Lexical Facility Size Recognition Speed and Consistency As Dimensions of Second Language Vocabulary Knowledge Michael Harrington Full Chapter

Lexical Facility : Size, Recognition
Speed and Consistency as Dimensions

of Second Language Vocabulary
Knowledge Michael Harrington
Visit to download the full and correct content document:
https://ebookmass.com/product/lexical-facility-size-recognition-speed-and-consistenc
y-as-dimensions-of-second-language-vocabulary-knowledge-michael-harrington/
MICHAEL HARRINGTON
the
be
to
of
and
a
in
t
tha
v e
it
ha
for
not with as youdo at

on
Lexical Facility
Size, Recognition Speed and Consistency
as Dimensions of Second Language
Vocabulary Knowledge
Lexical Facility
Michael Harrington
Lexical Facility
Size, Recognition Speed and
Consistency as Dimensions of Second
Language Vocabulary Knowledge
Michael Harrington
University of Queensland
Brisbane, QLD, Australia
ISBN 978-1-137-37261-1 ISBN 978-1-137-37262-8 (eBook)

DOI 10.1057/978-1-137-37262-8
Library of Congress Control Number: 2017946891
© The Editor(s) (if applicable) and The Author(s) 2018

The author(s) has/have asserted their right(s) to be identified as the author(s) of this work in accordance
with the Copyright, Designs and Patents Act 1988.
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and trans-
mission or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Cover design by Henry Petrides
Printed on acid-free paper
This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Macmillan Publishers Ltd.
The registered company address is: The Campus, 4 Crinan Street, London, N1 9XW, United Kingdom
This book is dedicated in loving memory to my parents, Frank and Dolores.
Front Cover
The image on the front cover is a stylized representation of what is known

as ‘Zipf ’s law’, which states that the frequency with which a word is used
is inversely proportional to its rank in a frequency table. The vertical
y-axis represents the frequency with which a word is used, and its rank
order is set out along the horizontal x-axis. The sloping function shows
that a small number of words account for the majority of uses. The
approach set out in this book assumes that frequency rank is a strong
predictor of vocabulary learning.
vii
Acknowledgments
I would first like to thank my wife Jan and daughter Bridget for their
forebearance. I am also greatly indebted to John Read for his advice and sup-
port throughout this project. He, of course, is not responsible for the final
outcome. Special thanks to collaborators Thomas Roche, Michael Carey, and
Akira Mochida, and colleagues Noriko Iwashita, Paul Moore, Wendy Jiang,
Mike Levy, Yukie Horiba, Yuutaka Yamauchi, Shuuhei Kadota, Ken Hashimoto,
Fred Anderson, Mark Sawyer, Kazuo Misono, John Ingram and Jenifer Larson-
Hall. Thanks also to Said Al-Amrani, Lara Weinglass, and Mike Powers.
Vikram Goyal programmed and has served as the long-standing sys-
tem administrator for the LanguageMAP online testing program used
to collect the data reported. He has been especially valuable to the proj-
ect. Special thanks also to Chris Evason, Director of the University of
Queensland’s (UQ) Foundation-Year program, who has provided encour-
agement and financial support for testing and program development.
Funding support is also acknowledged from Andrew Everett and the UQ
International Education Directorate.
The research reported here has been supported by the Telstra Broadband
Fund and a UQ Uniquest Pathfinder grant for the development of the
LanguageMAP program. Support was also provided by research contracts
from the Milton College (Chap. 9) and International Education Services–
UQ Foundation-Year (Chaps. 8, 9, and 10), and a grant, with Thomas
Roche, from the Omani Ministry of Research.
ix
Contents
Part 1 Introduction 1

References 2
1 Size as a Dimension of L2 Vocabulary Skill 3

1.1 Introduction 3
1.2 Estimating Vocabulary Size 5
1.3 Vocabulary Size as a Dimension of Learners’
Vocabulary Knowledge 13
1.4 Conclusions 21
References 22
2 Measuring Recognition Vocabulary Size 25

2.1 Introduction 25
2.2 Approaches to Measuring Recognition
Vocabulary Size 26
2.3 Uses of the Vocabulary Size Tests 34
2.4 Conclusions 39
References 40
xi
xii Contents
3 L2 Word Recognition Skill and Its Measurement 45

3.1 Introduction 45
3.2 Word Recognition Skill and Text Comprehension 46
3.3 Word Recognition Skill and L2 Text Comprehension 49
3.4 The LDT as a Measure of Word Recognition Skill 51
3.5 LDT Performance as a Window on Word Knowledge 57
3.6 Using the LDT Format to Measure L2 Word
Recognition Skill 60
3.7 Conclusions 61
References 61
4 Lexical Facility: Bringing Size and Speed Together 67

4.1 Introduction 67
4.2 Defining Lexical Facility 68
4.3 Lexical Facility as a Vocabulary Skill Construct 73
4.4 Lexical Facility as a Measurement Construct 76
4.5 Bringing Size and Speed Together 79
4.6 Recognition Vocabulary Size and Speed as a Vocabulary
Measure83
4.7 Establishing Lexical Facility: The Research Program 86
4.8 Conclusions 88
References 89
5 Measuring Lexical Facility: The Timed Yes/No Test 95

5.1 Introduction 95
5.2 The Timed Yes/No Test 96
5.3 Scoring the Timed Yes/No Test 99
5.4 Administering the Test 109
5.5 The Timed Yes/No Test as an L2 Vocabulary Task 112
5.6 Lexical Facility in English 115
5.7 Conclusions 116
References117
Contents
xiii
Part 2 Introduction 121

1.1 Overview 121
1.2 Aims of the Empirical Research 122
1.3 An Overview of Methods Used 122
References129
6 Lexical Facility as an Index of L2 Proficiency131

6.1 Introduction 131
6.2 Study 1: Lexical Facility as an Index
of English Proficiency 132
6.3 Study 1 Results 136
6.4 Sensitivity of the Lexical Facility Measures
to Frequency Levels 146
6.5 Discriminating Between Frequency Levels 148
6.6 Findings for Study 1 151
6.7 Conclusions 152
References153
7 Lexical Facility and Academic English Proficiency157

7.2 Study 2: Lexical Facility and University English
Entry Standards 158
7.4 Study 2 Findings 180
7.5 Conclusions 183
References185
8 Lexical Facility and IELTS Performance187

8.2 Study 3: Lexical Facility and IELTS Performance 188
xiv Contents
8.4 Findings for Study 3 IELTS Band-Scores 201

8.5 Conclusions 201
References203
9 Lexical Facility and Language Program Placement205

9.2 Study 4: Sydney Language School Placement Study 207
9.4 Findings for Study 4 Sydney Language Program
Placement216
9.5 Study 5: Singapore Language Program Study 217
9.7 Findings for Study 5 Singapore Language
Program Levels 223
9.8 Conclusions 224
References225
10 Lexical Facility and Academic Performance in English227

10.2 Study 6: Lexical Facility Measures and Academic
English Grades 228
10.4 Findings for Study 6 Lexical Facility and Academic
English Grades 235
10.5 Study 7: Lexical Facility and GPA 235
10.6 Findings for Study 7 Lexical Facility and GPA 236
10.7 Other GPA Studies 236
10.8 Conclusions 239
References240
Contents
xv
11 The Effect of Lexical Facility241

11.2 Sensitivity of Lexical Facility Measures by
Performance Domain 242
11.3 Key Findings 252
References258
12 The Future of Lexical Facility261

12.2 The Case for Lexical Facility 262
12.3 Measuring Lexical Facility: The Timed Yes/No Test
and Alternatives266
12.4 The Next Step in Lexical Facility Research 274
12.5 Uses of Lexical Facility in Vocabulary Assessment
and Instruction276
References279
References283
Index303
List of Figures
Fig. 1.1 Elements of vocabulary knowledge tapped by vocabulary

size tests 7
Fig. 1.2 A frequentist model of vocabulary learning 15
Fig. 1.3 Cumulative percentage of text coverage and corresponding
frequency bands 18
Fig. 1.4 Text coverage as the number of unfamiliar words and the
number of lines of text per unfamiliar word 18
Fig. 1.5 A sample reading text with 80% text coverage 19
Fig. 1.6 Text comprehension percentage as a function of vocabulary
coverage (Schmitt et al. 2011, p. 34) 20
Fig. 2.1 Instructions and example item for Vocabulary Levels Test
(Adapted from Nation 2013, p. 543) 27
Fig. 2.2 Sample item from Nation’s Vocabulary Size Test 28
Fig. 2.3 A simple checklist version of the original Yes/No Test 31
Fig. 2.4 Comparison of VLT and Yes/No Test Performance
(Mochida and Harrington 2006) 33
Fig. 3.1 Word recognition in the construction–integration model
of text comprehension (figure adapted from Perfetti and
Stafura (2014, p. 33)) 47
Fig. 3.2 A schematic diagram of the lexical decision task 54
Fig. 5.1 Yes/No Test response types 99
Fig. 5.2 Four Yes/No Test scoring formulas 101
Fig. 5.3 Composite measure formulas 108
xvii
xviii List of Figures
Fig. 5.4 Elements of the instruction set for the Timed Yes/No Test 111
Fig. 6.1 Lexical facility measures by English proficiency levels 140
Fig. 6.2 Median proportion of hits and 95% confidence intervals
for lexical facility measures by frequency levels and groups 149
Fig. 6.3 Median individual mnRT and 95% confidence intervals
for lexical facility measures by frequency levels and groups 150
Fig. 6.4 Median coefficient of variation (CV) and 95% confidence
intervals for lexical facility measures by frequency levels
and groups 150
Fig. 7.1 University entry standard study. Mean proportion of hits
by frequency levels for written and spoken test results 179
Fig. 7.2 University entry standard study. Mean response times by
frequency levels for written and spoken test results 180
Fig. 7.3 University entry standard study. Mean CV ratio by
frequency levels for written and spoken test results 181
Fig. 8.1 Combined IELTS dataset: Timed Yes/No Test scores by
IELTS overall band scores 194
Fig. 9.1 Sydney language program study. Comparison of VKsize
and mnRT scores with program placement grammar and
listening scores across four placement levels 213
Fig. 9.2 Singapore language program levels. Standardized scores
for the lexical facility measures (VKsize, mnRT, and CV)
for the VLT and BNC test versions 219
Fig. 9.3 Singapore language program study. Standardized scores
for the lexical facility measures (VKsize, mnRT, and CV)
for the combined test by level 221
Fig. 10.1 Oman university GPA study. Standardized VKsize,
mnRT, and CV scores by faculty 238
List of Tables
Table 1.1 Vocabulary size expressed in word families

and text coverage (written and spoken) across nine
corpora (Nation 2006, p. 79) 17
Table 3.1 A meta-analysis of factors affecting L2 reading skill
(Jeon and Yamashita 2014) 50
Table 6.1 Bivariate correlations and 95% confidence intervals
(within square brackets) for the three lexical facility
measures (VKsize, mnRT, and CV) and two composite
scores (VKsize_ mnRT and VKsize_ mnRT_ CV) 138
Table 6.2 Proficiency-level study. Means, standard deviations,
and confidence intervals for false-alarm rates and
the lexical facility measures, individual and
composite, for the three proficiency levels 139
Table 6.3 Proficiency-level study. One-way ANOVAs for
individual and composite lexical facility measures
as discriminators of English proficiency levels 143
Table 6.4 Proficiency-level study. Post hoc comparisons for
individual and composite measures, VKsize,
mnRT, CV VKsize_mnRT, and VKsize_mnRT_CV 143
Table 6.5 Proficiency-level study. Medians, interquartile ranges,
and 95% confidence intervals for the hits, mean response
time (in milliseconds), and mean proportion coefficient
of variation by frequency levels and groups 147
xix
xx List of Tables
Table 6.6 Frequency-level analysis. Comparing sensitivity of hits

(correct responses to words), mean RT, and CV to
frequency band differences using the omnibus Friedman
test and the follow-up Wilcoxon signed-rank test 149
Table 7.1 University entry standard study: written and spoken test
results. Pearson’s correlations for the three individual
measures (VKsize score, mnRT, and CV) and the two
composite scores (VKsize_mnRT and VKsize_mnRT_CV) 163
results. Means (M), standard deviations (SD), and
confidence intervals (95% CI) for the lexical facility
measures for the five English proficiency standard groups 165
results. Means (M), standard deviations (SD) and
confidence intervals (CI) for the composite scores
VKsize_mnRT and VKsize_mnRT_CV for
the five English entry standard groups 166
Table 7.4 Entry standard study. One-way ANOVA for individual
and composite lexical facility measures as discriminators
of English proficiency groups 172
Table 7.5 University entry standard group. Significant pairwise
comparisons for the VKsize measure for written and
spoken test results 173
Table 7.6 University entry standard study. Significant pairwise
comparisons for the mnRT and CV measures for written
and spoken test results 174
Table 7.7 University entry standard study. Significant pairwise
comparisons for composite VKsize_mnRT and VKsize_
mnRT_CV measures for written and spoken test results 175
Table 8.1 IELTS study data set. Years 1–3 means and standard
deviations, within brackets, for the VKsize, mnRT, and
CV measures by IELTS overall band score 191
Table 8.2 IELTS band-score study. Means, standard deviations, and
confidence intervals (CI) for the lexical facility measures,
individual and composite, for IELTS overall band scores 192
Table 8.3 IELTS study. IELTS band-score study. Bivariate
correlations with bootstrapped confidence intervals for
IELTS band scores and lexical facility measures 193
List of Tables
xxi
Table 8.4 IELTS band-score study. One-way ANOVAs for individual

and composite lexical facility measures as discriminators of
IELTS overall band scores 196
Table 8.5 IELTS study. Bandwise significant post hoc comparisons
for VKsize, mnRT, and CV 197
Table 8.6 IELTS band-score study. IELTS bandwise post hoc
comparisons for the VKsize_mnRT and VKsize_mnRT_
CV measures 198
Table 8.7 IELTS band-score study. Model summary (R2 and ΔR2)
for hierarchical regression analysis with proficiency level as
criterion and VKsize, mnRT, and CV as predictor variables
on written and spoken tests with complete and false-alarm-
trimmed (20 and 10%) data sets 200
Table 9.1 Sydney language program study. Bivariate Pearson’s
correlations for lexical facility measures, and listening and
grammar test scores 210
Table 9.2 Sydney language program study. Means, standard
deviations, and 95% confidence intervals for the lexical
facility measures at the four placement levels 211
Table 9.3 Sydney language program study. One-way ANOVAs for
individual and composite lexical facility measures and
placement test scores as discriminators of placement levels 214
Table 9.4 Sydney language program study. Significant post hoc
pairwise comparisons of the lexical facility measures and
listening test 215
Table 9.5 Singapore language program study. Means, standard
deviations, and confidence intervals for the lexical facility
measures for the four Singapore language program levels 220
Table 9.6 Singapore language program study. One-way ANOVAs
for individual and composite lexical facility measures as
discriminators of program levels 222
Table 9.7 Singapore language program study. Significant post hoc
comparisons for the lexical facility measures for the four
placement levels 223
Table 10.1 Australian university foundation-year study. Means,
standard deviations, and confidence intervals (CI) for
the individual and composite lexical facility measures, and
median and range values for academic grades and GPAs
for entry and exit groups 231
xxii List of Tables
Table 10.2 Bivariate correlations between lexical facility measures and

academic English performance measures for entry and exit
groups232
Table 10.3 Australian university foundation-year study. Model
summary of hierarchical regression analyses for entry and
exit groups using EAP grade percentage as criterion and
VKsize, mnRT, and CV as ordered predictor variables 234
Table 11.1 Summary of means (M) and standard deviations (SD) for
VKsize, hits, mnRT, and CV measures for Studies 1–5 242
Table 11.2 Summary of lexical facility measures’ effect sizes for
individual and composite measures 244
Introduction
Two bedrocks of fluent second language (L2) performance are an ade-

quate stock of words and the ability to access those words quickly.
Separately, the two have been shown to be reliable and sensitive correlates
of L2 proficiency both across and within user levels. The two are exam-
ined here jointly as a property of L2 vocabulary skill called lexical facility.
The book first makes the conceptual case for combining the two dimen-
sions and then provides empirical evidence for the sensitivity of the com-
bined measures to differences in proficiency and performance in common
domains of academic English. The main focus is on lexical facility in
written English, though some spoken language data are also presented.
Scope of the Book

The term lexical facility reflects how many words a learner knows and how
fast these words can be recognized. The term lexical is used to denote the
word-level focus, and the term facility the relative ease of accessing that
knowledge. A sizeable literature exists that relates vocabulary size to L2
performance. Researchers, including Bhatia Laufer, Paul Meara, Paul
Nation, and Norbert Schmitt, have sought to identify the kind and num-
ber of words an individual needs to function in various L2 domains, with
xxiii
xxiv Introduction
a particular interest in the vocabulary size needed for fluent performance

and its assessment in domains of academic English. A foundation of
vocabulary size research is the use of word frequency statistics as an index
for estimating an individual user’s vocabulary size. The resulting estimates
are then related to performance in various domains (e.g., Laufer and
Nation 1995). The vocabulary size research literature is the point of
departure for the lexical facility approach presented in the book.
A smaller body of research has also examined how L2 word processing
skill develops. Norman Segalowitz, Jan Hulstijn, and colleagues have
investigated the role that word recognition speed and consistency play in
fluent L2 performance, and in particular the development of automaticity.
“Word recognition speed is expressed throughout this book as the mean
recognition time (mnRT) it takes an individual to recognize a set of words
presented separately.” Faster recognition times have been shown to reliably
correlate with better performance both within and between users. In addi-
tion to the relative speed with which words are recognized, the overall
consistency of recognition speed is also of interest. Word recognition con-
sistency is captured in the coefficient of variation (CV), which is the ratio
of the standard deviation of the mnRT to the mnRT itself (SDmnRT/
mnRT). Segalowitz has proposed that the interaction of the mnRT and
the CV over the course of proficiency development can serve as an indica-
tor of automatization (Segalowitz and Segalowitz 1993). In the lexical
facility account, the CV is examined as an index of proficiency by itself
and in combination with the size and mnRT measures. As a measure of
response variability, the CV is examined as a window on vocabulary skill
development, as opposed to mere ‘noise’ that might otherwise obscure
experimental effects of interest. The interest in variability as a characteris-
tic of performance in its own right is attracting increasing attention in
cognitive science (Balota and Yap 2011; Hird and Kirsner 2010).
The two research areas differ in goals and method, but are in accord that
quantitative measures of vocabulary size and processing skill are important
indicators of L2 proficiency. Proficient learners have bigger vocabularies
and can access that knowledge more efficiently than their less proficient
counterparts. The book explores how the empirically established—and
intuitive—relationship between proficiency, and vocabulary size and pro-
cessing skill is manifested in various domains of academic English.
Introduction
xxv
The book is the first to investigate the value of treating vocabulary

size and processing skill (recognition speed and consistency) as a uni-
tary construct. The main empirical concern is the extent to which com-
bined measures of vocabulary size and processing skill are more sensitive
to performance differences than size alone. Sensitivity is reflected in
how reliably (as reflected in statistical significance) the measures dis-
criminate between levels in a given domain, and the magnitude of this
difference as reflected in the effect size. Evidence for the efficacy of a
composite measure combining static knowledge (size) and dynamic
processing skill (speed and consistency)—that is, for lexical facility—
has clear implications for L2 vocabulary research, testing, and assessment.
Lexical facility is a quantitative entity that captures a crucial facet of
lower-level L2 vocabulary knowledge skill. It is approached as a trait, that
is, as a user-internal, context-free property of L2 vocabulary knowledge
that is developed as a result of experience with the language and is avail-
able for use across contexts (Read and Chapelle 2001).
Research Goals
This book has three goals. The first is to make the theoretical case for lexi-
cal facility. The validity of the construct is established in the first four
chapters by first examining the crucial roles that vocabulary size (Chaps. 1
and 2) and word recognition skill (Chap. 3) play in L2 performance.
The rationale for characterizing size and processing skill jointly as an L2
vocabulary construct, that is, for lexical facility, is then set out in Chap. 4.
This chapter discusses key theoretical and methodological issues that arise
from the proposal. Primary among these is the attempt to treat size and
speed as parts of a unitary construct. Standard practice in the psychomet-
ric tradition has long been to treat the two as separate dimensions.
Human performance has been characterized either as knowledge (also
called power) or speed, the relative importance of each dependent on the
kind of performance being measured. Knowledge is seen as the critical
attribute of higher-level cognitive tasks such as educational testing, while
speed is paramount for mechanical tasks such as typing. The lexical
facility account proposes that size (knowledge) and processing skill (speed
xxvi Introduction
and consistency) can be productively considered together as indices of L2

vocabulary proficiency. As a result, the proposal has implications for the
broader incorporation of temporal measures in models of L2 learning
and use.
The second and third goals concern the empirical case for the con-
struct. The second goal is to assess the reliability and validity of an instru-
ment to measure lexical facility, the Timed Yes/No Test. In Part 2, seven
studies are presented that examine the sensitivity of the vocabulary size
and processing skill measures (size and consistency), individually and in
combination, to variability in proficiency and performance in various
academic English domains. All seven studies measure lexical facility using
the Timed Yes/No Test. The instrument is an online measure of recogni-
tion vocabulary knowledge based on the lexical decision task, a measure
of lexical access widely used in cognitive psychology. Chapter 5 describes
the Timed Yes/No Test and provides a rationale for its use. The use of
speed and consistency as measures of proficiency raises methodological
and technical issues. These are identified, and the implications for bring-
ing time as a performance measure out of the laboratory and into class-
room and testing contexts are discussed.
The third goal is to demonstrate the sensitivity of the lexical facility
measures to proficiency and performance differences in academic English.
Chapter 6 establishes the sensitivity of the size, speed, and consistency
measures to differences in proficiency levels in university-age users.
The chapter also demonstrates the validity of word frequency statistics to
index individual vocabulary knowledge. In Chap. 7, the sensitivity of the
measures to group differences in English entry standards used in an
Australian university is examined. Written and spoken versions of the
test are administered to evaluate differences in test performance due to
language mode. Chapter 8 investigates the measures as predictors of per-
formance by preuniversity students on one specific English entry stan-
dard, the International English Language Testing System (IELTS) test.
Performance on the lexical facility measures is compared with placement
Introduction
xxvii
testing outcomes in language schools in Sydney and Singapore in Chap. 9.

The last chapter, Chap. 10, investigates the measures as predictors of aca-
demic English grades and grade point average (GPA) in a university prep-
aration program in Australia. Also discussed are findings from other
studies that have addressed the same issues. Chapter 11 presents a sum-
mary of the findings from all the studies. The data reported in the various
studies are drawn from published and unpublished research by the author
and colleagues. Chapter 12 completes the book by considering the future
of the lexical facility proposal in light of the findings.
In summary, this book attempts to establish lexical facility as a quanti-
tative measure of L2 vocabulary proficiency that can serve as a context
independent index sensitive to learner performance in specific academic
English settings. The studies in Part 2 aim to
1. compare the three measures of lexical facility (vocabulary knowledge,

mean recognition time, and recognition time consistency) as stable indices
of L2 vocabulary skill;
2. evaluate the sensitivity of the three measures individually and as compos-
ites to differences in a range of academic English domains; and, in doing
so,
3. establish the degree to which the composite measures combining size with
processing skill (recognition speed and consistency) provide a more sensitive
indicator of L2 proficiency and performance differences than vocabulary
size alone.
The book is in two parts. Part 1 presents the theoretical foundation

and motivation for the lexical facility proposal. Part 2 reports on a set of
studies that provide empirical evidence for lexical facility and concludes
with a chapter that considers the place of lexical facility in the modeling
and measurement of L2 vocabulary.
xxviii Introduction
References
Balota, D. A., & Yap, M. J. (2011). Moving beyond the mean in studies of
mental chronometry: The power of response time distributional analyses.
Current Directions in Psychological Science, 20(3), 160–166.
Hird, K., & Kirsner, K. (2010). Objective measurement of fluency in natural
language production: A dynamic systems approach. Journal of Neurolinguistics,
23(5), 518–530. doi:10.1016/j.jneuroling.2010.03.001.
Laufer, B., & Nation, P. (1995). Vocabulary size and use: Lexical richness in L2
written production. Applied Linguistics, 16(3), 307–322.
Read, J., & Chapelle, C. A. (2001). A framework for second language vocabu-
lary assessment. Language Testing, 18(1), 1–32.
Segalowitz, N., & Segalowitz, S. J. (1993). Skilled performance, practice and
differentiation of speed-up from automatization effects: Evidence from sec-
ond language word recognition. Applied Psycholinguistics, 14(3), 369–385.
doi:10.1017/S0142716400010845.
Part 1
Introduction
Part 1 (Chaps. 1, 2, 3, 4, and 5) introduces the theoretical and method-

ological foundations of the lexical facility account. Chapter 1 introduces
the vocabulary size research program, including the frequency-based tests
of vocabulary knowledge that are used to estimate second language (L2)
vocabulary size in the individual user, which in turn has been related to
differences in L2 performance. Chapter 2 then presents different types of
vocabulary size tests, including the Vocabulary Levels Test (Nation 2013)
and the Yes/No Test (Meara and Buxton 1987). Test assumptions and
uses in testing and instruction are described and key findings surveyed.
Research on the development of speed and consistency in L2 word recog-
nition skill is examined in Chap. 3. The aims and methods of this research
paradigm are then described, as are key research findings. These two inde-
pendent lines of research provide the foundation for the lexical facility
proposal introduced in Chap. 4, which sets out the rationale for combin-
ing the two dimensions and discusses the key issues related to this under-
taking. Chapter 5 describes the Timed Yes/No Test, which is used in the
studies in Part 2 that provide evidence for the lexical facility account.
2 1 Introduction
References
Meara, P., & Buxton, B. (1987). An alternative to multiple choice vocabulary
tests. Language Testing, 4(2), 142–145.
Nation, I. S. P. (2013). Learning vocabulary in another language (2nd ed.).
Cambridge: Cambridge University Press
1
Size as a Dimension
of L2 Vocabulary Skill
Aims
• Introduce the vocabulary size research literature.

• Describe how vocabulary size is counted.
• Describe the use of word frequency statistics to estimate vocabulary size.
• Relate vocabulary size measures to second language (L2)
performance.
1.1 Introduction
This chapter introduces the field of what will be called vocabulary size
research, an approach based on the simple assumption that the overall
number of words a user knows—the breadth of an individual’s vocabulary
stock—provides an index of vocabulary knowledge. The focus on vocab-
ulary breadth means that little attention is given to what specific words
are known or the extent (or depth) to which any given word is used.
Rather, researchers in the area are interested in estimating the vocabulary
size needed to perform particular tasks in a target language. These tasks
© The Author(s) 2018 3

M. Harrington, Lexical Facility, DOI 10.1057/978-1-137-37262-8_1
4 1 Size as a Dimension of L2 Vocabulary Skill
can range from reading authentic texts (Hazenberg and Hulstijn 1996) to
coping with unscripted spoken language (Nation 2006). Size estimates
are used to propose vocabulary thresholds for second language (L2)
instruction, and more generally to provide a quantitative picture of an
individual’s L2 vocabulary knowledge (Laufer 2001; Laufer and
Ravenhorts-Kalovski 2010). The focus here, and in the book in general,
is on the size of recognition vocabulary and the role it plays in L2 use.
The main focus is on the recognition of written language.
Recognition vocabulary is acquired before productive vocabulary and
serves as the foundation for the learning of more complex language struc-
tures. The store of recognition vocabulary knowledge builds up over the
course of an individual’s experience with the language. This knowledge
ranges from the most minimal, as in the case of knowing only that a word
exists, to an in-depth understanding of its meaning and uses. A sparkplug
may be a thingamajig found in a car or, according to Wikipedia, ‘a device
for delivering electric current from an ignition system to the combustion
chamber of a spark-ignition engine to ignite the compressed fuel/air mix-
ture by an electric spark, while containing combustion pressure within
the engine’. Recognition vocabulary knowledge emerges from both
intentional learning and implicit experience, and even the most casual
experience can contribute to the stock of recognition vocabulary knowl-
edge. Repeated exposure to a word also has a direct effect on how effi-
ciently it is recognized.
The notion that knowing more words allows a language user to do
more in the language hardly seems controversial. However, many appar-
ently commonsensical assumptions in language learning are often diffi-
cult to specify in useful detail or to apply in practice (Lightbown and
Spada 2013). Even when evidence lends support to the basic idea, spe-
cific findings introduce qualifications that often diminish the scope and
power of the original insight. This chapter introduces and surveys the
vocabulary size research literature to see how the ‘greater size = better per-
formance’ assumption manifests itself. The methodology used for esti-
mating vocabulary size is first described, and then findings from key
studies are presented.
Size is a quantitative property and therefore requires some unit of mea-
surement. In the vocabulary size approach, it is the single word. Size
estimates reflect vocabulary breadth and have been related to L2 perfor-

mance in two ways. Researchers have sought to establish the minimum
size thresholds needed to perform specific tasks, such as reading an aca-
demic text (Schmitt et al. 2011), or related size to performance outcomes
in specific settings, as in placement testing (Meara and Jones 1988).
1.2 Estimating Vocabulary Size

The measurement of recognition vocabulary is a far more complex task
than might first appear. The first difficulty involves defining what to
count as a word. Criteria must also be established for deciding how a
given word is recognized for counting. Finally, a practical means must be
devised for obtaining a sufficient sample of the individual’s language from
which to make a valid size estimate. All three factors present challenges
for the researcher.
What to Count
The vocabulary size approach quantifies vocabulary knowledge as a col-

lection of single words. Characterizing vocabulary knowledge as a collec-
tion of individual words accords with how vocabulary knowledge is
popularly viewed. Single words are the means by which children learn to
spell and are the basis for dictionaries, spelling bees, and crossword puz-
zles. They also have a privileged place in vocabulary learning and teach-
ing, where word lists are a staple feature of any language textbook. And,
of course, multiword units (collocations, formulaic speech) are ultimately
made up of single words. Learning these forms involves either associating
a combination of known words to a new meaning or learning a new unit
in which some or all of the words are unknown (Wray 2008). In either
case, the single word represents a basic building block.
Single words are different from other kinds of language knowledge in
how they are acquired and represented in the brain. The L2 learner learns
a word (sound–meaning pair) consciously and that is stored as part of the
declarative memory system, a system open to reflection and explicit
odification. But this knowledge is only part of the lexicon, which con-
m
sists of these words in combination with the mostly implicit grammatical
properties that constrain how the words are used. These properties reside
in procedural memory, a system of implicit, unconscious knowledge.
Paradis (2009) makes a distinction between vocabulary and the lexicon to
capture this difference. Vocabulary is the totality of sound–meaning asso-
ciations and is typical of L2 learner knowledge, particularly in the early
stages. The lexicon characterizes the system of explicit and implicit
knowledge that the first language (L1) user develops as a matter of course
in development, and which is developed to varying degrees in more
advanced L2 users. In Paradis’s terms, the lexical facility account relates
strictly to vocabulary knowledge, its measurement, and its relationship to
L2 proficiency and performance.
Last, the pivotal role the single word plays in online processing also
reflects its importance. The word serves as the intersecting node for a
range of sentence and discourse processes that unfold in the process of
reading (Andrews 2008). It is where the rubber meets the road, as it were,
in text comprehension.
The focus on the recognition of single words means that the vocabu-
lary size approach captures only a small part of L2 vocabulary knowledge,
a multidimensional notion comprising knowledge of form, meaning, and
usage. Each word is part of a complex web of relationships with other
words, and this complex network is used to realize the wide range of
expressive, communicative, and instrumental functions encountered in
everyday use. Figure 1.1 depicts the basic elements of word knowledge in
a three-part model adapted from Nation (2013); see also Richards (1976).
The vocabulary size account reduces vocabulary knowledge to the sin-
gle dimension of the number of individual words a user knows, or more
precisely, recognizes. It is about the user’s ability to relate a form to a basic
meaning, whether by identifying the meaning from among a set of alter-
natives, as in the Vocabulary Levels Test (VLT), or merely recognizing a
word when it is presented alone, as in the Yes/No Test. This passive ‘rec-
ognition knowledge’ is assumed to be an internal property—a trait—of
the L2 user’s vocabulary stock that can be measured independently of a
given context.
Very high certainty

Written form, including orthography and
possible letter combinations
FORM
Spoken form, including the pronunciation of
individual sounds and connected speech
Word parts, including part of speech and

morphology
MEANING Referents, such as chair, sky, car
Concepts, such as truth, love, justice
Conceptual associations and links, including
metaphoric language such as life is a journey
Collocations: Tendency of two or more words

to occur together in discourse, both
grammatical
collocations (abide by, deal with) and semantic
collocations (spend money, cheerful
expression).
Associations: Comprised of links between
USE
words. Include syntagmatic associates
(abandon; hope, ship, me), paradigmatic
associates (abandon; neglect, give up, forsake).
Stylistic variations: Based on setting and
participants. This includes changes in use over
time, geographical or regional variation, social
class and social role variation; also includes Very low certainty
emotional valence.
Fig. 1.1 Elements of vocabulary knowledge tapped by vocabulary size tests
Single words are therefore of primary importance, primary used here

both in the sense of being crucial to understanding and in representing
the first stage of the comprehension process. However, it is also the case
that single words are typically used in combination with other words.
These combinations can be fixed, as in the case of collocations, or they
can be governed by grammatical and discourse constraints. The meaning
of a given word very often depends on the context, and ‘knowing’ a word
ultimately comes down to whether it facilitates comprehension in a par-
ticular context in an appropriate and timely manner. The measurement
of size alone says nothing about the depth of word knowledge, though
the two are not unrelated. Ultimately, greater vocabulary size correlates
with greater depth of vocabulary knowledge (Vermeer 2001).
The central question in the vocabulary size approach is the degree to
which this single form–meaning dimension relates to individual differ-
ences in L2 performance. Evidence of a reliable relationship between size
and performance has implications for the way L2 vocabulary knowledge
is conceptualized and, in turn, for L2 vocabulary assessment. The next
section will consider the challenging problem of how to count single
words.
Quantifying Vocabulary Size
There are alternative ways to calculate vocabulary size, all with their
advantages and disadvantages. The number of words on this page could
simply be counted by tallying the number of white spaces before each
word. These are all words in the simplest sense. But this method would
yield a very insensitive measure of vocabulary knowledge, given that
many words are repeated. For example, the word ‘the’ appears seven times
in this paragraph. The same word can also appear in different forms.
Does the researcher count ‘word’ and ‘words’ as one or two words? As a
result, although estimating vocabulary size is a quantitative process, the
researcher must make qualitative distinctions as to if and how individual
word forms are counted. Several alternatives are available.
Type/Token Ratio A basic distinction can be made between the first

appearance of a word in a text and that same word being repeated, or
what is termed the type/token ratio (TTR). Types refer to the various
unique words in a text, counted by the first appearance of a word. Tokens
refer to every subsequent appearance of that word type in the text. The
phrase ‘the big cat in the big hat’ has five types (the, big, cat, in, hat) out
of seven total tokens. The TTR is a measure of lexical diversity originally
developed for measuring L1 vocabulary development. It is an index of

lexical diversity and not a measure of absolute size, but it is reasonable to
assume that users who produce a wider variety of words—that is, have a
higher TTR—will also have larger vocabularies. In practice, however, the
measure has been shown to be relatively insensitive to differences in pro-
ficiency levels (Richards 1987).
Lemmas Estimating vocabulary size requires a way to identify the

number of distinct word meanings represented in all the tokens in a
text. One approach widely used in corpus linguistics is the lemma. A
lemma (or citation form) is a particular form of a word (or lexeme) that
is chosen by convention to represent the canonical form. These forms
are typically used in dictionaries as the headwords. Lemmas consist of
all regularly inflected forms sharing the same stem and belonging to
the same syntactic category. The lemma for the verb bank, for exam-
ple, includes the verb forms banks, banked, and banking. A separate
lemma is assumed for the noun bank (as an institution) and its plural,
banks. A further distinction is made between the lemmas representing
homonyms, such as in river bank and bank loan, though these can
pose a particular problem for computer- based corpus analyses.
(Aitchison 2012).
Word Families Related to the lemma is the word family, which is defined
as the base word form plus its inflections and most common derivational
variants, for example, invite, invites, inviting, invitation (Hirsh and Nation
1992, p. 692). English inflections include third person -s, past participle
-ed, present participle -ing, plural -s, possessive -s, and comparative -er and
superlative -est. Derivational affixes include -able, -er, -ish, -less, -ly, -ness,
-th, -y, non-, un-, -al, -ation, -ess, -ful, -ism, -ist, -ity, -ize, -ment, and in-
(Hirsh and Nation 1992, p. 692). As with the lemma, the underlying idea
is that a base word and its inflected forms express the same core meaning,
and thus can be considered learned words if a learner knows the base and
the affix rules. Bauer and Nation (1993) proposed seven levels of affixes,
which include derivations and inflections. Word families differ from lem-
mas in that they cross syntactic categories. In the example of bank, as
above, the noun and verb forms are counted as part of the same family.
As a result, a lemma count will always be larger than the word family
count, given the narrower range of forms counted as a single instance.
Milton identifies what he terms a ‘very crude’ equivalence of lemma to
word family involving multiplying the word family size by 1.6 to get the
approximate lemma size (Milton 2009, p. 12).
he Word Family as a Unit of Recognition

T
The word family has been widely used as the unit of counting in vocabu-
lary size studies (Schmitt 2010). Nation has argued that the word family
is a particularly appropriate unit for studying L2 recognition vocabulary
because it is primarily about meaning and meaning potential (Nation
2006, p. 76). It also has a degree of psycholinguistic reality regarding how
the different forms in a given family are stored in the mental lexicon
(Nagy et al. 1989). The basic assumption is that if the meaning of the
base word is known, the various inflections and derivations in which it
appears will also be potentially understood, at least to some degree. This
assumption has proved to be useful in relating individual vocabulary size
to L2 use, but is one that is not categorical. The assumption that a learner
who knows the meaning of build will understand the meaning of builder
on the first encounter is a probabilistic one. Schmitt and Zimmerman
(2002) show that university-level ESL students’ knowledge of the derived
forms of many stem words is far from complete, for example, not knowing
that persistent, persistently, and persistence all come from persist. However,
they also recognize that users will probably work out the meaning of per-
sistence faster if they knew persist than if they did not.
The word family construct also conflates the distinction that Paradis
(2009) makes between the stock of form–meaning associations stored in
declarative memory and morphological processes that are procedural in
nature. Widely used tests of vocabulary size, the VLT (Nation 2013) and
the Yes/No Test (Meara and Buxton 1987), always present the base form
as the test item, thus sidestepping any attempt to measure the morpho-
logical knowledge assumed in the word family construct.
While recognizing these limitations, the word family is nonetheless an

easy-to-understand and widely used measure of recognition vocabulary
size, and is used in the research studies reported in Part 2. From here
forward, word and word family will be used interchangeably when refer-
ring to size; that is, a reference to ‘the number of words’ will mean the
same as ‘the number of word families’.
ord Frequency Statistics as an Index of Vocabulary

W
Size
Figuring out how many words a user knows is the next challenge for the
vocabulary size researcher. While in theory it may be possible to identify
every single word a user knows, in practice, the process of fixing vocabu-
lary size is one of estimation. A vocabulary size estimate is based on a
finite sample of a user’s knowledge obtained in a specific task or set of
tasks. Recognition vocabulary knowledge is passive by nature, and evi-
dence for it must be elicited from the user. This is done by presenting a
set of words to a user and eliciting a response that indicates whether the
items are known. Time and resource limitations mean that any test can
present only a limited number of words, and it is from this limited sam-
ple that the user’s vocabulary size is estimated. Word frequency statistics
provide the vocabulary size researcher with a reliable and objective means
to index the size of recognition (and productive) vocabulary knowledge
(Laufer 2001).
Words greatly differ in how often they occur in a given language.
When the words in a large corpus of spoken or written English are rank-
ordered from the most to least frequently occurring, a highly distinctive
pattern emerges. The 2000–3000 most frequently occurring words
account for the vast majority of tokens that appear in the corpus. Beyond
these high-frequency words, the relative frequency of a given word
steadily decreases as a function of its relative order, until the very-low-
frequency words tail off and account for only a tiny proportion of tokens.
This frequency distribution, called Zipf ’s law (after one of its original
discoverers), provides an index for the measurement and interpretation of
vocabulary size. The law states that, for a corpus of natural language
utterances, the frequency of any word is in inverse proportion to its rank
in the frequency table. This inverse proportionality means that words

occur in a predictable and distinctive pattern. The most frequent word
occurs approximately twice as often as the second most frequent word,
three times as often as the third most frequent word, and so on. For
example, in the classic Brown Corpus, the word ‘the’ is the most fre-
quently occurring word, and by itself accounts for nearly 7% of all word
occurrences. Reflecting Zipf ’s law, the second-place word, ‘of ’, accounts
for slightly over 3.5% of words (36,411 occurrences), followed by ‘and’
at 2.8% (28,852). The first 135 words alone account for half the Brown
Corpus tokens (Biber et al. 1998).
When describing vocabulary size using these frequency counts,
researchers usually fix size in increments of 1000 words, that is, the 1000
most frequent (1–999th), the 2000 most frequent (1000–1999th), and
so on. The shorthand 1K, 2K, 3K, and so on is used to refer to these
bands throughout the book. Words sampled from selected bands are used
in testing instruments that systematically elicit word knowledge across a
range of frequency levels and quantify this knowledge in an objective and
context-independent way.
The use of word frequency statistics is a distinctive feature of the
vocabulary size approach. The development of very large and accessible
corpora has resulted in increasingly refined frequency counts of spoken
and written language use. These corpora include the British National
Corpus (BNC) (Leech et al. 2001), the Cambridge and Nottingham
Corpus of Discourse in English (CANCODE) (McCarthy 1998), and
the Corpus of Contemporary American English (COCA) (http://corpus.
byu.edu/coca/), among others. The increasing availability of vocabulary
software tools at high-quality websites such as the Complete Lexical
Tutor, www.lextutor.ca/, and Lancaster University, corpora.lancs.ac.uk,
permits researchers and practitioners to define and map learner vocabu-
lary size onto receptive and productive L2 performance in an increasingly
sophisticated manner.
In summary, the word family is an attested unit of recognition vocabu-
lary knowledge that has been widely used in vocabulary size research. It is
also used in the studies reported in this book, in conjunction with
word frequency statistics, to estimate user vocabulary size. The pat-
terns of word frequency have more than just a descriptive function, as
1.3 Vocabulary Size as a Dimension of Learners’ Vocabulary... 13
they also have direct implications for vocabulary learning and the repre-
sentation of this knowledge in the mental lexicon. This is discussed below.
1.3 ocabulary Size as a Dimension

V
of Learners’ Vocabulary Knowledge
The preceding has introduced the logic of vocabulary size research. The
next issue to consider is what individual differences in vocabulary size tell
us about a user’s underlying vocabulary knowledge and how it relates to
L2 learning and use. In other words, why measure vocabulary size?
ocabulary Size as a Dimension of L2

V
The vocabulary size approach focuses on the size, or breadth, of an indi-

vidual’s vocabulary knowledge. This breadth is established by identifying
how many words a user can recognize. This can be done by matching a
presented word with a basic definition, as is done in the VLT, or simply
by indicating that the word is known, as in the Yes/No Test. Both tests are
examined in the following chapter. Correct performance on the VLT
shows that a user knows a basic meaning for a word. It does not indicate
whether the user knows all, or even any other, meanings for the word.
Accuracy on the Yes/No Test shows that the test-taker can link some
meaning with the word form; however, the nature of that meaning is an
open question. At a minimum, it might just be that a particular word
exists in the language.
As is evident in Fig. 1.1, vocabulary size tests directly measure only a
very small part of what it means to know a word. Complete word knowl-
edge consists of form, meaning, and use. Form concerns the perceptual
and physical shape of a word, including both how it is pronounced and
how it is spelled. The meaning of a word includes knowledge of its basic
meaning and the words typically associated with it. Word meaning also
encompasses the range of lexical relationships that the word has with
other words in the mental lexicon. These links include polysemy,
antonymy, homonymy, synonymy, and other relational links such as

metonymy (Aitchison 2012). Knowledge of word use is the third part of
word knowledge and arguably the most important. Words are typically
used in particular combinations dictated by the requirements of the set-
ting, goal, and participants. Using a word appropriately also involves a
range of conceptual and world knowledge that goes beyond the mental
lexicon proper.
Performance on vocabulary size tests reflects only how many form–
meaning mappings the user knows, and these typically correspond to the
most basic meaning or meanings. These tests do not directly tap the user’s
knowledge of the range of word meanings or uses, that is, depth. Although
the relationship between breadth and depth is open to debate (Qian
1999; Read 2004), it is assumed here that the two are not independent
dimensions (Vermeer 2001). The assumption is that depth of knowledge
emerges with increasing size; that is, evidence that a user knows a large
number of these form–meaning mappings implies that vocabulary depth
knowledge is also present to some degree. However, vocabulary size tests
are ultimately not about qualitative differences in vocabulary knowledge,
but instead provide an estimate of the number of words (minimally
form–meaning mappings) the user has. The relative size of this vocabu-
lary stock is assumed to represent a basic constraint on performance in
the L2. How strong a constraint is a central focus of the research reported
in Part 2.
In addition to assumptions about the nature of the underlying vocabu-
lary knowledge, the vocabulary size approach also makes assumptions
about how this knowledge is acquired and used.
ocabulary Learning Assumptions in the Vocabulary

V
Size Approach
A defining characteristic of the vocabulary size approach is the use of

word frequency statistics to estimate individual vocabulary size. Schmitt
(2010) states that frequency ‘is arguably the single most important char-
acteristic of lexis that researchers must address’ (p. 64). Frequency is par-
ticularly important in the development of recognition vocabulary, which
is driven by exposure to the language. This exposure is both intended and

incidental. Vocabulary learning is thus assumed to be input driven (Ellis
2002). The likelihood that a given word is known can be predicted to a
large extent by the frequency with which it appears in the language. All
things being equal, words that occur more frequently will be learned
sooner. Over time, they will also be accessed more quickly. High-
frequency words are learned before mid-frequency words, which in turn
are acquired before low-frequency words. In short, vocabulary size is an
emergent property of L2 knowledge development, driven by the input the
learner receives.
Figure 1.2 depicts this frequentist model in schematic terms. Three
things should be noted. First and foremost, the more frequently a word
is used in a given language, the more likely it is that an individual will
know the word. Second, word knowledge (as reflected by accurate test
performance) forms a gradient across frequency bands such that knowl-
edge in specific bands increases proportionally from the high-frequency
to low-frequency bands. Proportionally, more high-frequency words are
known than mid-frequency words, and in turn, more mid-frequency
words are known than low-frequency words. Finally, the relationship
between size and learning is expressed in probabilistic terms. It is not
assumed that higher-frequency words are never learned before lower-fre-
quency ones. Individual differences in life experience mean that beginner
users will know some very-low-frequency words as a reflection of their
interests and experience. The key point is that word frequency statistics
Word Knowledge as a Function of the Frequency of

Occurrence of the Word in the Language
100
Likelihood of knowing
80
60
40
20
0
High Mid Low
Frequency of occurence
Fig. 1.2 A frequentist model of vocabulary learning

rovide an objective means to scale learner vocabulary size as probabi-

p
listic estimates that can, in turn, serve as sensitive discriminators of user
proficiency differences and performance outcomes.
Frequency of Occurrence and Diversity of Context
Frequency is a quantitative variable that ignores where words appear.

Quantity, not quality, is what matters. However, recent work indicates
that the effects of frequency on learning, and particularly on use, may not
merely be the function of the number of times a word occurs, but instead
reflect the range of contexts in which it is encountered. Research suggests
that frequency of occurrence is strongly correlated, if not confounded,
with contextual diversity when accounting for speed of lexical decision test
performance (Adelman et al. 2006) and L2 vocabulary learning (Crossley,
et al. 2013). This suggests that word frequency statistics may have a quali-
tative dimension, serving as surrogates for contextual variety. Frequency
may also have a qualitative dimension beyond mere token counts.
Frequency of Occurrence and Text Coverage
The primary focus of the vocabulary size approach is the relationship

between vocabulary size and L2 performance. Evidence for this relation-
ship takes two forms. One is the relationship between vocabulary size and
L2 proficiency as measured in proficiency standards such as the Common
European Framework (Milton 2009) and more localized performance
measures such as a placement test (Harrington and Carey 2009). This
approach is exemplified in the studies reported in Part 2 and will be dis-
cussed at length there. But there is another way in which vocabulary size
has been related to L2 performance. There has been a long-standing inter-
est in how user vocabulary size relates to the lexical demands of the text.
The rudimentary question driving this research is the number of words a
reader needs to know to ensure that all the words encountered in a given
text are recognized, that is, text coverage (Laufer 1992). Text coverage has
been widely used as a graded measure of the relative difficulty the reader
will have in comprehending a text and, in turn, as a means to identify
Table 1.1 Vocabulary size expressed in word families and text coverage (written
and spoken) across ninea corpora (Nation 2006, p. 79)
Knowledge of all Approximate written text Approximate spoken text
word in coverage (%) coverage (%)
1K 78–81 81.84
2K 8–9 5–6
3K 3–5 2–3
4K–5K 3 1.5–3
6K–9K 2 0.75–1
10K–14K <1 0.5
Proper nouns 2–4 1–1.5
+14K 1–3 1
a
Corpora analyzed: Lancaster–Oslo–Bergen (LOB) Corpus, Freiburg–LOB, Brown,
Frown, Kohlapur, Macquarie, Wellington Written, Wellington Spoken, and
Lund, available from the International Computer Archive of Modern and
Medieval English at http://gandalf.aksis.uib.no/icame.html (Nation 2006, p. 63).
vocabulary learning needs. A number of studies have examined the user

size–text coverage relationship in both spoken and written texts (Adolphs
and Schmitt 2003; Cobb 2007; Hazenberg and Hulstijn 1996; Hsueh-
Chao and Nation 2000; Laufer 1989, 1992; Laufer and Ravenhorst-
Kalovski 2010; Milton, 2009; Nation 2006; Schmitt et al. 2011; Webb and
Rodgers 2009; van Zeeland and Schmitt 2013).
Nation (2006) examined the relationship between size and text cover-
age in nine spoken and written corpora covering a range of text types.
Table 1.1 sets out the correspondence of vocabulary size with text cover-
age levels.
The data in Table 1.1 are presented visually in Fig. 1.3 to illustrate the
profound effect of bands 1K–3K on how much of the text will be recog-
nized. Knowledge of the 1K band alone allows a user to recognize around
80% of the words occurring in written or spoken texts.
The 1K band contains all of the function words (articles, prepositions,
pronominals, auxiliary verbs, conjunctions) that account for a significant
amount of coverage in any given text. Knowledge of the 2K band pro-
vides an additional 10% coverage in written texts and 5% coverage in
spoken texts. The 3K words provide an additional coverage of about 5%
and 3%, respectively. Past the 5K band, there is only a small increase in
text coverage as a function of increasingly lower-frequency bands. Proper
100
90
80
Percentage of coverage
70
60
50 Spoken Wrien
40
30
20
10
0
1K 2K 3K 4–5K 6–9K 10–14K
Frequency band
Fig. 1.3 Cumulative percentage of text coverage and corresponding frequency

bands
% of text Number of unfamiliar Number of text lines per 1
coverage words per 100 words unfamiliar word.
99 1 10
98 2 5
95 5 2
90 10 1
80 20 0.5
Note: Assumes ten words per line
Fig. 1.4 Text coverage as the number of unfamiliar words and the number of
lines of text per unfamiliar word
nouns are usually treated separately in size–text coverage discussions

because they are highly text and context specific.
So how do these various text coverage levels map onto text comprehen-
sion? Figure 1.4 sets out the relationship between text coverage levels and
the number of unfamiliar words a reader will encounter.
There are three ___ that must be considered when ______
to ______ how the most _______ of ____ coverage relates
to reading ____ ____. The first is the ___ of
reading, or ____ Reading a _____ is different from reading
a novel. The second is the nature of ______. Reading
for general _______ is different from reading for _____. The
third is the way in which ______ is _______.
All these _____must be taken into consideration when ______
to ___ how differences in ___ coverage might affect ______.
Fig. 1.5 A sample reading text with 80% text coverage
Learning the 1K band alone means that a reader should recognize

approximately 80% of the words in a text. While this might appear to be
enough to make some sense of a text, in fact, it is woefully inadequate for
even minimal understanding. The text in Fig. 1.5 has been modified so
that 20% of the words are missing. The omissions render the text almost
incomprehensible. Note that the unfamiliar words are usually the most
important for understanding a given text.
The various studies differ in aim, setting, and size, but a consensus has
emerged as to vocabulary size needs for key text coverage thresholds. For
written texts, it is generally agreed that 95% text coverage requires knowl-
edge of 1K–3K bands, and that this will result in only a basic level of
comprehension. For 98%–99% text coverage, up to the 8K–9K range is
needed, and it is only in this range that reading starts to become fluent.
These generalizations are relatively stable across text types and genres,
which can differ significantly in difficulty and purpose.
The relationship between size and spoken text coverage has received
somewhat less attention than the reading text research, but the findings
are generally the same. Adolphs and Schmitt (2003) found that 2K–3K
word families were sufficient for 95% text coverage, a number similar to
the written text research. In contrast, knowledge of only the 6K–7K
bands was needed for 98% text coverage, lesser than the 8K–9K sug-
gested as being necessary to read authentic texts with some degree of
fluency (Nation 2006). van Zeeland and Schmitt (2013) also reported
that listening comprehension required knowledge of fewer word families
than comparable reading levels.
A question remains as to whether these text coverage levels, particu-
larly the 95% and 98% levels, reflect a qualitative threshold that must be
met for adequate comprehension, or a continuum from lesser to greater
comprehension skill. Schmitt et al. (2011) examined this issue by plot-
ting text coverage levels against performance for 600 tertiary L2 English
readers from 12 different countries. The relationship between text cover-
age and comprehension was plotted at ten text coverage levels, ranging
from 90% to 100% coverage. See Fig. 1.6.
This figure is adapted from Schmitt et al. (2011, p. 34), with only
alternating text coverage levels reported here. A consistent linear relation-
ship is evident across the reading comprehension and vocabulary cover-
age levels. There is little suggestion of discrete thresholds at the 95% or
100
90
Comprehension percentage
80
70
60
50
40
30
1+SD Mean 1-SD
20
10
0
90% 92% 94% 96% 98% 99% 100%
(n=21) (n=39) (n=93) (n=176) (n=200) (n=186) (n=187)
Vocabulary Coverage and Number of Parcipants at Each Level
Fig. 1.6 Text comprehension percentage as a function of vocabulary coverage

(Schmitt et al. 2011, p. 34)
1.4 Conclusions 21
the 98% coverage level, with the comprehension slopes trending up in a

consistent way. It is important to note that the respective levels represent
averages that hide highly significant individual variability. For example,
comprehension scores at +1 standard deviation for the 92% coverage
group are nearly identical to the mean performance of the 98% group.
Although Schmitt et al. (2011) reject the idea of discrete thresholds, they
do endorse 98% text coverage as a target for achieving adequate compre-
hension, and with it the 8K–9K bands as the target identified by Nation
and others as needed for meeting this goal (Nation 2006). The significant
variability underscores the probabilistic nature of fixing vocabulary size
and its relationship to L2 performance.
1.4 Conclusions
The vocabulary size approach is based on the simple assumption that the
number of words an individual knows has a direct relationship to L2
proficiency. The focus here is on recognition vocabulary knowledge,
which is narrowly defined as the ability to recognize the association
between a single word form and a basic meaning. Vocabulary learning is
viewed as an input-driven process in which vocabulary size emerges from
the user’s experience with the language. Corpus-based word frequency
statistics provide a means of estimating the overall vocabulary size from
recognition performance on a limited set of words. These vocabulary size
estimates have been related to L2 proficiency and use in two ways.
Vocabulary size has been examined as a predictor of differences in L2
performance as measured by standardized and context-specific tests. As a
key component of the lexical facility construct, this use of vocabulary size
is the focus of the book. Considerable attention has also been given to the
relationship between vocabulary size and text coverage, the latter reflect-
ing the comprehension demands of written or spoken texts. Vocabulary
size thresholds have been proposed to meet the levels of text coverage
required for successful comprehension. Both uses demonstrate the utility
of vocabulary size as a dimension of L2 vocabulary knowledge and a cor-
relate of L2 proficiency.
Another random document with
no related content on Scribd:
Sec. 3. In the prosecution of slaves for crimes of higher grade than
petit larceny, the legislature shall have no power to deprive them of
an impartial trial by a petit jury.
Sec. 4. Any person who shall maliciously dismember, or deprive a
slave of life, shall suffer such punishment as would be inflicted in
case the like offence had been committed on a free white person, and
on the like proof, except in case of insurrection of such slave.
Free Negroes.
Bill of Rights, Sec. 23. Free negroes shall not be allowed to live in
this state under any circumstances.
Article VIII.—Elections and Rights of Suffrage.
Sec. 1. Every male citizen of the United States, above the age of
twenty-one years, having resided in this state one year, and in the
county, city, or town in which he may offer to vote, three months
next preceding any election, shall have the qualifications of an
elector, and be entitled to vote at all elections. And every male citizen
of the United States, above the age aforesaid, who may be a resident
of the state at the time this constitution shall be adopted, shall have
the right of voting as aforesaid; but no such citizen or inhabitant
shall be entitled to vote except in the county in which he shall
actually reside at the time of the election.
The Topeka Constitution.
The following are the political features of the Topeka constitution:
Slavery.
Bill of Rights, Sec. 6. There shall be no slavery in this state, nor

involuntary servitude, unless for the punishment of crime.
Amendments to the Constitution.
Sec. 1. All propositions for amendments to the constitution shall

be made by the General Assembly.
Sec. 2. A concurrence of two-thirds of the members elected to each
house shall be necessary, after which such proposed amendments
shall be again referred to the legislature elected next succeeding said
publication. If passed by the second legislature by a majority of two-
thirds of the members elected to each house, such amendments shall
be republished as aforesaid, for at least six months prior to the next
general election, at which election such proposed amendments shall
be submitted to the people for their approval or rejection; and if a
majority of the electors voting at such election shall adopt such
amendments, the same shall become a part of the constitution.
Sec. 3. When more than one amendment is submitted at the same
time, they shall be so submitted as to enable the electors to vote upon
each amendment separately. No convention for the formation of a
new constitution shall be called, and no amendment to the
constitution shall be, by the general assembly, made before the year
1865, nor more than once in five years thereafter.
Submission of Constitution to the People.
Schedule, Sec. 2. That this constitution shall be submitted to the

people of Kansas for ratification on the 15th day of December next.
That each qualified elector shall express his assent or dissent to the
constitution by voting a written or printed ticket, labelled
“Constitution,” or “No Constitution;” which election shall be held by
the same judges, and conducted under the same regulations and
restrictions as is hereinafter provided for the election of members of
the general assembly.
The Douglas Amendment.
The following is the Douglas amendment, which really formed the

basis of the bill for admission:
“It being the true intent and meaning of this act not to legislate
slavery into any state or territory, nor to exclude it therefrom, but to
leave the people thereof perfectly free to form and regulate their
domestic institutions in their own way, subject only to the
Constitution of the United States.”
The bill which passed on the 4th of May was known as the English
bill, and it met the approval of Buchanan. To the measure was
attached “a fundamental condition precedent,” which arose from the
fact that the ordinance of the convention accompanying the
constitution claimed for the new State a cession of the public lands
six times greater than had been granted to other States, amounting
in all to 23,500,000 acres. In lieu of this Congress proposed to
submit to a vote of the people a proposition specifying the number of
acres and the purposes for which the money arising from their sale
were to be used, and the acceptance of this was to be followed by a
proclamation that “thereafter, and without further proceedings from
Congress the admission of the State of Kansas, into the Union, upon
an equal footing with the original States in all respects whatever,
shall be complete and absolute.” The condition was never fulfilled,
for the people at the election on the 2d of August, 1858, rejected it by
a majority of 9,513, and Kansas was not admitted under the
Lecompton constitution.
Finally, and after continued agitation, more peaceful, however,
than that which characterized the earlier stages of the struggle, the
territorial legislature of Kansas called an election for delegates to
meet and form a constitution. They assembled in convention at
Wyandot, in July, 1859, and reported a constitution prohibiting
slavery. This was adopted by a majority exceeding 4000, and under it
Kansas was admitted to the Union on the 29th of January, 1861.
The comparative quiet between the rejection of the English
proposition and the adoption of the Wyandot constitution, was at
one time violently disturbed by a raid made by John Brown at
Harper’s Ferry, with a view to excite the slaves to insurrection. This
failed, but not before Gov. Wise, of Virginia, had mustered his
militia, and called for the aid of United States troops. The more
radical anti-slavery men of the North were at first shocked by the
audacity of an offense which many looked upon as an act of treason,
but the anxiety oi Virginia to hang Brown and all his followers who
had been captured alive, changed a feeling of conservatism in the
North to one of sympathy for Brown and deeper hatred of slavery. It
is but fair to say that it engendered hostility to the Union in the
South. The right and wrong of slavery was thereafter more generally
discussed than ever. The talent of the South favored it; while, with at
least a large measure of truth it can be said that the talent of the
North opposed it. So bitter grew the feeling that soon the churches of
the sections began to divide, no other political question having ever
before disturbed the Union.
We have not pretended to give a complete history of the Kansas
trouble either in that State or in Congress, nor yet a full history of the
many issues raised on questions which were but subsidiary to the
main one of slavery. Our object is to show the relation of the political
parties throughout that struggle, for we are dealing with the history
of parties from a national view, and not with battles and the minor
questions or details of parliamentary struggles. The contest had
cemented the Democrats of the South as it had the Republicans of
the North; it divided both the Democrats of the North and the
Americans in all sections. John Bell, of Tennessee, and Sam Houston
of Texas, recognized leaders of the Americans, had shown their
sympathy with the new stand taken by Douglas, as early as 1854.
Bell, however, was less decided than Houston, and took his position
with many qualifications. Houston opposed even the repeal of the
Missouri Compromise, and made the last speech against it in the
Senate. He closed with these words:
“In the discharge of my duty I have acted fearlessly. The events of
the future are left in the hands of a wise Providence, and, in my
opinion, on the decision which we make upon this question must
depend union or disunion.”
These sentiments were shared by many Americans, and the great
majority of them drifted into the Republican party. The Abolitionists
from the beginning of the struggle, allied themselves with the
Republicans, a few of their leaders proclaiming, however, that this
party was not sufficiently advanced in its views.
The Charleston Convention.
Such was the condition of the parties when the Democratic

national convention met at Charleston, S. C., on the 23d of April,
1860, it being then the custom of the Democratic party, as it is of all
majority parties, to call its convention first. It was composed of
delegates from all the thirty-three States of the Union, the whole
number of votes being 303. After the example of former Democratic
conventions it adopted the two-third rule, and 202 votes were
required to make nominations for President and Vice-President.
Caleb Cushing, of Mass., presided. From the first a radical difference
of opinion was exhibited among the members on the question of
slavery in the Territories. Almost the entire Southern and a minority
of the Northern portion believed in the Dred Scott decision, and held
that slave property was as valid under the constitution as any other
class of property. The Douglas delegates stood firmly by the theory of
popular sovereignty, and avowed their indifference to the fact
whether it would lead to the protection of slave property in the
territories or not. On the second day a committee on resolutions
consisting of one member from each State, selected by the State
delegates, was named, and then a resolution was resolved
unanimously “that this convention will not proceed to ballot for a
candidate for the Presidency until the platform shall have been
adopted.” On the fifth day the committee on resolutions presented
majority and minority reports.
After a long discussion on the respective merits of the two reports,
they were both, on motion of Mr. Bigler, of Pennsylvania,
recommitted to the Committee on Resolutions, with a view, if
possible, to promote harmony; but this proved to be impracticable.
On the sixth day of the Convention (Saturday, April 28th,) at an
evening session, Mr. Avery, of North Carolina, and Mr. Samuels, of
Iowa, from the majority and minority of the committee, again made
opposite and conflicting report on the question of slavery in the
Territories. On this question the committee had divided from the
beginning, the one portion embracing the fifteen members from the
slaveholding States, with those from California and Oregon, and the
other consisting of the members from all the free States east of the
Rocky Mountains. On all other questions both reports substantially
agreed.
The following is the report of the majority made on this subject by
Mr. Avery, of North Carolina, the chairman of the committee:
“Resolved, That the platform adopted by the Democratic party at
Cincinnati be affirmed with the following explanatory resolutions:
1st. That the Government of a Territory, organized by an act of
Congress, is provisional and temporary, and during its existence all
citizens of the United States have an equal right to settle with their
property in the Territory, without their rights, either of person or
property, being destroyed or impaired by Congressional or
Territorial legislation. 2d. That it is the duty of the Federal
Government, in all its departments, to protect, when necessary, the
rights of persons and property in the Territories, and wherever else
its constitutional authority extends. 3d. That when the settlers in a
Territory having an adequate population form a State Constitution,
the right of sovereignty commences, and being consummated by
admission into the Union, they stand on an equal footing with the
people of other States, and the State thus organized ought to be
admitted into the Federal Union whether its constitution prohibits or
recognizes the institution of slavery.”
The following is the report of the minority, made by Mr. Samuels,
of Iowa. After reaffirming the Cincinnati platform by the first
resolution, it proceeds: “Inasmuch as differences of opinion exist in
the Democratic party, as to the nature and extent of the powers of a
Territorial Legislature, and as to the powers and duties of Congress,
under the Constitution of the United States, over the institution of
slavery within the Territories, Resolved, That the Democratic party
will abide by the decisions of the Supreme Court of the United States
upon questions of constitutional law.”
After some preliminary remarks, Mr. Samuels moved the adoption
of the minority report as a substitute for that of the majority. This
gave rise to an earnest and excited debate. The difference between
the parties was radical and irreconcilable. The South insisted that the
Cincinnati platform, whose true construction in regard to slavery in
the Territories had always been denied by a portion of the
Democratic party, should be explained and settled by an express
recognition of the principles decided by the Supreme Court. The
North, on the other hand, refused to recognize this decision, and still
maintained the power to be inherent in the people of a Territory to
deal with the question of slavery according to their own discretion.
The vote was then taken, and the minority report was substituted for
that of the majority by a vote of one hundred and sixty-five to one
hundred and thirty-eight. The delegates from the six New England
States, as well as from New York, Ohio, Indiana, Illinois, Michigan,
Wisconsin, Iowa, and Minnesota, fourteen free States, cast their
entire vote in favor of the minority report. New Jersey and
Pennsylvania alone among the free States east of the Rocky
Mountains, refused to vote as States, but their delegates voted as
individuals.
The means employed to attain this end were skillfully devised by
the minority of the Pennsylvania delegation in favor of nominating
Mr. Douglas. The entire delegation had, strangely enough, placed
this power in their hands, by selecting two of their number, Messrs.
Cessna and Wright, to represent the whole on the two most
important committees of the Convention—that of organization and
that of resolutions. These gentlemen, by adroitness and
parliamentary tact, succeeded in abrogating the former practice of
casting the vote of the State as a unit. In this manner, whilst New
York indorsed with her entire thirty-five votes the peculiar views of
Mr. Douglas, notwithstanding there was in her delegation a majority
of only five votes in their favor on the question of Territorial
sovereignty, the effective strength of Pennsylvania recognizing the
judgment of the Supreme Court, was reduced to three votes, this
being the majority of fifteen on the one side over twelve on the other.
The question next in order before the Convention was upon the
adoption of the second resolution of the minority of the committee,
which had been substituted for the report of the majority. On this
question Georgia, Louisiana, Alabama, Arkansas, Texas, Florida, and
Mississippi refused to vote. Indeed, it soon appeared that on the
question of the final adoption of this second resolution, which in fact
amounted to nothing, it had scarcely any friends of either party in
the Convention. The Douglas party, without explanation or addition,
voted against it. On the other hand, the old Democracy could not
vote for it without admitting that the Supreme Court had not already
placed the right over slave property in the Territories on the same
footing with all other property, and therefore they also voted against
it. In consequence the resolution was negatived by a vote of only
twenty-one in its favor to two hundred and thirty-eight. Had the
seven Southern States just mentioned voted, the negatives would
have amounted to two hundred and eighty-two, or more than
thirteen to one. Thus both the majority and the minority resolutions
on the Territorial question were rejected, and nothing remained
before the Convention except the Cincinnati platform.
At this stage of the proceedings (April 30th), the States of
Louisiana, Alabama, South Carolina, Mississippi, Florida, Texas, and
Arkansas, having assigned their reasons for the act, withdrew in
succession from the Convention. After these seven States had retired,
the delegation from Virginia made an effort to restore harmony. Mr.
Russell, their chairman, addressed the Convention and portrayed the
alarming nature of the crisis. He expressed his fears that we were on
the eve of a revolution, and if this Convention should prove a failure
it would be the last National Convention of any party which would
ever assemble in the United States. “Virginia,” said he, “stands in the
midst of her sister States, in garments red with the blood of her
children slain in the first outbreak of the ‘irrepressible conflict.’ But,
sir, not when her children fell at midnight beneath the weapon of the
assassin, was her heart penetrated with so profound a grief as that
which will wring it when she is obliged to choose between a separate
destiny with the South, and her common destiny with the entire
Republic.”
Mr. Russell was not then prepared to answer, in behalf of his
delegation, whether the events of the day (the defeat of the majority
report, and the withdrawal of the seven States) were sufficient to
justify her in taking the irrevocable step in question. In order,
therefore, that they might have time to deliberate, and if they
thought proper make an effort to restore harmony in the Convention,
he expressed a desire that it might adjourn and afford them an
opportunity for consultation. The Convention accordingly adjourned
until the next day, Tuesday, May 1st; and immediately after its
reassembling the delegation from Georgia, making the eighth State,
also withdrew.
In the mean time the Virginia delegation had consulted among
themselves, and had conferred with the delegation of the other
Southern States which still remained in the Convention, as to the
best mode of restoring harmony. In consequence Mr. Howard, of
Tennessee, stated to the Convention that “he had a proposition to
present in behalf of the delegation from Tennessee, whenever, under
parliamentary rules, it would be proper to present it.” In this
Tennessee was joined by Kentucky and Virginia. He should propose
the following resolution whenever it would be in order: ‘Resolved,
That the citizens of the United States have an equal right to settle
with their property in the Territories of the United States; and that,
under the decision of the Supreme Court of the United States, which
we recognize as the correct exposition of the Constitution of the
United States, neither the rights of person nor property can be
destroyed or impaired by Congressional or Territorial legislation.’
On a subsequent day (May 3d), Mr. Russell informed the
Convention that this resolution had, “he believed, received the
approbation of all the delegations from the Southern States which
remained in the Convention, and also received the approbation of the
delegation from New York. He was informed there was strength
enough to pass it when in order.”
Mr. Howard, however, in vain attempted to obtain a vote on his
resolution. When he moved to take it up on the evening of the day it
had been offered, he was met by cries of “Not in order,” “Not in
order.” The manifest purpose was to postpone its consideration until
the hour should arrive which had been fixed by a previous order of
the Convention, in opposition to its first order on the same subject,
for the balloting to commence for a Presidential candidate, when it
would be too late. This the friends of Mr. Douglas accomplished, and
no vote was ever taken upon it either at Charleston or Baltimore.
Before the balloting commenced Mr. Howard succeeded, in the
face of strong opposition, with the aid of the thirty-five votes from
New York, in obtaining a vote of the Convention in re-affirmance of
the two-thirds rule. On his motion they resolved, by 141, to 112 votes,
“that the President of the Convention be and he is hereby directed
not to declare any person nominated for the office of President or
Vice-President, unless he shall have received a number of votes equal
to two-thirds of the votes of all the electoral colleges.” It was well
known at the time that this resolution rendered the regular
nomination of Mr. Douglas impossible.
The balloting then commenced (Tuesday evening, May 1st), on the
eighth day of the session. Necessary to a nomination, under the two-
thirds rule, 202 votes. On the first ballot Mr. Douglas received 145½
votes; Mr. Hunter, of Virginia, 42; Mr. Guthrie, of Kentucky, 35½;
Mr. Johnson, of Tennessee, 12; Mr. Dickinson, of New York, 7; Mr.
Lane, of Oregon, 6; Mr. Toucey, of Connecticut, 2½; Mr. Davis, of
Mississippi, 1½; and Mr. Pearce, of Maryland, 1 vote.
The voting continued until May 3d, during which there were fifty-
four additional ballotings. Mr. Douglas never rose to more than
152½, and ended in 151½ votes, 202 votes being necessary to a
nomination.
Until 1824 nominations had been made by Congressional caucus.
In these none participated except Senators and Democratic States,
and Representatives from Democratic Congressional districts. The
simple majority rule governed in these caucuses, because it was
morally certain that, composed as they were, no candidate could be
selected against the will of the Democratic States on whom his
election depended. But when a change was made to National
Conventions, it was at once perceived that if a mere majority could
nominate, then the delegates from Anti-Democratic States might be
mainly instrumental in nominating a candidate for whom they could
not give a single electoral vote. Whilst it would have been harsh and
inexpedient to exclude these States from the Convention altogether,
it would have been unjust to confer on them a controlling power over
the nomination. To compromise this difficulty, the two-thirds rule
was adopted. Under its operation it would be almost impossible that
a candidate could be selected, without the votes of a simple majority
of delegates from the Democratic States. This was the argument of its
friends.
It had now become manifest that it was impossible to make a
nomination at Charleston. The friends of Mr. Douglas adhered to
him and would vote for him and him alone, whilst his opponents,
apprehending the effect of his principles should he be elected
President, were equally determined to vote against his nomination.
In the hope that some compromise might yet be effected, the
Convention, on the motion of Mr. Russell, of Virginia, resolved to
adjourn to meet at Baltimore on Monday, the 18th June; and it was
“respectfully recommended to the Democratic party of the several
States, to make provision for supplying all vacancies in their
respective delegations to this Convention when it shall reassemble.”
The Convention reassembled at Baltimore on the 18th June, 1860,
according to its adjournment, and Mr. Cushing, the President, took
the chair.
Immediately after the reorganization of the Convention, Mr.
Howard, of Tennessee, offered a resolution, “that the President of
this Convention direct the sergeant-at-arms to issue tickets of
admission to the delegates of the Convention, as originally
constituted and organized at Charleston.” Thus the vitally important
question was distinctly presented. It soon, however, became manifest
that no such resolution could prevail. In the absence of the delegates
who had withdrawn at Charleston, the friends of Mr. Douglas
constituted a controlling majority. At the threshold they resisted the
admission of the original delegates, and contended that by
withdrawing they had irrevocably resigned their seats. In support of
this position, they relied upon the language of the resolution
adjourning the Convention to Baltimore, which, as we have seen,
“recommended to the Democratic party of the several States to make
provision for supplying all vacancies in their respective delegations
to this Convention, when it shall reassemble.” On the other hand, the
advocates of their readmission contended that a simple withdrawal
of the delegates was not a final renunciation of their seats, but they
were still entitled to reoccupy them, whenever, in their judgment,
this course would be best calculated to restore the harmony and
promote the success of the Democratic party; that the Convention
had no right to interpose between them and the Democracy of their
respective States; that being directly responsible to this Democracy,
it alone could accept their resignation; that no such resignation had
ever been made, and their authority therefore continued in full force,
and this, too, with the approbation of their constituents.
In the mean time, after the adjournment from Charleston to
Baltimore, the friends of Mr. Douglas, in several of these States, had
proceeded to elect delegates to take the place of those who had
withdrawn from the Convention. Indeed, it was manifest at the time,
and has since been clearly proved by the event, that these delegates
represented but a small minority of the party in their respective
States. These new delegates, nevertheless, appeared and demanded
seats.[7]
After a long and ardent debate, the Convention adopted a
resolution, offered by Mr. Church, of New York, and modified on
motion of Mr. Gilmore, of Pennsylvania, as a substitute for that of
Mr. Howard, to refer “the credentials of all persons claiming seats in
this Convention, made vacant by the secession of delegates at
Charleston, to the Committee on Credentials.” They thus prejudged
the question, by deciding that the seats of these delegates had been
made and were still vacant. The Committee on Credentials had been
originally composed of one delegate from each of the thirty-three
States, but the number was now reduced to twenty-five, in
consequence of the exclusion of eight of its members from the States
of Georgia, Alabama, Mississippi, South Carolina, Texas, Louisiana,
Arkansas, and Florida. The committee, therefore, now stood 16 to 9
in favor of the nomination of Mr. Douglas, instead of 17 to 16 against
it, according to its original organization.
The committee, through their chairman, Mr. Krum, of Missouri,
made their report on the 21st June, and Governor Stevens, of
Oregon, at the same time presented a minority report, signed by
himself and eight other members.
It is unnecessary to give in detail these conflicting reports. It is
sufficient to state that whilst the report of the majority maintained
that the delegates, by withdrawing at Charleston, had resigned their
seats, and these were still vacant; that of the minority, on the
contrary, asserted the right of these delegates to resume their seats in
the Convention, by virtue of their original appointment.
On the next day (June 22), the important decision was made
between the conflicting reports. Mr. Stevens moved to substitute the
minority report for that of the majority, and his motion was rejected
by a vote of 100½ to 150. Of course no vote was given from any of
the excluded States, except one-half vote from each of the parties in
Arkansas.
The resolutions of the majority were then adopted in succession.
Among other motions of similar character, a motion had been made
by a delegate in the majority to reconsider the vote by which the
Convention had adopted the minority report, as a substitute for that
of the majority, and to lay his own motion on the table. This is a
common mode resorted to, according to parliamentary tactics, of
defeating every hope of a reconsideration of the pending question,
and rendering the first decision final.
Mr. Cessna with this view called for a vote on laying the motion to
reconsider on the table. Should this be negatived, then the question
of reconsideration would be open. The President stated the question
to be first “on laying on the table the motion to reconsider the vote by
which the Convention refused to amend the majority report of the
Committee on Credentials by substituting the report of the
minority.” On this question New York, for the first time since the
meeting at Baltimore, voted with the minority and changed it into a
majority. “When New York was called,” says the report of the
proceedings, “and responded thirty-five votes” (in the negative) “the
response was greeted with loud cheers and applause.” The result of
the vote was 113½ to 138½—“so the Convention refused to lay on
the table the motion to reconsider the minority report.” The
Convention then adjourned until evening, on motion of Mr.
Cochrane, of New York, amidst great excitement and confusion.
This vote of New York, appearing to indicate a purpose to
harmonize the party by admitting the original delegates from the
eight absent States, was not altogether unexpected. Although voting
as a unit, it was known that her delegation were greatly divided
among themselves. The exact strength of the minority was
afterwards stated by Mr. Bartlett, one of its members, in the
Breckinridge Convention. He said: “Upon all questions and
especially upon the adoption of the majority report on credentials, in
which we had a long contest, the line was strictly drawn, and there
were thirty on one side and forty on the other.”
The position of New York casting an undivided vote of thirty-five,
with Dean Richmond at their head, had been a controlling power
from the commencement.
Strong expectations were, therefore, now entertained that after the
New York delegation had recorded their vote against a motion which
would have killed the minority report beyond hope of revival, they
would now follow this up by taking the next step in advance and
voting for its reconsideration and adoption. On the evening of the
very same day, however, they reversed their course and voted against
its reconsideration. They were then cheered by the opposite party
from that which had cheered them in the morning. Thus the action of
the Convention in favor of the majority report became final and
conclusive.
Mr. Cessna, of Pennsylvania, at once moved “that the Convention
do now proceed to nominate candidates for President and Vice-
President of the United States.”
Mr. Russell rose and stated, “It has become my duty now, by
direction of a large majority of the delegation from Virginia,
respectfully to inform you and this body, that it is not consistent with
their convictions of duty to participate longer in its deliberations.”
Mr. Lander next stated “that it became his duty, as one of the
delegates from North Carolina, to say that a very large majority of the
delegation from that State were compelled to retire permanently
from this Convention, on account, as he conceived, of the unjust
course that had been pursued toward some of their fellow-citizens of
the South. The South had heretofore relied upon the Northern
Democracy to give them the rights which were justly due them; but
the vote to-day had satisfied the majority of the North Carolina
delegation that these rights were now refused them, and, this being
the case, they could no longer remain in the Convention.”
Then followed in succession the withdrawal of the delegations
from Tennessee, Kentucky, Maryland, California, Oregon, and
Arkansas. The Convention now adjourned at half-past-ten o’clock
until the next morning at ten.
Soon after the assembling of the Convention, the President, Mr.
Cushing, whilst tendering his thanks to its members for their candid
and honorable support in the performance of his duties, stated that
notwithstanding the retirement of the delegations of several of the
States at Charleston, in his solicitude to maintain the harmony and
union of the Democratic party, he had continued in his post of labor.
“To that end and in that sense,” said he, “I had the honor to meet
you, gentlemen, here at Baltimore. But circumstances have since
transpired which compel me to pause. The delegations of a majority
of the States have, either in whole or in part, in one form or another,
ceased to participate in the deliberations of the Convention. * * * In
the present circumstances, I deem it a duty of self-respect, and I
deem it still more a duty to this Convention, as at present
organized, * * * to resign my seat as President of this Convention, in
order to take my place on the floor as a member of the delegation
from Massachusetts. * * * I deem this above all a duty which I owe to
the members of this Convention, as to whom no longer would my
action represent the will of a majority of the Convention.”
Governor Tod, of Ohio, one of the Vice-Presidents, then took the
vacant chair, and was greeted with hearty and long-continued cheers
and applause from members of the Convention.
Mr. Butler, of Massachusetts, now announced that a portion of the
Massachusetts delegation desired to retire, but was interrupted by
cries of “No,” “No,” “Call the roll.” Mr. Cessna called for the original
question, to wit, that the Convention now proceed to a nomination
for President and Vice-President.
The President here ordered the Secretary to call the States. Maine,
New Hampshire, and Vermont were called, and they gave an
unbroken vote for Stephen A. Douglas. When Massachusetts was
called, Mr. Butler rose and said he had a respectful paper in his hand
which he would desire the President to have read. A scene of great
confusion thereupon ensued, cries of “I object” being heard upon all
sides. Mr. Butler, not to be baffled, contended for his right at this
stage to make remarks pertinent to the matter, and cited in his
support the practice of the Conventions at Baltimore in 1848 and
1852, and at Cincinnati in 1856. He finally prevailed, and was
permitted to proceed. He then said he “would now withdraw from
the Convention, upon the ground that there had been a withdrawal,
in whole or in part, of a majority of the States; and further, which
was a matter more personal to himself, he could not sit in a
convention where the African slave trade, which was piracy
according to the laws of his country, was openly advocated.”
Mr. Butler then retired, followed by General Cushing and four
others of the Massachusetts delegation. All of these had voted with
the South and against Douglas.
The balloting now proceeded. Mr. Douglas received 173½ votes;
Mr. Guthrie 9; Mr. Breckinridge 6½; Mr. Bocock and Mr. Seymour
each 1; and Mr. Dickerson and Mr. Wise each half a vote. On the next
and last ballot Mr. Douglas received 181½ votes, eight of those in the
minority having changed their votes in his favor.
To account for this number, it is proper to state that a few
delegates from five of the eight States which had withdrawn still
remained in the Convention. On the last ballot Mr. Douglas received
all of their votes, to wit: 3 of the 15 votes of Virginia, 1 of the 10 votes
of North Carolina, 1½ of the 3 votes of Arkansas, 3 of the 12 votes of
Tennessee, 3 of the 12 votes of Kentucky, and 2½ of the 8 votes of
Maryland, making in the aggregate 14 votes. To this number may be
added the 9 votes of the new delegates from Alabama and the 6 from
Louisiana, which had been admitted to the exclusion of the original
delegates.
Mr. Douglas was accordingly declared to be the regular nominee of
the Democratic party of the Union, upon the motion of Mr. Church,
of New York, when, according to the report of the proceedings, “The
whole body rose to its feet, hats were waved in the air, and many
tossed aloft; shouts, screams, and yells, and every boisterous mode of
expressing approbation and unanimity, were resorted to.”
Senator Fitzpatrick, of Alabama, was then unanimously nominated
as the candidate for Vice-President; and the Convention adjourned
sine die on the 23d June, the sixth and last day of its session. On the
same day, but after the adjournment, Mr. Fitzpatrick declined the
nomination, and it was immediately conferred on Mr. Herschel V.
Johnson, of Georgia, by the Executive Committee. Thus ended the
Douglas Convention.
But another Convention assembled at Baltimore on the same 23d
June, styling itself the “National Democratic Convention.” It was
composed chiefly of the delegates who had just withdrawn from the
Douglas Convention, and the original delegates from Alabama and
Louisiana. One of their first acts was to abrogate the two-third rule,
as had been done by the Douglas Convention. Both acted under the
same necessity, because the preservation of this rule would have
prevented a nomination by either.
Mr. Cushing was elected and took the chair as President. In his
opening address he said: “Gentlemen of the Convention, we
assemble here, delegates to the National Democratic Convention,
duly accredited thereto from more than twenty States of the Union,
for the purpose of nominating candidates of the Democratic party for
the offices of President and Vice-President of the United States, for
the purpose of announcing the principles of the party, and for the
purpose of continuing and re-establishing that party upon the firm
foundations of the Constitution, the Union, and the co-equal rights
of the several States.”
Mr. Avery, of North Carolina, who had reported the majority
resolutions at Charleston, now reported the same from the
committee of this body, and they “were adopted unanimously, amid
great applause.”
The Convention then proceeded to select their candidates. Mr.
Loring, on behalf of the delegates from Massachusetts, who with Mr.
Butler had retired from the Douglas Convention, nominated John C.
Breckinridge, of Kentucky, which Mr. Dent, representing the
Pennsylvania delegation present, “most heartily seconded.” Mr.
Ward, from the Alabama delegation, nominated R. M. T. Hunter, of
Virginia; Mr. Ewing, from that of Tennessee, nominated Mr.
Dickinson, of New York; and Mr. Stevens, from Oregon, nominated
General Joseph Lane. Eventually all these names were withdrawn
except that of Mr. Breckinridge, and he received the nomination by a
unanimous vote. The whole number of votes cast in his favor from
twenty States was 103½.
General Lane was unanimously nominated as the candidate for
Vice-President. Thus terminated the Breckinridge Convention.
The Chicago Republican Convention.
The Republicans had named May 16th, 1860, as the date and
Chicago as the place for holding their second National Convention.
They had been greatly encouraged by the vote for Fremont and
Dayton, and, what had now become apparent as an irreconcilable
division of the Democracy, encouraged them in the belief that they
could elect their candidates. Those of the great West were especially
enthusiastic, and had contributed freely to the erection of an
immense “Wigwam,” capable of holding ten thousand people, at
Chicago. All the Northern States were fully represented, and there
were besides partial delegations from Delaware, Maryland,
Kentucky, Missouri and Virginia, with occasional delegates from
other Slave States, there being none, however, from the Gulf States.
David Wilmot, of Penna., author of the Wilmot proviso, was made
temporary chairman, and George Ashmun, of Mass., permanent
President. No differences were excited by the report of the committee
on platform, and the proceedings throughout were characterized by
great harmony, though there was a somewhat sharp contest for the
Presidential nomination. The prominent candidates were Wm. H.
Seward, of New York; Abraham Lincoln, of Illinois; Salmon P. Chase,
of Ohio; Simon Cameron, of Pennsylvania, and Edward Bates, of
Missouri. There were three ballots, Mr. Lincoln receiving in the last
354 out of 446 votes. Mr. Seward led the vote at the beginning, but
he was strongly opposed by gentlemen in his own State as prominent
as Horace Greeley and Thurlow Weed, and his nomination was
thought to be inexpedient. Lincoln’s successful debate with Douglas
was still fresh in the minds of the delegates, and every addition to his
vote so heightened the enthusiasm that the convention was finally
carried “off its feet,” the delegations rapidly changing on the last
ballot. Lincoln had been a known candidate but a month or two
before, while Seward’s name had been everywhere canvassed, and
where opposed in the Eastern and Middle States, it was mainly

Lexical Facility Size Recognition Speed and Consistency As Dimensions of Second Language Vocabulary Knowledge Michael Harrington Full Chapter

Uploaded by

Copyright:

Available Formats

You might also like

Lexical Facility Size Recognition Speed and Consistency As Dimensions of Second Language Vocabulary Knowledge Michael Harrington Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lexical Facility Size Recognition Speed and Consistency As Dimensions of Second Language Vocabulary Knowledge Michael Harrington Full Chapter

Uploaded by

Copyright:

Available Formats

Lexical Facility : Size, Recognition

Speed and Consistency as Dimensions

not with as youdo at

ISBN 978-1-137-37261-1 ISBN 978-1-137-37262-8 (eBook)

Library of Congress Control Number: 2017946891

© The Editor(s) (if applicable) and The Author(s) 2018

Cover design by Henry Petrides

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

The image on the front cover is a stylized representation of what is known

Part 1 Introduction 1

1 Size as a Dimension of L2 Vocabulary Skill 3

2 Measuring Recognition Vocabulary Size 25

3 L2 Word Recognition Skill and Its Measurement 45

4 Lexical Facility: Bringing Size and Speed Together 67

5 Measuring Lexical Facility: The Timed Yes/No Test 95

Part 2 Introduction 121

6 Lexical Facility as an Index of L2 Proficiency131

7 Lexical Facility and Academic English Proficiency157

8 Lexical Facility and IELTS Performance187

8.4 Findings for Study 3 IELTS Band-Scores 201

9 Lexical Facility and Language Program Placement205

10 Lexical Facility and Academic Performance in English227

11 The Effect of Lexical Facility241

12 The Future of Lexical Facility261

Fig. 1.1 Elements of vocabulary knowledge tapped by vocabulary

Table 1.1 Vocabulary size expressed in word families

Table 6.6 Frequency-level analysis. Comparing sensitivity of hits

Table 8.4 IELTS band-score study. One-way ANOVAs for individual

Table 10.2 Bivariate correlations between lexical facility measures and

Two bedrocks of fluent second language (L2) performance are an ade-

Scope of the Book

a particular interest in the vocabulary size needed for fluent performance

The book is the first to investigate the value of treating vocabulary

and consistency) can be productively considered together as indices of L2

testing outcomes in language schools in Sydney and Singapore in Chap. 9.

1. compare the three measures of lexical facility (vocabulary knowledge,

The book is in two parts. Part 1 presents the theoretical foundation

Part 1 (Chaps. 1, 2, 3, 4, and 5) introduces the theoretical and method-

• Introduce the vocabulary size research literature.

© The Author(s) 2018 3

estimates reflect vocabulary breadth and have been related to L2 perfor-

1.2 Estimating Vocabulary Size

The vocabulary size approach quantifies vocabulary knowledge as a col-

Very high certainty

Word parts, including part of speech and

Collocations: Tendency of two or more words

Fig. 1.1 Elements of vocabulary knowledge tapped by vocabulary size tests

Single words are therefore of primary importance, primary used here

Quantifying Vocabulary Size

Type/Token Ratio A basic distinction can be made between the first

developed for measuring L1 vocabulary development. It is an index of

Lemmas Estimating vocabulary size requires a way to identify the

 he Word Family as a Unit of Recognition

While recognizing these limitations, the word family is nonetheless an

 ord Frequency Statistics as an Index of Vocabulary

in the frequency table. This inverse proportionality means that words

1.3  ocabulary Size as a Dimension

Part 1 Introduction 1

1 Size as a Dimension of L2 Vocabulary Skill 3

2 Measuring Recognition Vocabulary Size 25

3 L2 Word Recognition Skill and Its Measurement 45

4 Lexical Facility: Bringing Size and Speed Together 67

5 Measuring Lexical Facility: The Timed Yes/No Test 95

Part 2 Introduction 121

6 Lexical Facility as an Index of L2 Proficiency131

7 Lexical Facility and Academic English Proficiency157

8 Lexical Facility and IELTS Performance187

8.4 Findings for Study 3 IELTS Band-Scores 201

9 Lexical Facility and Language Program Placement205

10 Lexical Facility and Academic Performance in English227

11 The Effect of Lexical Facility241

12 The Future of Lexical Facility261

Scope of the Book

1.2 Estimating Vocabulary Size

Quantifying Vocabulary Size

he Word Family as a Unit of Recognition

ord Frequency Statistics as an Index of Vocabulary

1.3 ocabulary Size as a Dimension

ocabulary Size as a Dimension of L2

antonymy, homonymy, synonymy, and other relational links such as

ocabulary Learning Assumptions in the Vocabulary

rovide an objective means to scale learner vocabulary size as probabi-

Frequency of Occurrence and Diversity of Context

Frequency of Occurrence and Text Coverage

There are three _ that must be considered when ____

to __ how the most _ of __ coverage relates

to reading __ . The first is the _ of

reading, or Reading a _ is different from reading

for general ___ is different from reading for _. The

third is the way in which is _.

All these _must be taken into consideration when __

to _ how differences in _ coverage might affect ______.