre of the plat, at its
southern end reached the Delaware near the confluence with the
Schuylkill. This street is one hundred and thirteen feet wide, Broad
Street, a magnificent thoroughfare stretching for miles and bordered
with handsome buildings. Upon the Centre Square was built a Quaker
meeting-house, the Friends, while yet occupying the caves on the
bluff banks of the Delaware that were their earliest dwellings,
showing anxiety to maintain their forms of religious worship. This
meeting-house has since multiplied into scores in the city and
adjacent districts; for the sect, while not increasing in numbers, holds
its own in wealth and importance, and has great influence in modern
Philadelphia. Afterwards the Centre Square was used for the city
water-works, and finally it was made the site of the City Hall.
The bronze statue of the founder, surmounting the City Hall tower at
five hundred and fifty feet elevation, clad in broad-brimmed hat and
Quaker garb, carrying the city charter, and gazing intently
northeastward towards the "neutral land of Shackamaxon," is the
prominent landmark for many miles around Philadelphia. A blaze of
electric light illuminates it at night. This City Hall, the largest edifice
in America, and almost as large as St. Peter's Church in Rome, has
fourteen acres of floor space and seven hundred and fifty rooms, and
cost $27,000,000. It is a quadrangle, built around a central court
about two hundred feet square, and measures four hundred and
eighty-six by four hundred and seventy feet. The lower portion is of
granite, and the upper white marble surmounted by Louvre domes
and Mansard roofs. This great building is the official centre of
Philadelphia, but the centre of population is now far to the northwest,
the city having spread in that direction. The City Hall, excepting its
tower, is also being dwarfed by the many enormous and tall store and
office buildings which have recently been constructed on Broad and
other streets near it. Closely adjacent are the two vast stations of the
railways leading into Philadelphia, the Broad Street Station of the
Pennsylvania system, and the Reading Terminal Station, which serves
the Reading, Baltimore and Ohio and Lehigh Valley systems. Also
adjoining, to the northward, is the Masonic Temple, the finest

Masonic edifice in existence, a pure Norman structure of granite two
hundred and fifty by one hundred and fifty feet, with a tower two
hundred and thirty feet high, and a magnificent carved and decorated
granite Norman porch, which is much admired.
The great founder not only started his City of Brotherly Love upon
principles of the strictest rectitude, but he was thoroughly rectangular
in his ideas. He laid out all the streets on his plan parallel to the two
prominent ones, so that they crossed at right angles, and his map
was like a chess-board. In the newer sections this plan has been
generally followed, although a few country roads in the outer
districts, laid upon diagonal lines, have been converted into streets in
the city's growth. Penn's original city also included four other squares
near the outer corners of his plan, each of about seven acres, and
three of them were long used as cemeteries. These are now
attractive breathing-places for the crowded city, being named after
Washington and Franklin, Logan and Rittenhouse. The east and west
streets Penn named after trees and plants, while the north and south
streets were numbered. The chief street of the city is Chestnut
Street, a narrow highway of fifty feet width, parallel to and south of
Market Street. Its western end, like Walnut Street, the next one
south, is a fashionable residential section, both being prolonged far
west of the Schuylkill River. In the neighborhood of Broad Street, and
for several blocks eastward, Chestnut Street has the chief stores. Its
eastern blocks are filled largely with financial institutions and great
business edifices, some of them elaborate structures.
INDEPENDENCE HALL.
Upon the south side of Chestnut Street, occupying the block between
Fifth and Sixth Streets, is Independence Square, an open space of
about four acres, tastefully laid out in flowers and lawns, with
spacious and well-shaded walks. Upon the northern side of the
square, and fronting Chestnut Street, is the most hallowed building of
American patriotic memories, Independence Hall, a modest brick
structure, yet the most interesting object Philadelphia contains. It

was in this Hall, known familiarly as the "State-house," that the
Continental Congress governing the thirteen revolted colonies met
during the American Revolution, excepting when driven out upon the
British capture, after the battle of the Brandywine. The Declaration of
Independence was adopted here July 4, 1776. The old brick building,
two stories high, plainly built, and lighted by large windows, was
begun in 1732, taking three years to construct, having cost what was
a large sum in those days, £5600, the population then being about
ten thousand. It was the Government House of Penn's Province of
Pennsylvania. There has recently been a complete restoration, by
which it has been put back into the actual condition at the time
Independence was declared. In the central corridor stands the
"Independence bell," the most sacred relic in the city. This Liberty
bell, originally cast in England, hung in the steeple, and rang out in
joyous peals the news of the signing of the Declaration. Running
around its top is the significant inscription: "Proclaim Liberty
throughout the Land unto all the Inhabitants Thereof." This bell was
cracked while being rung on one of the anniversaries about sixty
years ago. In the upper story of the Hall, Washington delivered his
"Farewell Address" in closing his term of office as President. The
eastern room of the lower story is where the Revolutionary Congress
met, and it is preserved as then, the old tables, chairs and other
furniture having been gathered together, and portraits of the Signers
of the Declaration hang on the walls. The old floor, being worn out,
was replaced with tiles, but otherwise the room, which is about forty
feet square, is as nearly as possible in its original condition. Here are
kept the famous "Rattlesnake flags," with the motto "Don't Tread on
Me," that were the earliest flags of America, preceding the Stars and
Stripes. Of the deliberations of the Congress which met in this
building, William Pitt wrote: "I must declare that in all my reading
and observation, for solidity of reasoning, force of sagacity, and
wisdom of conclusion, under such a complication of difficult
circumstances, no body of men could stand before the National
Congress of Philadelphia." In this building is Penn's Charter of
Philadelphia, granted in 1701, and West's noted painting of "Penn's
Treaty with the Indians." There are also portraits of all the British

kings and queens from Penn's time, including a full-length portrait of
King George III., representing him, when a young man, in his
coronation robes, and painted by Allan Ramsay.
Other historic places are nearby. To the westward is Congress Hall,
where the Congress of the United States held its sessions prior to
removal to Washington City. To the eastward is the old City Hall,
where the United States Supreme Court sat in the eighteenth century.
Adjoining is the Hall of the American Philosophical Society, founded
by Benjamin Franklin, and an outgrowth of his Junto Club of 1743. It
has a fine library and many interesting relics. Franklin, who was the
leading Philadelphian of the Revolutionary period, came to the city
from Boston when eighteen years old, and died in 1790. His grave is
not far away, in the old Quaker burying-ground on North Fifth Street.
A fine bronze statue of Franklin adorns the plaza in front of the Post-
office building on Chestnut Street. Farther down Chestnut Street is
the Hall of the Carpenters' Company, standing back from the street,
where the first Colonial Congress met in 1774, paving the way for the
Revolution. An inscription appropriately reads that "Within these
walls, Henry, Hancock and Adams inspired the delegates of the
colonies with nerve and sinew for the toils of war." On Arch Street,
east from Franklin's grave, is the house where Betsy Ross made the
first American flag, with thirteen stars and thirteen stripes, from a
design prepared by a Committee of Congress and General
Washington in 1777. In this committee were Robert Morris and
Colonel George Ross, the latter being the young woman's uncle. It
appears that she was expert at needlework and an adept in making
the handsome ruffled bosoms and cuffs worn in the shirts of those
days, and had made these for General Washington himself. She had
also made flags, and there is a record of an order on the Treasury in
May, 1777, "to pay Betsy Ross fourteen pounds, twelve shillings
twopence for flags for the fleet in the Delaware River." She made the
sample-flag, her uncle providing the means to procure the materials,
and her design was adopted by the Congress on June 14, 1777, the
anniversary being annually commemorated as "Flag Day." Originally
there was a six-pointed star suggested by the committee, but she

proposed the five-pointed star as more artistic, and they accepted it.
The form of flag then adopted continues to be the American
standard. She afterwards married John Claypole, whom she survived
many years, and she died in January, 1836, aged 84, being buried in
Mount Moriah Cemetery, on the southwestern border of the city.
GIRARD COLLEGE.
The name of Girard is familiar in Philadelphia, being repeated in
streets, buildings, and financial and charitable institutions. On Third
Street, south from Chestnut Street, is the fine marble building of the
Girard Bank, which was copied after the Dublin Exchange. This,
originally built for the first Bank of the United States, was Stephen
Girard's bank until his death. One of the greatest streets in the
northern part of the city is Girard Avenue, over one hundred feet
wide, stretching almost from the Delaware River westward far beyond
the Schuylkill River, which it crosses upon a splendid iron bridge. In
its course through the northwestern section, this fine street diverges
around the enclosure of Girard College, occupying grounds covering
about forty-two acres. Stephen Girard, before the advent of Astor in
New York, amassed the greatest American fortune. He was born in
Bordeaux in 1750, and, being a sailor's son, began life as a cabin boy.
He first appeared in Philadelphia during the Revolution as a small
trader, and after some years was reported, in 1790, to have an estate
valued at $30,000. Subsequently, through trading with the West
Indies, and availing of the advantages a neutral had in the warlike
period that followed, he rapidly amassed wealth, so that by 1812,
when he opened his bank, he had a capital of $1,200,000; and so
great was the public confidence in his integrity that depositors
flocked to his institution. He increased its capital to $4,000,000; and
when the war with England began in that year he was able to take,
without help, a United States loan of $5,000,000. He was a
remarkable man, frugal and even parsimonious, but profuse in his
public charities, though strict in exacting every penny due himself. He
contributed liberally to the adornment of the city and created many
fine buildings. He despised the few relatives he had, and when he

died in 1831 his estate, then the largest known in the country, and
estimated at $9,000,000, was almost entirely bequeathed for charity.
Stephen Girard left donations to schools, hospitals, Masonic poor
funds, for fuel for the poor, and other charitable purposes; but the
major part of his fortune went in trust to the city of Philadelphia,
partly to improve its streets and the Delaware River front, but the
greater portion to endow Girard College. This was in the form of a
bequest of $2,000,000 in money and a large amount of lands and
buildings, together with the ground whereon the College has been
built. He gave the most minute directions about its construction, the
institution to be for the support and instruction of poor white male
orphans, who are admitted between the ages of six and ten years,
and between the ages of fourteen and eighteen years are to be
bound out as apprentices to various occupations. A clause in the will
provides that no ecclesiastic, missionary or minister of any sect
whatever is to hold any connection with the College, or even be
admitted to the premises as a visitor; but the officers are required to
instruct the pupils in the purest principles of morality, leaving them to
adopt their own religious beliefs. The College building is of white
marble, and the finest specimen of pure Grecian architecture in the
United States. It is a Corinthian temple, surrounded by a portico of
thirty-four columns, each fifty-five feet high and six feet in diameter.
The building is one hundred and sixty-nine by one hundred and
eleven feet, and ninety-seven feet high, the roof being of heavy slabs
of marble, from which, as the College stands on high ground, there is
a grand view over the city. Within the vestibule are a statue of Girard
and his sarcophagus. The architect, Thomas U. Walter, achieved such
fame from this building that he was afterwards employed to extend
the Capitol at Washington. There are many other buildings in the
College enclosure, some being little less pretentious than the College
itself. This comprehensive charity has been in successful operation
over a half-century, and it supports and educates some sixteen
hundred boys, the endowment, by careful management, now
exceeding $16,000,000.

Philadelphia is great in other charities, and notably in hospitals.
Opposite Girard College are the magnificent buildings of the German
Hospital and the Mary J. Drexel Home for the education of nurses,
established by the munificence of John D. Lankenau, the widowed
husband of the lady whose name it bears. The Drexel Institute,
founded by Anthony J. Drexel, is a fine building in West Philadelphia,
with $2,000,000 endowment, established for "the extension and
improvement of industrial education as a means of opening better
and wider avenues of employment to young men and women," and it
provides for about two thousand students. The Presbyterian,
Episcopal, Jewish, Methodist and Roman Catholic hospitals, all under
religious care, are noted. Philadelphia is also the great medical school
of the country, and the University, Jefferson, Hahnemann and
Women's Colleges, each with a hospital attached, have world-wide
fame. The oldest hospital, the Pennsylvania Hospital, occupying an
entire block between Spruce and Pine and Eighth and Ninth Streets,
was founded in 1752, and is supported almost entirely by voluntary
contributions. In 1841 it established in West Philadelphia a separate
Department for the Insane. The Medico-Chirurgical Hospital is a
modern foundation which has grown to large proportions. There are
many libraries—not only free libraries, with branches in various parts
of the city, for popular use, supported by the public funds, but also
the Philadelphia Library, founded by Franklin and his friends of the
Junto Literary Club in 1731, and its Ridgway Branch, established, with
$1,500,000 endowment, by Dr. James Rush—a spacious granite
building on Broad Street, which cost $350,000. One of the restrictions
of his gift, however, excludes newspapers, he describing them as
"vehicles of disjointed thinking." The Pennsylvania Historical Society
also has a fine library pertaining to early Colonial history, and many
valuable relics and manuscripts, including the first Bible printed in
America, and the original manuscripts of Home, Sweet Home, and
the Star-Spangled Banner.
NOTABLE PHILADELPHIA BUILDINGS.

There are many notable structures in Philadelphia. The United States
Mint, opposite the City Hall, and fronting on Chestnut Street, has
executed nearly all the coinage of the country since its establishment
in 1792, the present building having been completed in 1833. It
contains a most interesting collection of coins, including the "widow's
mite." A fine new mint is now being erected on a much larger scale in
the northwestern section of the city. The Bourse, on Fifth Street near
Chestnut, erected in 1895 at a cost of $1,500,000, is the business
centre, its lower hall being the most spacious apartment in the city,
and the edifice is constructed in the style of Francis I. The white
marble Custom House, with fine Doric portico, was originally erected
in 1819, at a cost of $500,000, for the second United States Bank,
this noted bank, which ultimately suspended, having been for many
years a political bone of contention. On the opposite side of the
street, covering a block, is a row of a half-dozen wealthy financial
institutions, making one of the finest series in existence, granite and
marble being varied in several orders of architecture. The Post-office
building, also on Chestnut Street, a grand granite structure in
Renaissance, with a façade extending four hundred feet, cost over
$5,000,000. The plain and solid Franklin Institute, designed to
promote the mechanical and useful arts, is not far away.
Down nearer the river is the venerable Christ Church, with its tall
spire, built in 1727, the most revered Episcopal church in the city, and
the one at which General Washington and all the Government officials
in the Revolutionary days worshipped. William White, a native of the
city, was the rector of this church and chaplain of the Continental
Congress, and in 1786 was elected the Episcopal Bishop of
Pennsylvania, being ordained by the Archbishop of Canterbury at
Lambeth in February, 1787. He presided over the Convention, held in
this church in 1789, which organized the Protestant Episcopal Church
in the United States. Christ Church still possesses the earliest chime
of bells sent from England to America, and the spire, rising nearly
two hundred feet, is a prominent object seen from the river. Bishop
White died in 1836, aged 88. He was also, in his early life, the rector
of St. Peter's Church, another revered Episcopal church at Third and

Pine Streets. In its yard is the grave of Commodore Stephen Decatur,
the famous American naval officer, who, after all his achievements
and victories, was killed in a duel with Commodore Barron in 1820,
his antagonist also dying. The most ancient church in Philadelphia is
Gloria Dei, the "Old Swedes'" Church, a quaint little structure near the
Delaware River bank in the southern part of the city, built in 1700.
The early Swedish settlers, coming up from Fort Christina, erected a
log chapel on this site in 1677, at which Jacob Fabritius delivered the
first sermon. After he died, the King of Sweden in 1697 sent over
Rev. Andrew Rudman, under whose guidance the present structure
was built to replace the log chapel; and it was dedicated, the first
Sunday after Trinity, 1700, by Rev. Eric Biorck, who had come over
with Rudman. Many are the tales told of the escapades of the early
Swedes in the days of the log chapel. The Indians on one occasion
undermined it to get at the congregation, as they were afraid of the
muskets which the men shot out of the loopholes. The women,
however, scenting danger, brought into church a large supply of soft-
soap, which they heated piping hot in a cauldron. When the redskins
made their foray and popped their heads up through the floor, they
were treated to a copious bath of hot soap, and fled in dismay. This is
the "Old Swedes'" Church at Wicaco of which Longfellow sings in
Evangeline. The poet, in unfolding his story, brings both Evangeline
and Gabriel from Acadia to Philadelphia in the enforced exodus of
1755, and thus graphically describes the Quaker City:

"In that delightful land which is washed by the Delaware's waters,
Guarding in sylvan shades the name of Penn, the Apostle,
Stands on the banks of its beautiful stream the city he founded.
There all the air is balm, and the peach is the emblem of beauty,
And the streets still re-echo the names of the trees of the forest,
As if they fain would appease the Dryads whose haunts they
molested.
There, from the troubled sea, had Evangeline landed an exile,
Finding among the children of Penn a home and a country.
Something, at least, there was, in the friendly streets of the city,
Something that spake to her heart and made her no longer a
stranger;
And her ear was pleased with the Thee and Thou of the Quakers,
For it recalled the past, the old Acadian country,
Where all men were equal, and all were brothers and sisters."
In Philadelphia it is said Evangeline lived many years as a Sister of
Mercy, and it was thus that she visited the ancient almshouse to
minister to the sick and dying on a Sabbath morning:
"As she mounted the stairs to the corridors, cooled by the east wind,
Distant and soft on her ear fell the chimes from the belfry of Christ
Church,
While intermingled with these, across the meadows were wafted
Sounds of psalms that were sung by the Swedes in their church at
Wicaco."
There she found the dying Gabriel, and both, according to the
tradition, are buried in the yard of the Roman Catholic Church of the
Holy Trinity, at Sixth and Spruce Streets:

"Still stands the forest primeval; but far away from its shadow,
Side by side, in their nameless graves, the lovers are sleeping.
Under the humble walls of the little Catholic churchyard,
In the heart of the city, they lie unknown and unnoticed.
Daily the tides of life go ebbing and flowing beside them,
Thousands of throbbing hearts, where theirs are at rest and forever;
Thousands of aching brains, where theirs no longer are busy;
Thousands of toiling hands, where theirs have ceased from their
labors;
Thousands of weary feet, where theirs have completed their journey."
In the ancient graveyard of "Old Swedes" is buried Alexander Wilson,
the American ornithologist, who was a native of Scotland, but lived
most of his life in Philadelphia, dying in 1813. The largest church in
the city is the Roman Catholic Cathedral of St. Peter and St. Paul,
fronting on Logan Square, an imposing Roman Corinthian structure of
red sandstone, two hundred and sixteen by one hundred and thirty-
six feet, and crowned by a dome rising two hundred and ten feet.
The chief institution of learning is the University of Pennsylvania, the
most extensive and comprehensive College in the Middle States,
dating from 1740, and munificently endowed, which occupies, with its
many buildings, a large surface in West Philadelphia, and has three
thousand students. This great institution originated from a building
planned in 1740 for a place in which George Whitefield could preach,
which was also used for a charity school. This building was conveyed
to trustees in 1749 to maintain the school, and they were in turn
chartered as a college in 1753 "to maintain an academy, as well for
the instruction of poor children on charity as others whose
circumstances have enabled them to pay for their learning." This
charitable feature is still maintained in the University by free
scholarships.
Philadelphia is eminently a manufacturing city, and its two greatest
establishments are the Cramp Shipbuilding yards in the Kensington

district and the Baldwin Locomotive Works on North Broad Street,
each the largest establishment of its kind in America. The city has
spread over a greater territory than any other in the United States,
and sixteen bridges span the Schuylkill, with others in contemplation,
its expansion beyond that river has been so extensive. The enormous
growth of the town has mainly come from the adoption of the
general principle that every family should live in its own house,
supplemented by liberal extensions of electrical street railways in all
directions. Hence, Philadelphia is popularly known as the "City of
Homes." As the city expanded over the level land, four-, six-, eight-
and ten-room dwellings have been built by the mile, and set up in
row after row. Two-story and three-story houses of red brick, with
marble steps and facings, make up the greater part of the town, and
each house is generally its owner's castle, the owner in most cases
being a successful toiler, who has saved his house gradually out of his
hard earnings, almost literally brick by brick. There is almost
unlimited space in the suburbs yet capable of similar absorption, and
the process which has given Philadelphia this extensive surface goes
on indefinitely. The population is also regarded as more
representative of the Anglo-Saxon races than in most American cities,
though the Teuton numerously abounds and speedily assimilates. The
greatest extent of Philadelphia is upon a line from southwest to
northeast, which will stretch nearly twenty miles in a continuous
succession of paved and lighted streets and buildings.
FAIRMOUNT PARK AND SUBURBS.
Philadelphia, excepting to the southward, is surrounded by a broad
belt of attractive suburban residences, the semi-rural region for miles
being filled with ornamental villas and the tree-embowered and
comfortable homes of the well-to-do and middle classes. Down the
Schuylkill is "Bartram's Garden," now a public park, where John
Bartram established the first botanic garden in America, and where
his descendants in 1899 celebrated the two hundredth anniversary of
his birth on June 2, 1699. His grandfather was one of the companions
of William Penn, and John Bartram, who was a farmer, mastered the

rudiments of the learned languages, became passionately devoted to
botany, and was pronounced by Linnæus the greatest natural
botanist in the world. Bartram bought his little place of about seven
acres in 1728, and built himself a stone house, which still exists,
bearing the inscription, cut deep in a stone, "John and Ann Bartram,
1731." He wrote to a friend describing how he became a botanist:
"One day I was very busy in holding my plough (for thou seest I am
but a ploughman), and being weary, I ran under a tree to repose
myself. I cast my eyes on a daisy; I plucked it mechanically and
viewed it with more curiosity than common country farmers are wont
to do, and I observed therein many distinct parts. 'What a shame,'
said my mind, or something that inspired my mind, 'that thou
shouldst have employed so many years in tilling the earth and
destroying so many flowers and plants without being acquainted with
their structure and their uses.'" He put up his horses at once, and
went to the city and bought a botany and Latin grammar, which
began his wonderful career. He devoted his life to botany, travelled
over America collecting specimens, and died in 1777. At the mouth of
the Schuylkill River is League Island, where the United States has an
extensive navy yard, and a reserve fresh water basin for the storage
of naval vessels when out of commission. The attractive Philadelphia
suburban features spread westward across the Schuylkill, and are
largely developed in the northwestern sections of Germantown and
Chestnut Hill, Jenkintown and the Chelten-hills. In this extensive
section the wealth of the people has of late years been lavishly
expended in making attractive homes, and the suburban belt for
miles around the city displays most charming scenery, adorned by
elaborate villas, pleasant lanes, shady lawns and well-kept grounds.
The chief rural attraction of Philadelphia is Fairmount Park, one of the
world's largest pleasure-grounds. It includes the lands bordering both
sides of the Schuylkill above the city, having been primarily
established to protect the water-supply. There are nearly three
thousand acres in the Park, and its sloping hillsides and charming
water views give it unrivalled advantages in delicious natural scenery.
At the southern end is the oldest water reservoir of six acres, on top

of a curious and isolated conical hill about ninety feet high, which is
the "Fair Mount," giving the Park its name. The Schuylkill is dammed
here to retain the water, and the Park borders the river for several
miles above, and its tributary, the Wissahickon, for six miles farther.
The Park road entering alongside the Fairmount hill passes a colossal
equestrian statue of George Washington, and beyond a fine bronze
statue of Abraham Lincoln, and also an equestrian statue of General
Grant. The roadways are laid on both sides of the river at the water's
edge, and also over the higher grounds at the summits of the sloping
bordering hills, thus affording an almost endless change of routes
and views. The frequent bridges thrown across the river, several of
them carrying railroads, add to the charm. An electric railway is
constructed through the more remote portions, and displays their
rustic beauty to great advantage. All around this spacious Park the
growing city has extended, and prosperous manufacturing suburbs
spread up from the river, the chief being the carpet district of the
Falls and the cotton-mills of Manayunk, the latter on the location of
an old-time Indian village, whose name translated means "the place
of rum." In this Park, west of the Schuylkill, was held the Centennial
Exposition of 1876, and several of the buildings remain, notably the
Memorial Art Gallery, now a museum, and the Horticultural Hall,
where the city maintains a fine floral display. William Penn's "Letitia
House," his original residence, removed from the older part of the
city, now stands near the entrance to the West Park.
A large part of the northeastern bank of the Schuylkill adjoining the
Park is the Laurel Hill Cemetery. Its winding walks and terraced
slopes and ravines give constantly varying landscapes, making it one
of the most beautiful burial-places in existence. In front, the river far
beneath curves around like a bow. Some of its mausoleums are of
enormous cost and elaborate ornamentation, but generally the
grandeur of the location eclipses the work of the decorator. Standing
on a jutting eminence is the Disston Mausoleum, which entombs an
English sawmaker who came to Philadelphia without friends and
almost penniless, and died at the head of the greatest sawmaking
establishment on the Continent. At one place, as the river bends, the

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Language In Use Georgetown University Round Table On Languages And Linguistics Andrea Tyler

More Related Content

Similar to Language In Use Georgetown University Round Table On Languages And Linguistics Andrea Tyler (20)

Recently uploaded (20)

Language In Use Georgetown University Round Table On Languages And Linguistics Andrea Tyler