MENTAL IMAGERY AND IDIOM U1\TDERSTANDING IN ADULTS:
EXAMINING DUAL CODING THEORY
by
PEl-FANG HUNG
A DISSERTATION
Presented to the Department of Special Education
and Clinical Sciences
and the Graduate School ofthe University of Oregon
in partial fulfillment ofthe requirements
for the degree of
Doctor ofPhilosophy
June 2010
11
University of Oregon Graduate School
Confirmation of Approval and Acceptance of Dissertation prepared by:
Pei-Fang Hung
Title:
"Mental Imagery and Idiom Understanding in Adults: Examining Dual Coding Theory"
This dissertation has been accepted and approved in partial fulfillment ofthe requirements for
the degree in the Department of Special Education and Clinical Sciences by:
Marilyn Nippold, Chairperson, Special Education and Clinical Sciences
Roland Good, Member, Special Education and Clinical Sciences
Deborah Olson, Member, Special Education and Clinical Sciences
Nathaniel Teich, Outside Member, English
and Richard Linton, Vice President for Research and Graduate Studies/Dean ofthe Graduate
School for the University ofOregon.
June 14,2010
Original approval signatures are on file with the Graduate School and the University of Oregon
Libraries.
? 2010 PEl-FANG HUNG
111
Pei-Fang Hung
An Abstract ofthe Dissertation of
for the degree of
iv
Doctor ofPhilosophy
in the Department ofSpecial Education and Clinical Sciences
to be taken June 2010
Title: MENTAL IMAGERY AND IDIOM UNDERSTANDING IN ADULTS:
EXAMINING DUAL CODING THEORY
Approved:
Dr. Marilyn A. Nippold
This study examined idiom understanding in 120 neurologically healthy adults,
ages 20-29 (20s Group), 40-49 (40s Group), 60-69 (60s Group), and 80-89 (80s Group)
years old. Each participant was administered a familiarity task, definition explanation task,
mental imagery task, and forced-choice comprehension task. Twenty idioms, 10
transparent and 10 opaque, were used with no supporting contexts. Participants were asked
to rate the familiarity ofeach idiom, to provide a definition ofeach, to generate a mental
image ofeach, and to select the best definition ofeach from among four options. It was
predicted that younger and older adults would perform equally well on the comprehension
task but that older adults would perform poorer than younger adults on the explanation
task. Additionally, mental imagery ofidioms was expected to become more figurative with
v
advancing age, and participants were expected to perfOlm better on highly familiar and
transparent idioms than on less familiar and opaque ones.
Participants rated all 20 idioms as highly familiar, with the lowest familiarity
rating for participants in the 20s Group. No significant differences were found on the
forced-choice comprehension task across the four age groups although the 20s Group
scored the lowest among all age groups. The 60s Group performed significantly better
than the 20s Group on the definition explanation task, but no significant differences were
found between the other age groups. No significant differences were found in generating
mental images between transparent and opaque idioms, and mental images tended to be
figurative rather than literal for both types ofidioms.
The present study adds to our knowledge of idiom understanding across
adulthood. Familiarity seemed to playa stronger role than transparency in idiom
understanding in adults. Once an idiom was learned and stored as a lexical unit, people
used the idiomatic meaning and generated figurative mental imagery immediately
without accessing the literal meaning or the literal mental image.
CURRICULUM VITAE
NAME OF AUTHOR: Pei-Fang Hung
PLACE OF BIRTH: Taipei, Taiwan
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon, Eugene, OR, U.S.A.
Chung Shan Medical University, Taichung, Taiwan
DEGREES AWARDED:
Doctor ofPhilosophy in Communication Disorders and Sciences, 2010
University ofOregon.
Master ofScience in Communication Disorders and Sciences, 2004
University ofOregon.
Bachelor of Science in Communication Disorders, 1999
Chung Shan Medical University
AREAS OF SPECIAL INTEREST:
Language Performance in Geriatric Population
Adult Language Disorders and Aphasia
Figurative Language
Research Design and Data Analysis
Cognitive Impairment
Assistive Technology
PROFESSIONAL EXPERIENCE:
Speech-Language Pathologist, Country Villa Health Care, December 2007 to
Present.
Graduate Teaching Fellow, Communication Disorders and Sciences Program,
University of Oregon. March, 2004 to December, 2007.
VI
Vll
Speech Language Pathologist, Rehabilitation Center, Department ofPhysical
Medicine & Rehabilitation, National Taiwan University Hospital, Taipei,
Taiwan. July 1999 to June 2002.
GRANTS, AWARDS AND HONORS:
Clare Wilkins Chamberlin Memorial Research Award, 2009
University of Oregon Graduate School Research Award. April, 2009
PUBLICATIONS:
Sohlberg, M. M., Fickas, S., Lemoncello, R., & Hung, P. (2009) Validation ofthe
Activities ofCommunity Transportation (ACTs) model ofcommunity
navigation for individuals with cognitive impairments. Disability and
Rehabilitation, 31(11),887-897
Fickas, S., Sohlberg, M. M., & Hung, P. (2008). Route-following assistance for travelers
with cognitive impairments: A comparison offour prompt modes. International
Journal ofHuman-Computer Studies, 66(12),876-888.
Sohlberg, M. M., Fickas, S., Hung, P, & Fortier, A. (2007). A comparison offour prompt
modes for route finding for community travelers with severe cognitive
impairments. Brain Injury, 21(5), 531-538.
Sohlberg, M. M., Todis, B., Fickas, S., Hung, P., & Lemoncello, R. (2005). A profile of
community navigation in adults with chronic cognitive impairments. Brain
Injury, 19(14), 1249-1259.
Vlll
ACKNOWLEDGMENTS
I would like to especially thank Dr. Marilyn Nippold, my advisor and the
committee chair. Without her persistent help and guidance, this dissertation would not
have been possible. I am thankful for her mentoring and assistance, so I could utilize
critical~thinkingand problem-solving skills to successfully complete my dissertation. I
am especially grateful and thankful to my committee members, Dr. Roland Good, Dr.
Deborah Olson, and Dr. Nathaniel Teich, for their encouragement and support throughout
the process and the insightful comments and suggestions.
I wish to express my warm and sincere thanks to all the participants ofthis study.
Thank you all for your time and effort. I would also like to thank Kim, Ron, Michelle, and
Pearl for their help and friendship. Special thanks are due to the Culver City Senior Center
and the Department ofCommunicative Disorders ofthe California State University Long
Beach. I appreciate all the staffand faculty for their help.
I wish to thank my family for their constant support and everlasting love throughout
my life. Thank my father and mother for their financial support and my two little brothers
for sharing my feelings.
Finally, I would like to show my gratitude to my partner and soul mate, Lei. I thank
her for believing in me and always being there for me, cheer me up, and help me through
these difficult years. Without her unconditional love and continued support, I would never
accomplish this much.
IX
TABLE OF CONTENTS
Chapter Page
1. BACKGROUND AND INTRODUCTION 1
Purpose ofthe Present Study 10
II. LITERATURE REVIEW 13
Idiom Comprehension 13
Theory in the Idiom Comprehension Process 14
Right Hemisphere or Left Hemisphere? 27
Idiom Comprehension in People with Brain Damage.............................. 32
People with Widespread Brain Damage 33
People with Focal Brain Damage 38
Factors Involved in Idiom Comprehension............................................. 48
Summary of Idiom Comprehension......................................................... 54
Dual Coding Theory 55
Concreteness Effect 56
Dual Coding Theory in Neuroimaging Studies 60
Mental Imagery and Figurative Language 69
Summary ofDual Coding Theory........................................................... 78
Language and Aging 80
Language Production and Aging 81
Language Comprehension and Aging 87
Factors Affect Language Performance in Normal Aging 91
Language Difficulties in People with Dementia...................................... 96
Figurative Language in Normal Aging 101
Summary of Language and Aging 106
Hypotheses ofthe Current Study................................................................. 108
III. METHOD 110
Chapter
Participants .
Recruiting and Sampling .
Participants' Characteristics ..
Materials .
Stimuli ..
Background Questionnaire .
FT' R .
amI Ianty atlng .
Idiom Explanation/Mental Imagery Task ..
Idiom Comprehension Task .
Procedures .
Coding System .
Idiom Comprehension Task .
Idiom Explanation and Mental Imagery Scoring ..
Interrater Reliability .
IV. RESlTLTS .
Age versus Idiom Comprehension ..
Age versus Idiom Explanation .
Age versus Mental Imagery .
Familiarity Rating versus Idiom Comprehension, Explanation,
and Mental Imagery .
Age versus Familiarity Rating ..
Transparent Idioms versus Opaque Idioms ..
Education Years versus Idiom Comprehension, Explanation,
and Mental Imagery .
Relationships among Idiom Familiarity, Idiom Comprehension,
Idiom Explanation, and Mental Imagery .
V. DISCUSSION .
Conclusions .
Age Differences in Idiom Comprehension ..
Dual Coding Theory and Idioms ..
x
Page
110
110
112
114
114
115
115
116
116
117
118
118
118
120
121
121
123
124
126
126
128
129
130
135
136
136
141
Xl
Chapter Page
Factors Involved in Idiom Comprehension 145
Familiarity 145
Transparency..................................................................................... 147
Education 148
Limitations and Future Research 149
Limitations................ 149
Future Research 151
APPENDICES
A. THE LIST OF IDIOMS USED IN THE PRESENT STUDY
(NIPPOLD & DUTHIE, 2003)........................................................................ 153
B. BACKGROUND QUESTIONNAIRE............................................................ 154
C. FAMILIARITY RATING -- IDIOM FAMILIARITY TASK........................ 155
D. IDIOM EXPLANATION TASK 161
E. IDIOM COMPREHENSION TASK............................................................... 170
F. SCORING EXAMPLES OF IDIOM EXPLANATION TASK
REPORTED BY PARTICIPANTS ACROSS FOUR AGE GROUPS........... 175
G. GOOD MENTAL IMAGES REPORTED BY PARTICIPANTS
ACROSS FOUR AGE GROUPS... 178
H. THE ACCURACY OF EACH IDIOM IN IDIOM COMPREHENSION
TASK ACROSS FOUR AGE GROUPS 184
1. THE FREQUENCY OF EACH IDIOM IN IDIOM FAMILIARITY
RATING TASK ACROSS FOUR AGE GROUPS 186
1. THE ACCURACY OF EACH IDIOM IN IDIOM EXPLANATION
AND MENTAL IMAGERY TASK ACROSS FOUR AGE GROUPS......... 189
BIBLIOGRAPHy........... 193
xu
LIST OF FIGURES
Bg~~~
3.1. The Education Distributions for Each Age Group.................................... 114
Xlll
LIST OF TABLES
Table
Page
3.1. The Converting Method for Total Education Years 113
3.2. The Education Distributions for Each Age Group.................................... 113
3.3. Idiom Explanation Scoring System 119
3.4. Mental Imagery Scoring System............................................................... 119
4.1. Performance on Idiom Comprehension Task 122
4.2. One-Way, Between Subjects Analysis ofVariance on Idiom
Comprehension......................................................................................... 122
4.3. The Performance on Idiom Explanation Task 123
4.4. One-Way Between Subjects Analysis ofVariance on Idiom
Explanation Task....................................................................................... 124
4.5. The Performance on Mental Imagery Task............................................... 125
4.6. One-Way Between Subjects Analysis ofVariance on Mental Imagery
Task........................................................................................................... 125
4.7.Pe~rsonCorrelation Coefficients among Familiarity Rating and Idiom
Explanation, Mental Imagery, and Idiom Comprehension Task 126
4.8. The Idiom Familiarity Rating across Four Age Groups 127
4.9. One-Way Between Subjects Analysis ofVariance on Familiarity
Rating Scores............................................................................................ 128
4.10. One-Way, Within-Subjects Analysis ofVariance on Mental Imagery
Task........................................................................................................... 129
4.11. Pearson Correlation Coefficients among Education Years and Idiom
Comprehension, Idiom Explanation, Mental Imagery, and Idiom
Familiarity 130
4.12. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for
20s Group.................................................................................................. 131
4.13. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for
40s Group.................................................................................................. 132
4.14. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for
60s Group.................................................................................................. 133
4.15. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for
80s Group.................................................................................................. 134
1
CHAPTER I
BACKGROU1\1D AND INTRODUCTION
Figurative language uses "figures ofspeech," a way ofsaying something other
than the literal meaning ofthe words, to express people's underlying goals and intentions.
Figures ofspeech includes proverbs (Silence is golden, All roads lead to Rome), idioms
(turn back the clock, skate on thin ice), similes (Jamie runs as fast as the wind) and
metaphors (Life is a rollercoaster, Friends are a shadow). What makes figurative language
special is that communicators need to not only understand the meaning ofthe phrase but
also recognize the implied meaning ofthe phrase. People use different types offigurative
language to achieve different goals. For example, people frequently use irony to guide
another's action or to protect the self, and use idioms to be humorous and to clarify
something (Roberts & Kreuz, 1994). Kempler, Van Lancker, and Read (1988) suggested
that abstract verbal reasoning, concept formation, and quality ofthinking are critical in
figurative language processing since it is commonly used to express abstract concepts in
a concise and concrete manner (Zempleni, Haverkort, Renken, & Stowe, 2007).
Therefore, understanding the abstract and figurative meanings ofwords is an important
way to appreciate everyday speech (Bottini, Corcoran, Sterzi, Paulesu, Schenone, Scarpa,
Frackowiak, & Frith, 1994).
2
Among different types of figurative language, idioms are one <,If the most
common forms of figurative language that are frequent in everyday language and are
prevalent with diverse forms across languages (Bortfeld, 2002). Use ofidioms can be
seen in everyday speech. For example, someone might say "put your foot down" to
encourage you to be decisive without hesitation, or people often say "it is a piece of
cake" to emphasize how easy the job is. Idioms contain conventional meaning and are
easily recognized because oftheir unique linguistic features and ample exposure in daily
life. Although idioms vary on syntactically flexibility, decompositionality, transparency,
and familiarity, given that arbitrary link between idioms and their figurative meanings,
idioms are viewed as frozen phrases that act like mental lexicons or giant lexical units
(Nippold, 2007). Therefore, understanding idioms requires holistic understanding ofnot
only literal but also figurative interpretations (Nippold & Duthie, 2003), which makes
idiom comprehension complicated and not always precise and clear.
Since the underlying meaning offigurative expressions is not understood like
regular phrases, the question how young people acquire figurative competence is worth
further exploration. Kempler and Van Lancker (1993) used The Familiar and Novel
Language Comprehension (FANL-C) protocol to test the acquisition offamiliar language
such as proverbs and idioms on participants, age 3 to 18 years. They found that
participants as young as eight years old comprehended novel sentences, such as "He's
racing a truck against a horse" at adult level; however, not until 12 years of age for
familiar phrases, such as "Rome wasn't built in one day" or "That's enough to drive a
man to drink." Gibbs (1987) proposed that around the ages of eight or nine, children
3
consistently view idiomatic expressions as having figurative meanings. Furthermore,
Kempler and colleagues (Kempler, Van Lancker, Marchman, & Bates, 1999) indicated
that the comprehension of idiomatic and literal meaning followed two different
developmental paths with literal interpretations developing faster and reaching adult level
around age nine or 10 years, but idiomatic interpretations reaching adult levels not until
the age of 15 years old. Young children were able to understand idioms literally during
early childhood (Abkarian, Jones, & West, 1992), but the understanding becomes more
figurative during the school-age and adolescent years with growing exposure to idioms,
language experience, and world knowledge (Nippold & Rudzinski, 1993). Therefore, it
was evident that literal meanings ofidioms were acquired earlier than idiomatic meanings.
Although idiom comprehension was expected to improve throughout the school-age years
and well into the adulthood, even adults did not master all the idioms used in the study
(Nippold & Duthie, 2003).
How idioms are processed is complicated and controversial. Common suggestions
ofidiom comprehension are literal meaning constructed first (Bobrow & Bell, 1973),
idiomatic meaning established first (Gibbs, 1980), or both meanings activated
simultaneously (Swinney & Cutler, 1979). Four hypotheses are generated according to
the common suggestions ofidiom comprehension. The idiom list hypothesis (Bobrow &
Bell, 1973) claims that idioms are listed separately in memory. When people are not able
to generate the meaning based on its grammatical analysis, people then look up the idiom
list to find the correct interpretation. Different from the idiom list hypothesis, the lexical
representation hypothesis (Swinney & Cutler, 1979) suggests that both literal and
4
idiomatic meanings are activated simultaneously, and it views idioms as long words that
are stored and accessed like words. The third assumption, the configurational hypothesis
(Cacciari & Tabossi, 1988), propose that both figurative and literal interpretations are
processed in parallel, but people have to recognize the configuration ofthe idioms first
through the idiomatic key, the point that figurative meaning emerged. Finally, the direct
access hypothesis (Gibbs, 1980) states that people recognize the idiomatic meanings
directly without accessing literal interpretations.
Idioms vary widely in difficulty depending onfamiliarity, a measure ofhow
frequency an expression occurs in the language, transparency, a measure ofthe degree of
correlation between its literal and nonliteral meanings of an idiom (Nippold & Tylor,
2002), and decompositionality, a measure ofthe degree of generating the meaning from
each individual word ofan idiom. Familiar idioms were reported to be better recalled
(Schweigert, 1991), read faster (Cronk & Schweigert, 1992), and comprehended more
easily (Nippold & Taylor, 2002) than less familiar idioms. Schweigert (1986) also
suggested that the familiarity ofthe idiom needed to be taken into account in the models
of idiom processing since it might play an important role in inconsistent results (Cronk &
Schweigert, 1992; Schweigert, 1991). Nippold and Rudzinski (1993) conducted a
developmental study to explore the effect offamiliarity and transparency in idiom
explanation. They pointed out that familiarity and transparency were tightly related to the
wide differences in difficulty ofindividual idioms for both school-age children and
adolescents. In addition, exposure and active analysis contributed to the learning of
5
figurative meanings ofidioms, and opaque idioms were learned in a more holistic way
while analyzing strategies were commonly used in learning transparent idioms.
Because ofthe wide variability among idioms, results from idiom comprehension
studies were often inconclusive due to the types ofidioms used in the study, different
measurements utilized in idiom comprehension, the amount and the type ofcontext that
the idioms embedded in, and the focus ofstudy tapping on different levels ofprocess.
Therefore, how to access the idiom comprehension process through a deep understanding
method and how to establish the behavioral evidence of developing the figurative
meaning ofan idiom are challenging. Gibbs and O'Brien (1990) proposed that the active
role ofconceptual metaphors strengthened the link between idioms and their nonliteral
meanings in idiom comprehension. Additionally, researchers believed that conceptual
metaphors facilitated what the proverbs or idioms really meant according to the high
consistency ofpeople's mental imagery (Bortfeld, 2002; Gibbs, Strom, &
Spivey-Knowlton, 1997). Gibbs and O'Brien (1990) pointed out that the best way to
reveal tacit knowledge of idioms is through a detailed examination ofpeople's mental
images ofthe expressions. Nippold and Duthie (2003) also suggested that mental imagery
may serve as an indicator of deep understanding of figurative meanings in idiom
comprehension. Therefore, exploring mental images may be an excellent way to
understand the underlying mechanism in the idiom comprehension process.
According to the dual coding theory (Pavio, 1971, 1986), there are two
independent but interconnected systems, the verbal and imaginal system, attributed to the
different retrieval effects between concrete and abstract words. A concrete concept is
6
better recalled and retained than an abstract concept because the concrete concept could
be encoded both verbally and nonverbally, which results in the concreteness effect.
Several behavioral studies (Sadoski, 1995; Sadoski, Goetz, & Fritz, 1993; Sadoski, Goetz,
& Rodriguez, 2000; Sabsevitz, Meddler, Seidenberg, & Binder, 2005) and neuroimaging
studies (Jessen, Heun, Erb, Granath, Klose, Papassotiropoulos, & Grodd, 2000; Kounios
& Holcomb, 1994; Nittono, Suehiro, & Hori, 2002; Sabsevitz, Meddler, Seidenberg, &
Binder, 2005; Swaab, Baynes, & Knight, 2002; West & Holcomb, 2000) showed that
advantage existed in processing concrete words and concrete information in comparison
to abstract information. In addition, the processing of abstract concept is left-Iateralized
and the processing ofconcrete concepts is bilateral. The dual coding theory suggests that
the additional creation ofmental imagery helps people comprehend and remember the
concrete information; therefore, the activation ofthe imaginal system facilitates concrete
language and information processing in the brain. Since the transparency ofthe idiom
was one ofthe main factors in idiom comprehension and a key variable in the present
study, it was important to explore the role ofmental imagery in comprehending
transparent and opaque idioms. Therefore, this study intended to examine ifthe dual
coding theory is applicable to idiom comprehension and ifthe concreteness effect exists
in the process oftransparent idioms. In addition, examining the dual coding theory in
idiom comprehension may add knowledge in understanding the relationship between
mental imagery and figurative language.
Different from developmental studies in children and clinical studies in adults
with language and cognitive impairments, the present study focused on idiom
7
comprehension in adults across the lifespan. Language performance in normal aging
populations is rarely studied and has received less attention in comparison to adults with
dementia, Alzheimer's disease, or cognitive impairments. However, some studies showed
that both language production and comprehension declined with advancing age due to
decrement in processing speed, working memory, and inhibitory efficiency (Burke, 1997;
Carpenter, Miyake, & Just, 1994; Gunter, Jackson, & Mulder, 1995; KwongSee & Ryan,
1995; Waters & Caplan, 2005). In addition, difficulties in naming, reduced syntactic
complexity, and decrement in verbal fluency are commonly observed during the age of70
years old (Ardila & Rosselli, 1996), and word finding/lexical retrieval difficulties, tip of
the tongue (TOT), disfluency in speech, and ambiguous references were reported to be
common deterioration observed in elder's language production (Burke & MacKay, 1997).
Several studies showed that older adults tended to produce less complex sentences
compared to younger people (Kemper, Greiner, Marquis, Prenovost, & Mitzner, 2001;
Kemper, Herman, & Lian, 2003; Kemper, Herman, & Liu, 2004) and seniors tended to
use simpler grammatical forms to avoid imposing high memory demand (Kynette &
Kemper, 1986)
Language comprehension, similar to language production, also showed declines
with advancing age. Cohen (1979) reported that elders maintained surface comprehension
but had deficits in language comprehension involving integration, inference, and
construction. Additionally, normal aging affected comprehension of complex sentences,
and the decline began after 60 years of age (Davis & Ball, 1989; Feier & Gerstman, 1980;
Obler et aI., 1991). Although the difficulties revealed in each study depending on the
8
tasks that researchers used to assess the performance (Glosser & Deser, 1992), most of
complex linguistic abilities were affected by deterioration in attention mechanism ofthe
working memory system and an increasing lack of inhibition capacity, a disturbance in
the cerebral activity which activated multiple cortex areas during the complex tasks
(Juncos-Rabadan & Iglesias, 1994; Kemper, Herman, & Liu, 2004). Generally speaking,
in language production and comprehension, the greater the linguistic complexity, the
more difficulty in language processing was found in normal older people (Emery, 1986).
However, education (Ardila & Rosselli, 1996; Harris, Rogers, & Qualls, 1998;
luncos-Rabadan & Iglesias, 1994; luncos-Rabadan, 1996; Mackenzie, 2000), gender
(Ardila & Rosselli, 1996), verbal ability (Harris et aI., 1998; Petros, Norgaard, Olson, &
Tabor, 1989), and other factors, such a the influence oftext genre (Harris et aI., 1998;
Petros et aI., 1989), should be taken into account while studying language performance in
normal aging populations, since these variables were reported to be strongly correlated to
the language performance in older people.
Studies on how figurative language was processed in normal older people showed
inconclusive results. Obler and Albert (1989) found no changes in metalinguistic tasks
such as interpretation ofproverbs and idioms in normal older people; however, proverb
explanation showed slightly decline over 60 years of age and the decline reached
significance around age of 70 in the Nippold, Uhden, and Schwarz study (1997).
Metaphor interpretation performance was also found to be well maintained through the
later adult years (Szuchman & Erber, 1990) although plausible executive function
impairment was found to be related to the proverb comprehension deficits in normal
9
elderly in the study ofUekeermann, Thoma, and Daum (2008). Examining language
abilities in normal aging populations was challenging because of different variables, such
as education and verbal ability existing in individuals, different tasks, and types of
figurative language tapping different cognitive and linguistic functions (Qualls & Harris,
2003). However, figurative language provided a functional and natural way to examine
language performance, which may be more sensitive and reliable than standardized tests.
Figurative language is substantial in all language and is essential to everyday
communication (Qualls & Harris, 2003); however, it is rarely studied and has received
less attention in comparison with other aspects oflanguage. Given that figurative
language is less studied and rarely examined in normal aging populations, it is necessary
to have ground knowledge ofidiom processing in adults across the lifespan. Studies
showed that language performance, both production and comprehension, declined with
advancing age due to decrement in cognition and linguistic processing (KwongSee &
Ryan, 1995; Waters & Caplan, 2005). However, with the growing population with
dementia and Alzheimer's disease, the understanding of language performance in the
normal aging population is relatively limited. In addition, standardized language tests
may not be sensitive enough to distinguish language decline that results from normal
aging or mild cognitive impairment, such as the early-stage dementia. Therefore, the
purpose ofconducting the current study was to explore idiom understanding in four age
groups of adults by examining their explanation ofidioms, the mental images they had
while encountering idioms, and their comprehension ofidioms using a multiple-choice
task. This study was expected to provide normative data on idiom comprehension in
10
neurologically normal adults and elderly, which would be valuable and useful for
clinicians to evaluate possible language deterioration with a more functional and practical
approach. Additionally, both explanation and comprehension were examined in the
present study to obtain a comprehensive understanding ofhow idioms were processed in
adults across the lifespan.
Purpose ofthe Present Study
Given the need for research evidence in idiom comprehension in adults across the
lifespan, the present study was designed to extend the study ofNippold and Duthie
(2003). In their study, they examined the use ofmental imagery in idiom comprehension
in school-age children and 40 neurologically nonnal adults. Similar to their study, the
present study also examined the relationship between idiom comprehension and mental
imagery in understanding oftransparent and opaque idioms. The results ofthe Nippold
and Duthie study (2003) showed that transparent idioms were easier understood than
opaque ones, and mental imagery was associated with the comprehension ofboth
transparent and opaque idioms for both school-age children and adults. One recent study
by Qualls and Harris (2003) examined idiom comprehension in normal elders. They
examined the comprehension ofdifferent types of figurative language, including idioms,
metaphors, and metonyms, with the possible effect ofage, working memory, and reading
ability. No age-related decline was found in figurative language comprehension in older
adults in their study when a selection ofresponse choices was provided. In addition, older
adults performed better on idioms than metaphors and metonyms with the greatest
11
difficulty in metonyms. However, Qualls and Harris (2003) only used a force-choice task
to test idiom comprehension and did not examine the effect offamiliarity and
transparency in idiom comprehension.
This study intended to include the effect offamiliarity and transparency that are
essential in idiom comprehension process and to examine idiom comprehension through
different types oftasks -- explanations, mental imagery, and multiple choices -- in order
to provide a more comprehensive understanding ofidiom comprehension in the normal
aging populations. Figurative language, especially idioms, was rarely studied in
neurologically normal elderly. The results ofproverb understanding in aging populations
showed that proverb explanation reached peak around the age 20s and remained stable
during the age 30s, 40s, and 50s, but slightly declined in the age 60s and the decline
reached significant around the age 70s O'Jippold, Uhden, & Schwarz, 1997). In addition,
the Uekeermann, Thoma, and Daum study (2008) indicated that older people had
impaired proverb comprehension and tended to choose literal meanings rather than
figurative ones. Therefore, whether idiom comprehension declined like proverb
comprehension with advancing age or it remained intact as suggested by Qualls and
Harris (2003) was worth further study.
The purposes and rationales ofthe present study can be summarized as follows:
1. Figurative language is substantial in all language and is essential to everyday
communication. Among different types offigurative language, idioms are one the
most common forms in everyday language and are prevalent with diverse forms
across languages.
12
2. Figurative language is less studied in comparison with other aspects of language
in children and is even rarely examined in normal aging populations.
3. Studies showed that the best way to reveal tacit knowledge was through a detailed
examination ofpeople's mental images ofidioms; therefore, exploring mental
images in idiom comprehension is essential for researchers to understand the
underlying mechanisms.
4. Mental imagery in figurative language development is expected to become more
figurative with advancing age; however, no studies up to date examined the
mental imagery generated during idiom comprehension in normal elderly
populations.
5. Findings in language performance ofnormal older people are inconclusive in both
language production and language comprehension tasks. Since different tasks
yield different results, the explanation task and the comprehension task would be
used in this study to provide a better understanding ofidiom comprehension in the
normal aging population.
6. Understanding language ability in normal aging is essential to distinguish
language decline in normal aging, dementia, or other age-related cognitive
declines.
7. Figurative language provides a realistic and practical way to evaluate the change
ofeveryday speech, which provides useful, functional, and natural means for
clinicians to evaluate language performance in the elderly.
13
CHAPTER II
LITERATURE REVIEW
The main focus ofthis study was to examine the applicability ofthe dual coding
theory in idiom comprehension in adults across the lifespan. Therefore, literature in 1)
idiom comprehension, 2) the dual coding theory, and 3) language in aging populations
was reviewed to provide comprehensive and solid background knowledge ofthis study.
Idiom Comprehension
Idioms are one ofthe most common forms of figurative language that are frequent
in everyday language and are prevalent with diverse forms across languages (Bortfeld,
2002). Idioms are complex and difficult to define; however, some characteristics are
commonly seen and recognized in idioms. Four features in idiom processing were
proposed by Oliveri, Romero, and Papagno (2004). First, the use of idiom is conventional.
Secondly, idioms comprehension varies based on its transparency, and some ofthe idioms
involve figuration. Thirdly, some ofthe idioms can be interpreted though their parts, but
some ofthem are not decomposable and are stored as a complete phrase. The last feature
is its syntactic frozenness. Gibbs (1987) pointed out that some idioms are syntactically
14
frozen, but some are syntactically flexible. Therefore, some idioms do not retain their
figurative meaning ifthey are transformed into the passive, but some idioms retain
figurative meaning even after transformed into the passive. However, none ofthese
features applies to all idioms, which makes understanding ofidiom comprehension even
more complicated. According to Nunberg, Sag, and Wasow (1994), idioms are
conventionalized and appear only in a limited number ofsyntactic frames. Idioms
typically involve figuration and are used to describe and implicitly explain a recurrent
situation ofparticular social interest or to imply an affective standpoint. Thus, several
variables, such as decompositionality, familiarity, syntactical flexibility, transparency, and
use ofcontext, must be considered when examining idiom process.
Theories in the Idiom Comprehension Process
The process ofidiom comprehension seems like muddy water in which
researchers hold different views and hypotheses regarding how people process the literal
and figurative meanings ofidioms. Many idioms have both literal and figurative
interpretations; therefore, how these two interpretative processes relate to each other
raises researchers' interests (Needham, 1992). Bowbrow and Bell (1973) proposed the
Idiom List Hypothesis that idiomatic meanings were acquired by combining words into a
complex "idiom word" and searching through a mental "idiom word dictionary" to find
the meaning ofthe idiom. In their study, participants were presented a set ofsentences
containing either four literal or four idiomatic sentences followed by one idiomatic
sentence serving as the test sentence. Participants were asked to report the meaning they
perceived first. The results showed that the number offirst perceived idiomatic meaning
15
reported by participants increased when the preceding sentences were idiomatic sentences
and that a similar increase was found for the number offirst perceived literal meaning
after participants saw a set ofliteral sentences. Therefore, they suggested that distinct
idiomatic and literal processing modes existed and that idioms are listed separately in
memory. When people's attempts to generate the idiomatic meaning from its grammatical
analysis failed, people then looked up the idiom list for the correct interpretation
(Bowbrow & Bell, 1973).
However, the Idiom List Hypothesis was questioned by Swinney and Cutler
(1979). In contrast to the Idiom List Hypothesis, the Lexical Representation Hypothesis,
proposed by Swinney and Cutler (1979), proposed that both literal and idiomatic
meanings were activated simultaneously and idioms were viewed as long words that were
stored and accessed like lexicon. Swinney and Cutler (1979) asked 20 undergraduates to
make an on-line judgment ofwhether the phrase was meaningful on 23 grammatical
idiomatic phrases, 23 matched grammatical controls with one word replaced from the
idiomatic phrases, 30 non-idiomatic and grammatical phrases, and 76 ungrammatical
word strings. It was found that idiomatic phrases were recognized as acceptable English
phrases significantly faster than grammatical controls. Therefore, they claimed that
idioms were stored and accessed as lexical items, neither from the idiom list nor by
special processing mode.
Sixty college students participated in Estill and Kemper's study (1982) in order to
examine the immediate comprehension process during the interpretation ofidioms.
Twenty-four sentence sets designed around idioms were used and each contained four
16
sentences: one sentence used the idiom figuratively, one sentence used the idiom literally,
one sentence was ambiguous between the literal and figurative meanings, and one
sentence was constructed with the final word ofthe idiom with non-idiomatic expression.
Three types ofcues, including identity cues, rhyme cues, and category cues, were defined
and announced. Each participant was asked to press the button as fast as possible when
hearing the word in the sentences that corresponded to the cues. It was revealed that
participants responded more rapidly to the target words in idiomatic sentences than
non-idiomatic control sentences. It showed that idioms were automatically processed as
lexical items. In addition, this study supported that literal and figurative meanings were
processed simultaneously because there was no reaction time advantage for either
condition.
In order to test the idiom list hypothesis and the lexical representation hypothesis,
18 undergraduates participated in Glass's first study (1983). Thirty idioms were used to
construct 120 pairs, including 30 literal paraphrases followed by their idioms, 30 literal
paraphrases followed by irrelevant idioms, 30 figurative paraphrases followed by their
idioms, and 30 figurative paraphrases followed by irrelevant idioms. Participants were
asked to press a button ifthey understood the paraphrases and ifthe idioms had the same
literal and figurative meanings as provided by preceding paraphrases. The results
supported the lexical representation hypothesis, not the idiom list hypothesis. In order to
clarify the confusion between literal and figurative paraphrases for interpretation, each
paraphrase was labeled as either literal or figurative in the second experiment. It was
revealed that literal and figurative interpretations were both accessed during idiom
17
comprehension. The overall findings indicated that the comprehension time for idioms
was faster than for paraphrases and that the literal and figurative meaning ofthe idiom
were always generated during idiom comprehension. Therefore, this study suggested that
the figurative interpretation ofthe idiom was always retrieved along with the literal
interpretation when people encounter an idiom.
Another study, conducted by Tompkins, Boada, and McGarry (1992), also
supported that literal and figurative meanings ofthe idioms were activated
simultaneously in people with brain damage. Twenty adults with
right-hemisphere-damaged (RHD), 20 with left-hemisphere-damaged (LHD), and 20
neurologically normal controls performed an on-line word-monitoring task and an
off-line definition task. In the on-line word-monitoring task, participants listened to a
spoken sentence and pressed the button when the specified target word was presented.
Response time was recorded in three experimental conditions, including idiomatic, literal,
and control. In the off-line definition task, 12 highly familiar idioms, six from the
word-monitoring task, were presented auditorily and visually. The findings from the
online task indicated that all three groups responded more quickly on idiomatic phrases
than on the same nouns in control context, and the idiomatic phrases were processed
automatically regardless the context. The similar reaction time among three groups
suggested that the initial activation and retrieval offamiliar idioms remained intact
regardless ofthe site ofcerebral damage. Moreover, since the reaction time for literal and
idiomatic context was equally fast, it was proposed that literal and figurative meanings of
the idioms were activated simultaneously.
18
Different from the previous two hypotheses, Gibbs (1980) asserted that the literal
interpretation was not necessary for idiom comprehension; therefore, people would not
automatically compute the literal meaning ofidioms. The Direct Access
Hypothesis/Idiom Processing Hypothesis, proposed by Gibbs, stated that people
recognized the idiomatic meanings directly without accessing literal interpretations. Four
types ofexperimental stimuli were used in Gibbs' study (1980), including with or without
context and literal or idiomatic context. In the with-context condition, each story ended in
a target sentence followed by a paraphrase that was either literal or idiomatic. In the
no-context condition, only the target sentence and one type ofparaphrases were presented.
In addition to eight stories with literal target sentences and eight stories with idiomatic
target sentences, another 12 filler stories with false paraphrases were included.
Participants were asked to make a judgment regarding paraphrases, and the time between
onset ofthe display and the response button being pressed was recorded. It was found
that it took less time for participants to process idiomatic interpretations than to process
literal ones.
In the second experiment of Gibbs' study, participants were asked to listen to the
same stories used in the first study and then to come back after 24 hours to write down
what they remembered, especially the last line of each story. The findings showed that
familiar and literal interpretations were better recalled. Therefore, in regular conversation,
people remember literal and unconventional meanings better than idiomatic conventional
meanings. It was suggested that the conventional meaning ofthe sentence was first
analyzed and rejected, which resulted in longer time to process unconventional meanings
19
of idioms. In order to clarify ifpeople needed longer time to process unconventional
interpretations, the third experiment was conducted. The only difference between second
and third study was two types ofrecall prompts were provided, including literal and
idiomatic prompts. In addition, 40 sentences, including 16 idioms, were presented in a
no-context condition. It was revealed that participants recalled better on literal
interpretations than on idiomatic interpretations under either literal or idiomatic prompts
in the with-context condition. Therefore, this study pointed out that whether the sentence
was stated literally or idiomatically, the most important factor in language processing was
the conventionality ofthe sentence. It was suggested that people would automatically
analyze the conventional meaning ofthe sentence before deciding ifthe unconventional
meaning was necessary to be processed.
In order to provide more evidence for the theory, 18 stories supporting the literal
interpretations and 18 stories supporting idiomatic interpretations were used in another
Gibbs' study (1986). The final sentence ofeach story served as a prime sentence, and
each story was followed by three target sentences, literal interpretation, idiomatic
interpretation, or unrelated sentence. Nine stories that did not have either idiomatic or
literal meanings and were followed by non-meaningful sentences were served as fillers.
Participants were instructed to press a button ifthey understood each sentence. After each
story, a target string appeared on the screen. Participants were asked to press different
buttons ifthe sentence was meaningful or not meaningful to them. It was revealed that
participants responded faster on idioms than on literal interpretations. In the second
experiment, the same stories and target sentences were used as materials, but the prime
20
sentences used in the second study only repeated major content words from the story
without idiomatic expressions. It was found that nonliteral target sentences were strongly
facilitated when participants read stories with the idiomatic interpretation. Different from
the first experiment, literal target sentences were facilitated when the preceding prime
sentences were literal. Based on these two experiments, it was concluded that the literal
interpretation was not the default to understanding idioms since there was no difference
in reaction time in judging paraphrases ofthe literal meanings and those ofumelated
meanmgs.
Schweigert and Moates (1988) considered the influence of context and familiarity
and further examined idiom comprehension by presenting familiar idioms in either literal
or idiomatic sentences preceded by a short paragraph. Twenty idioms were written into
40 short stories ending with one idiom target sentence, either representing the literal or
the idiomatical meaning. Twenty short stories ending with a control sentence without an
idiomatic meaning and 40 sentences containing ambiguous words serving as fillers were
also included. Fifty-seven undergraduates first were presented with sentences only and
then read a short paragraph before being presented with the ending sentences. The
measurement was the number ofpresentations that the participant needed to read each
sentence correctly. Participants completed a cued recall task in which participants had to
fill the missing words for the sentences presented 24 hours earlier in the experiment. It
was found that idiom-literal sentences required more presentations than both
idiom-figurative and control sentences. However, idioms presented literally were better
recalled than idioms presented figuratively. The findings supported the Direct Access
21
Model that figurative meaning was retrieved and processed first and the literal meaning
was processed only ifthe figurative meaning was inappropriate. Schweigert and Moates
(1988) suggested that familiar idioms were more likely processed as lexical units than
less familiar idioms. Therefore, familiar idioms were perceived as wholes instead of
individual words. According to the direct access model, idioms used literally were first
processed figuratively and then literally; however, idioms presented figuratively were
only processed figuratively. Therefore, the recall advantage observed in literally
presented idioms might result from extra processing time than idioms used figuratively.
Sixty-three undergraduates were recruited in the Needham study (1992), and
stimuli were 18 experimental passages. Each passage contained a title, a passage with
five full sentences ofthe figurative meanings ofthe idiomatic phrase, and one final
partial sentence with three different versions, including idiom, anaphor, and control
conditions. Three partial sentences were identical except the final noun ofthe target
phrase. For example, she spilled the beans, she spilled the carrots, and she spilled the
beer were created for the title "Carol lets out a secret". Eighteen experimental passages,
32 filler passages, and target words were presented to participants, and they had to decide
whether or not the test word had appeared in the passage. The findings showed that
facilitation was found in the anaphor condition, which had shortest response time, but the
referents ofpotential anaphor in the idiom phrase were not activated. In the second
experiment, materials and procedures were identical to the first experiment except the
verbs in the target phrases were replaced. For example, she dropped the beans. Although
the results also showed that facilitation was found in the anaphor condition, different
22
from the first experiment, the response time for the idiom condition was similar to the
time for the anaphor condition. Therefore, the finding ofthe study supported Gibbs'
Direct Access Hypothesis/Idiom Processing Hypothesis. Needham (1992) claimed that
people did not compute full literal interpretation ofidiomatic expressions when an
idiomatic phrase was interpreted figuratively. Ifthe figurative meaning was recognized at
the early stage, then the processing ofthe literal meaning might be terminated at a very
early stage.
The observed reaction time advantage for idioms found in Gibbs' studies (1980,
1986) was questioned by not measuring on-line processing, not reflecting how people
process idioms during comprehension, and plausible reflecting integration process
without testing the retrieval ofmeaning (Cacciari & Tabossi, 1988; Estill & Kemper,
1982; Needham, 1992). The Configurational Hypothesis, proposed by Cacciari and
Tabossi (1988), argued that both figurative and literal interpretations were processed in
parallel, but people had to recognize the configuration ofthe idioms first through the key
point. In the first experiment conducted by Cacciari and Tabossi (1988), participants were
visually presented with 60 sentences, nine idioms, and 51 filler sentences. Each sentence
was followed by three target words, one semantically associated with the meaning ofthe
idiom (idiom target), one associated with the meaning ofthe last word in the phase
(literal target), and the last one was an unrelated word serving as the control. Participants
had to judge ifthe string showed on the screen was an actual word while listening to a
sentence. Although the results showed that idiom targets were responded to faster than
literal targets and controls, it was questionable due to measuring post-perceptual events
23
instead ofthe online idiom comprehension process. There was possible awareness of
experimental purposes by participants, and idioms used in the experiment could have
unequally induced literal and figurative interpretations. Therefore, after the experimental
biases were controlled, participants were presented 12 idiom sentences, 12 informal
sentences, and 48 formal sentences that each were paired with three target words. It was
found that the figurative meaning was not automatically activated when the sentences
were not biased towards idiomatic interpretations. Different from the first experiment,
idiomatic sentences were first processed only literally. It appeared that sometime after the
last word in the sentence, the idiomatic interpretation might be detected. The results
clearly showed that the interpretation ofidiomatic expressions was a slow process and
required some time to be activated. Therefore, Cacciari and Tabossi (1988) asserted that
for non-predictable idioms, only the literal meaning was accessed first and then the
recognition ofthe idiomatic configuration would take place after the key word has been
accessed. In other words, processing ofan idiomatic string was literal until sometime
after the configuration emerged. Therefore, the Configurational Hypothesis was not
consistent with either the lexical representation hypothesis or the direct access
hypothesis.
The Graded Salience Hypothesis (GSH), suggested by Giora (1997), claimed that
comprehension offigurative and literal language should be governed by a more general
principle of salience depending on which salient meanings were processed first. Ifthe
literal and figurative meanings ofthe conventional metaphor were equally salient, both
meanings would be processed in parallel. Giora (1997) stated that the salience ofa word
24
or an utterance was a function ofits conventionality, familiarity, and given context.
According to the GSH, the idiomatic meaning ofthe familiar idiom should be activated in
both idiomatically-biased context and literally-biased context since idiomatic meanings
ofthe familiar idiom was more salient than its literal meaning. However, for the less
familiar idioms, literal meanings should only be activated in the literally-biased context
(Giora, 1999). Therefore, salient meanings have the privilege that they were always
activated and accessed initially regardless ofthe type or length ofcontext.
In one ofthe experiments of Giora and Fein (1999), the hypothesis that familiarity
might affect the activation ofliteral and idiomatic meanings ofthe idiom was examined.
Sixty primary school students, aged 12-13 years, participated in the study. Twelve
familiar and twelve less familiar idioms were constructed in either literally-biased
sentences or idiomatically-biased sentences. Participants were instructed to read the
sentence and then complete one fragmented word that first came to mind. It was revealed
that for the familiar idioms within the idiomatically-biased context, the salient idiomatic
meaning was highly activated while the less salient literal meaning was hardly activated.
On the other hand, both idiomatic and literal meanings were activated for familiar idioms
in literally-biased context. For the less familiar idioms, both meanings were evoked in the
idiomatically-biased context; however, the more salient literal meaning was highly
activated, the less salient idiomatic meaning was activated in the literally-biased context.
The GSH focused on not only time courses ofthe idioms processing that were
addressed in the lexical representation and configuration model but also the brain regions
that were activated during the process. According to the GSH, salient meanings were
25
easier to access than less salient ones regardless of literal or nonliteral interpretations.
Therefore, the degree ofmeaning salience decided which meaning was retrieved. While
applying to idiom comprehension, the GSH argued that the right hemisphere was
activated during the comprehension ofnonsalient (literal) meaning, and the
comprehension of salient (idiomatic) meaning was processed by the left hemisphere. In
idiom comprehension, the idiomatic meaning was salient because it can not be
decomposed but restored as a whole like mental lexicon; however, its literal meaning was
compositional.
In order to investigate the role ofthe right hemisphere in understanding
idiomatic sentences, Mashal and colleagues (Mashal, Faust, Hendler, & lung-Beeman,
2008) conducted a behavioral divided visual field study using functional Magnetic
Resonance Imaging (fMRI). Seventy-six ambiguous idioms, 38 followed by the target
words related to the idiomatic meaning ofthe idiom, and 38 followed by the target words
related to the literal meaning ofthe idiom, were presented to 32 college students in either
the right visual field/left hemisphere (RVF/LH) or the left visual field/right hemisphere
(LVF/RH). Additional 60 filler sentences followed by nonword targets were also
presented. The results revealed that the reaction time was faster when target words related
to the literal meaning ofthe idioms were presented in LVFIRH than in RVF/LH.
Therefore, the right hemisphere was more sensitive than the left hemisphere on
nonsalient, literal meanings ofambiguous idioms. In order to determine which regions of
the right hemisphere were activated during the process ofnonsalient meaning of
ambiguous idioms, another fMRI study was conducted on 14 young adults, ages 21-31
26
years. Stimuli were 50 ambiguous idioms and 25 literal sentences differing from idioms
in one or two words. For example, kick the bucket versus lick the jar. The findings
indicated that neural activities increased in the right-Iateralized brain regions when literal
interpretations of idioms were processed. In sum, this study supported that literal,
nonsalient interpretations ofidioms were processed in the right hemisphere while the
increased activation in the left hemisphere was observed during processing ofidioms.
The results ofthe behavioral tasks and fMRI studies were in agreement with the Graded
Salience Hypothesis (GSH) that the right hemisphere was sensitive to nonsalient
interpretations of idioms.
Laurent and colleagues (Laurent, Denhieres, Passerieux, Iakimova, &
Hardy-Bayle, 2006) attempted to examine Giora's graded salience hypothesis through the
event-related potentials (ERPs). Thirty adults, age between 21 to 50 years old, were
recruited, and 240 experimental sentences followed by one target word either
semantically related or not were used as stimuli. Six different stimulus categories were
created, including 20 strongly salient idioms with related target words, 20 weakly salient
idioms with related target words, 40 idiomatic fillers with related target words, 40
idiomatic fillers with non-related target words, 40 literal fillers with related target words,
and 40 literal fillers with non-related target words. Participants were instructed to decide
ifthe target word was semantically related to the meaning ofthe utterance. The findings
revealed that the shortest reaction time was found in response to figurative targets that
followed highly salient idioms and in response to literal targets preceded by weakly
salient idioms. Therefore, the results supported Giora's Graded Salience Hypothesis that
27
more salient meanings were accessed faster as a result ofconventionality, frequency,
familiarity, or prototypicality and reached a sufficient level of activation more easily
compared to less salient meanings.
Different theories in idiom processing were reviewed in order to provide a
foundation ofunderstanding how people comprehend idioms. In addition to these theories
and hypotheses, how idioms were processed in the brain also raised researchers' interests.
Literature in idiom processing was reviewed to provide further understanding in idiom
processing.
Right Hemisphere or Left Hemisphere?
Language is commonly assumed to be processed in the left hemisphere with
production in Broca's Area and comprehension in Wernicke's Area. However, growing
evidence indicated that both hemispheres contribute to the comprehension of
semantically related words (Chiarello, Burgess, Richards, & Pollock, 1990), indirect
semantic priming (Kiefer, Weisbrod, Kern, Maier, & Spitzer, 1998 ), semantic priming
(Belanger & Cimino, 2002; Mcdonald, Bauer, Filoteo, Grande, Roper, Buchanan, &
Gilmore, 2005; Richards & Chiarello, 1995), the understanding the meanings of
ambiguous words (Atchley, Burgess, Audet, & Arambel, 1996; Burgess & Simpson,
1988), the understanding ofindirect speech (Wapner, Hamby, & Gardner, 1981), the
comprehension ofjoke (Coulson & Williams, 2005), and making inferences (Beeman,
1993). The right hemisphere was assumed to playa critical role in figurative language
comprehension, such as proverbs and idioms. Burgess and Chiarello (1996) reviewed
studies related to cerebral asymmetries in metaphor and idiom comprehension and further
28
concluded that the right hemisphere played an important role in inferencing and in
understanding the figurative language. They claimed that the right hemisphere was
sensitive to the pragmatic information and activated a broad range ofsemantic
information, which was essential for comprehending figurative language. Likewise, the
left hemisphere was also important in figurative language comprehension, especially
those idioms that involved in syntactic analysis and with preceding context. Since
figurative language contains literal and figurative expressions and requires both semantic
and syntactic knowledge, how figurative language is processed in the brain is worth
further studied.
Van Lancker and Kempler (1987) examined the comprehension of familiar
phrases, such as idioms, and novel sentences on 28 people with left hemisphere damage
(LHD), mean age of62.3 years, and 11 people with right hemisphere damage (RHD),
mean age of63.4 years. Stimuli were 10 concrete nouns, 10 familiar phrases, and 10
novel sentences with matching lengths and syntactic structures. Each phrase contained
four line drawings. The foils were semantically-related words for nouns and variation of
grammatical roles for novel sentences. Foils for the familiar phrases were one literal
expression, one related or opposite expression, and one irrelevant expression. The results
showed that the comprehension offamiliar phrases was less impaired than syntactic
abilities in participants with LHD. Participants with LHD performed better than those
with RHD in familiar phrases but not words and novel sentences. The results supported
that familiar phrases were stored and processed differently from the newly generated
language, and the right hemisphere was specialized in comprehending inferential
29
meanings. Given that the opposite results was obtained from people with LHD and
people with RHD, the findings ofthis study supported the critical role ofthe right
hemisphere in the comprehension offamiliar phrases, such a idioms, and formulaic
speech.
Bottini and colleagues (Bottini, Corcoran, Sterzi, Paulesu, Schenone, Scarpa,
Frackowiak, & Frith, 1994) measured the performance ofsix neurologically normal
adults, age between 21 and 35 years, on three tasks, including metaphor task, sentence
task, and lexical decision task, through the positron emission tomography (PET). In the
metaphorical sentence task, new and unusual metaphors were used to explore the specific
cognitive process. Participants were instructed to decide whether a sentence was
metaphorical or not. In the literal sentence task, participants were asked to decide
whether sentences were plausible or not at the literal level. In the lexical decision task,
participants were asked to identify the non-word in a sentence-like string. It was revealed
that processing ofmetaphors activated several areas in the right hemisphere compared to
the processing of literal sentences. Additionally, an extensive activation in the left
hemisphere was observed in the processing ofcomplex sentences than single words. The
results supported that the right hemisphere played a specific role in complex language
that normally required a holistic or integrated approach to understand it. It was worth
noting that mental images might be crucial to understand metaphors since four out of six
participants reported that they used mental imagery during the metaphors task.
Anaki, Faust, and Kravetz (1998) investigated how lexical metaphors were
processed in the two hemispheres by 56 undergraduates. The stimuli consisted of30
30
priming Hebrew words, 90 target words, and 90 target non-words. The priming words
had literal and metaphoric meanings, such as rolling and feeble. Thirty target words were
related to the metaphoric meaning ofthe priming word, 30 target words were related to
the literal meaning ofthe priming word, and 30 words were unrelated to the priming
word. Stimuli were presented to either left or right visual fields, and participants were
asked to determine whether the stimulus was a word or a non-word. The different
activations ofmetaphoric and literal meanings were found in the cerebral hemispheres.
The results supported the enhanced role ofthe right hemisphere in metaphoric
comprehension ofsingle word metaphors.
In order to understand the neural network underlying idiom comprehension and to
identify which brain areas were responsible for different cognitive functions, Zempleni
and colleagues (Zempleni, Haverkort, Renken, & Stowe, 2007) conducted a fMRI study
on 15 participants, mean age at 30.8 years old. Four experimental conditions, including
literal sentences with an ambiguous or an unambiguous idiom verses figurative sentences
with an ambiguous or an unambiguous idiom, were created using 64 ambiguous and 32
unambiguous Dutch idioms. Each sentence was presented visually on the computer
screen, and participants were instructed to read the sentence silently. Participants made
judgments about the relatedness between the sentence and a word that appeared after the
sentence. After the fMRI scan, a behavioral idiom comprehension task was carried out.
Stimuli were 30 literally plausible idioms selected from the idioms used in the fMRI
experiment with two extra new contexts that represented either the idiomatic
interpretation or literal interpretation. Six conditions were created: an idiomatic sentence
31
followed by the idiomatically-related target word; an idiomatic sentence followed by the
literally-related target word; an idiomatic sentence followed by the unrelated target word;
a parallel condition created for the literal sentences. Similar to the fMRI study,
participants were instructed to decide ifthe sentence-target pairs were related or not. It
was revealed that idiom comprehension activated both hemispheres, and figurative
language comprehension was more effortful in language areas compared to the literal
processing. Moreover, the findings indicated that idiomatic sentences activated more
areas that were typically for language processing compared to literal sentences, and the
right hemisphere was more sensitive in ambiguous idioms than unambiguous ones.
Different from previous studies, Huber-Okrainec, Blaser, and Dennis (2005)
examined idiom comprehension in relation to the literality, compositionality, and
contextual bias in children with spina bifida meningomyelocele (SBM), a
neurodevelopment disorder associated with agenesis and/or hyperplasia ofthe corpus
callosum. Participants were 38 children with SBM and 38 typically developing children,
age between seven to eighteen years old. Forty-eight highly familiar idioms varying on
the literality and compositionality were stimuli. Two pictures, one representing figurative
meaning and one representing literal meaning, were created for each idiom. Halfofthe
idioms were presented verbally in isolated sentences and halfofthe idioms were
presented verbally with a contextual sentence that was biased towards the figurative
meaning. Participants were instructed to press the yes/no button as fast as they could to
judge whether the picture represented the idiom. The accuracy and response time were
measured. The results showed that children with SBM were impaired in idiom
32
comprehension, and the compositionality, literality, and context bias affected how
children with SBM understood idioms. It was evident that children with SBM had great
difficulties making inferences and understanding non-decomposable idioms.
Huber-Okrainec and colleagues (2005) pointed out that the difficulties were associated
with congenital agenesis ofthe corpus callosum, which further suggested that
understanding idioms required the inter-hemispheric integration. The results were also
supported by a recent ERP study done by Proverbio, Crotti, Zani, and Adomi (2009) on
15 undergraduate students in Italy. They proposed a bilateral involvement in idiom
comprehension, and a direct access to the idiomatic meaning ofthe idiom without
suppression ofits literal meaning.
Whether idiom comprehension was processed bilaterally or unilaterally yielded
inconclusive results. In addition to neuroimaging studies reviewed above, studies on
people with brain damage also provided the direct and robust clinical evidence for the
understanding ofidiom comprehension in the brain. Therefore, literature reviewed in the
following section focused on how people with widespread brain damage, such as
dementia and Alzheimer's disease, and people with left or right brain damage
comprehended idioms.
Idiom Comprehension in People with Brain Damage
Observations and clinical evidence from people with clinical dysfunctions
provided valuable insights ofwhich factors play critical roles during language processing.
However, it is still not clear whether the right hemisphere or the left hemisphere is
involved in idiom comprehension. Given that limited research in idiom comprehension
33
was studied in neurologically normal aging populations, literature review in people with
brain damage may offer a window to explore the relationship between aging, brain
activity, and figurative language processing.
Comprehension offigurative language, such as metaphors and proverbs, were
examined in people with brain damage (Winner & Gardner, 1977; Brundage, 1996),
people with traumatic brain injury (Moran, Nippold, & Gillon, 2006; Towne & Entwisle,
1993), people with aphasia (Brundage & Brookshire, 1995; Chapman, Ulatowska,
Franklin, Shobe, Thompson, & McIntire, 1995; Ulatowska, Chapman & Johnson, 1995),
people with dementia (Brundage, 1996; Chapman et aI., 1997), and people with
Alzheimer's disease (Amanzio, Geminiani, Leotta, & Cappa, 2008; Chapman et aI.,1997).
Since the focus ofthe present study was idiom comprehension, literature associated with
idiom comprehension in widespread brain damage and focal brain damage was reviewed
to understand idiom comprehension from different aspects.
People with Widespread Brain Damage
The abstract verbal reasoning, concept formation, and quality ofthinking are
critical in figurative language processing. However, people with widespread brain
damage often lose the ability of abstract thinking and frequently substitute it with
concrete interpretation (Kempler, Van Lancker, & Read, 1988). Obler and Albert (1984)
pointed out that people with dementia had difficulties interpreting idioms and proverbs,
and their interpretations were fairly concrete even in people with higher education.
Individuals with early-stage Alzheimer's disease would have no difficulty explaining the
idioms, but people with middle-stage Alzheimer's disease would have a hard time
34
explaining an idiom. However, ifmultiple choice responses were provided, people with
middle-stage Alzheimer's disease were able to identify idiomatic interpretations.
Papagno and colleagues (Papagno, Lucchelli, Muggia, & Rizzo, 2003)
investigated the relationship between idiom comprehension and the executive function on
individuals with Alzheimer's disease (AD). In the first experiment, 15 people with
probable AD were recruited; 40 unambiguous and familiar verbal idioms were used as
stimuli. Participants were asked to choose a picture from two possible drawings when the
experimenter read a sentence corresponding to an idiom. One ofthe drawings represented
the idiomatic/figurative meaning and the other one displayed the literal meaning ofthe
idiom. Participants with AD knew the meanings ofthe idioms but were unable to
suppress the literal interpretation when it was overtly represented. Papagno and
colleagues (2003) claimed that the activation ofliteral meanings might be faster and
stronger than the activation offigurative meanings in people with AD, which was
consistent with the results ofthe Stroop Test used to measure the inhibition function.
In the second study, fifteen people with AD different from the first study were
asked to perform a sentence-to-picture matching task for idiom comprehension and also
orally explain the meaning ofeach idiom. The same idioms from the first experiment
were used. It was revealed that people with AD performed better in the oral task than in
the sentence-to-picture matching task. According to the results ofthe two experiments,
the literal meaning interference existed whenever the literal meaning ofthe sentence was
available for people with AD. In addition to ineffective inhibition ofthe literal meaning,
activation ofthe figurative meaning was not fast and sufficient during idiom
35
comprehension. Papagno and colleagues (2003) argued that since suppression played a
key role in the idiom comprehension task, the executive function was critical in figurative
language processing.
Kempler, Van Lancker, and Read (1988) administered three tasks on twenty-nine
people diagnosed with AD and forty-three neurological normal older adults, age between
50 to 82 years old. Three tasks included single words, familiar phrases with idioms and
proverbs, and novel phrases, and each task contained 10 stimuli. In the familiar phrases
task, three foils, one contains the referential representation, one contains meanings that
were opposite or related to the idiomatic meanings, and one irrelevant choice, were
created. In the novel phrases task, all three foils were ungrammatical, such as
subject-object reversal. Participants were instructed to choose a picture from four line
drawings that best represented the stimuli. The results indicated that people with
dementia had difficulty understanding abstract language, such as idioms and proverbs,
compared to the neurologically normal elders. In addition, people with AD had difficulty
recognizing the overall pattern offamiliar phrases and matching the phrases to its
complex meanings. It was suggested that people with AD had impairments in the holistic,
integrative, and global processing and had difficulty processing familiar phrases that have
holistic forms and integrated meanings. According to the error analysis, people with AD
often chose concrete instead ofrelated or irrelevant responses. It reflected a deficit in the
processing ofabstract meanings. The results also found that the performance ofpeople
with AD in figurative language was similar to people with the right hemisphere damage
who also performed better in novel phrases than in familiar phrases.
36
A recent study done by Rassiga and colleagues (Rassiga, Lucchelli, Crippa, &
Papagno, 2009) also investigated the comprehension ofambiguous literally plausible
idioms in people with mild AD. Fifteen participants with mild probable AD and 15
neurologically normal participants with matched age, education, and gender performed
two tasks, including a sentence-to-word matching task and a picture matching task. A
total of28 ambiguous idioms containing both figurative and literal meanings were used.
In the sentence-to-word matching task, each idiom was paired with four words and
participants needed to point to the word that matched the idiomatic meaning ofthe
sentence. In the picture-matching task, each idiom was represented by four line-drawings
containing one figurative interpretation, one semantic associated meaning, and two foil
unrelated semantic meanings. The results showed that participants performed
significantly better in the sentence-to-word matching task than in the picture-matching
task. It was evident that people with AD were impaired with ambiguous idioms and were
unable to suppress the literal interpretation ofan idiom when the pictorial representation
was available. In addition, a high variability in the picture-matching task in both AD
participants and healthy elders was observed. Therefore, Rassiga and colleagues (2009)
hypothesized that elders with good executive function would perform well on the idiom
comprehension task; however, these with lower executive function would perform at a
lower level.
However, the result ofthe study done by Papagno (2001) was not in line with
previous studies. Thirty-nine people with mild dementia but with normal comprehension
ofeasy commands were recruited. Metaphor and idiom comprehension tasks were
37
administered. In the idiom comprehension task, participants were asked to explain 20
opaque idiomatic phrases. Twenty-three out ofthirty-nine participants were evaluated
again after six to eight months. It was revealed that figurative language comprehension
was preserved in most ofthe people with mild AD even at the later stage. However, the
participants ofthis study had good comprehension in order to perform the tasks which
created a bias in the findings. In the error analysis, the literal interpretation was
commonly observed for idiom comprehension, but an attempt to generate the figurative
meanings ofmetaphors was frequently presented.
In addition to people with brain damage, people with schizophrenia also have
difficulty understanding nonliterallanguage, such as idioms and proverbs. The plausible
explanations for the deficits were the failure to use the relevant context information to
facilitate their understanding and to inhibit unrelated contextual information during
encoding contextual-relevant information. Titone, Holzman, and Levyn (2002) recruited
32 individual diagnosed with schizophrenia, mean age at 38.8 years. The stimuli
consisted of24 literally plausible idioms and 24 literally implausible idioms which were
further embedded in neutral sentence contexts as sentence primes. The idioms were
non-decomposable, moderately predictable, and moderately frequent. Visual target words
related to idiomatic and literal meanings ofthe sentence primes were selected, and fillers
were randomly mixed with experimental items. Four experimental conditions, including
idiom prime-idiom target, idiom prime-literal target, control prime-idiom target, and
control prime-literal target, were designed. The latency and accuracy were measured.
Participants were instructed to listen to the sentence primes and to make lexical decisions
38
to the visual target words. It was found that idioms primed significantly for literally
implausible idioms but not for literally plausible idioms in people with schizophrenia;
however, idiom priming was found in both literally plausible and implausible idioms for
normal controls. Moreover, people with schizophrenia showed literal word priming for
literally plausible idioms only, which reflected the possible interference ofactivation of
both literal and idiomatic interpretations. It was evident that people with schizophrenia
used idiomatic context to generate idiomatic meanings when no other interpretation ofthe
sentence was available. Additionally, this population only showed impairments in idiom
comprehension when they had to choose between plausible interpretations of idioms.
People with Focal Brain Damage
In addition to people with widespread brain damage, such as Alzheimer's disease
and dementia, how people with focal brain damage processed idioms provided the unique
clinical evidence for idiom comprehension. Studying idiom comprehension in this
population not only extended our understanding ofidiom processing in the brain but also
provided the evidence for clinical assessments in figurative language comprehension after
brain damage. The left hemisphere is commonly referred as a central place oflanguage
processing. However, studies (Bottini et aI., 1994; Burgess & Chiarello, 1996; Van
Lancker & Kempler, 1987) indicated that the right hemisphere in fact plays a critical role
in figurative language processing. Therefore, studies in people with left or/and right
hemisphere damage offer a path to understand how the left or right hemisphere is
involved in the idiom comprehension process.
39
Bush and Drummond (1985) attempted to understand the comprehension and
production ofidioms in people with aphasia. Five people with fluent aphasia and five
people with nonfluent aphasia, age between 23 to 69 years old, were presented 15 sets of
pictures, each containing four pictures, in the comprehension task. Four pictures included
one literal description, one idiomatic description, one literal variation, and one unrelated
picture. Participants had to choose the picture that best explained the phrase. In the
production task, participants were asked to provide the explanation for 15 idioms that
were read by the experimenter. The findings showed that the production and
comprehension ofidioms were difficult for all people with aphasia regardless ofthe
severity and the type ofaphasia although participants with more severe aphasia
performed less well on both comprehension and production tasks. Overall, the
comprehension task was performed better than the production task. It was worthy noting
that people with aphasia tended to select literal pictures over the literal variation or the
unrelated pictures although they identified relatively more idiomatic representations in
general.
Papagno and Genoni (2003) conducted serial tests on 10 people with aphasia to
examine the relationship between idiom comprehension and syntactic processing.
Participants suffered with aphasia and had the syntactic but not semantic deficits. Forty
familiar verbal idioms were selected, and three pictures that each represented idiomatic
meanings, figurative meanings, and unrelated situations were created. In the idiom
comprehension test, participants chose a picture matching the idiom that was read by the
experimenter. In the literal sentence comprehension task, participants chose from two
40
pictures that better matched the sentence. Lastly, in the grammaticality judgment test, 40
literal sentences and 40 idioms used in the idiom comprehension test were used. Halfof
the literal sentences and idioms were presented in their correct forms, but the other halfof
the stimuli contained an omission or substitution which violated syntactic rules.
Participants were asked to judge ifthe word string was grammatically correct or not. The
findings indicated that the idiom comprehension underwent syntactical analysis since the
syntactic deficit interfered with idiom processing.
Whether comprehension ofidioms was impaired or intact in people with left brain
damage was also investigated in the study ofPapagno et al. (Papagno, Tabossi, Colombo,
& Zampetti, 2004) for which 10 people with aphasia were recruited. Thirty-four
unambiguous idioms were used, and three line drawing pictures were created for each
idiom, including one representing the possible literal expression, one showing idiomatic
interpretations, and one representing the unrelated situation. Three tasks were
administered, including a string-to-picture matching task, a single word comprehension
task, and a literal sentence comprehension task. It was found that comprehension of
idioms, measured by the string-to-picture matching task, was severely impaired in people
with aphasia. However, this deficit could not be reduced to the impaired comprehension
ofeither words or literal sentences since they performed better on word and sentence
comprehension tasks than on the idiom comprehension task. To avoid the plausible
interpretation bias from the picture matching task, Papagno and colleagues (2004) asked
participants to provide verbal explanations for each idiom. It was revealed that idiom
comprehension in people with aphasia was severely impaired. Moreover, the inference
41
was strong when the literal interpretation ofan idiom was available for people with
aphasia.
Cacciari and colleagues (Cacciari, Reatio, Colombo, Padovani, Rizzo, & Papagno,
2006) used the string-to-word matching task instead ofthe picture matching and
definition tasks on 15 patients with aphasia to examine idiom comprehension.
Twenty-three familiar ambiguous idioms, each pairing with four target words, were used,
including one target word corresponding to the idiomatic expression ofthe string, one foil
that was semantically related to the last word ofthe idiom string, and two unrelated foils.
One ofthe unrelated foils was either an abstract or a concrete word depending on the
nature ofthe idiom string, and the other unrelated foil was a noun that could plausibly
complete the verb in the verb phrase. Participants were asked to point to the word that
matched the idiomatic meaning ofthe string and to rate the familiarity ofthe idiom at the
end ofthe experiment. The results showed that people with aphasia were more impaired
in idiom comprehension compared to normal controls. Cacciari and colleagues (2006)
claimed that the idiom comprehension in people with aphasia exceeded single-word level,
which was in line with Papagno et aI's findings (2004).
Papagno and Caporali (2007) intended to examine the effects oftasks and idiom
types during idiom comprehension on people with aphasia. Fifteen individuals with
aphasia were given the word comprehension task, oral naming task, and dual task that
assessed the executive function. In the first experiment, forty familiar, opaque, and
unambiguous verbal idioms were selected to further construct four line-drawing pictures
for each idiom, including idiomatic and literal interpretations, content word related
42
situations, and unrelated situations. Participants were asked to choose the corresponding
picture from the four options. Each idiom was also paired with four target words,
including words corresponding to the idiomatic interpretation, words associated with the
last word ofthe idiom string, unrelated concrete or abstract words depending on the
nature ofthe idiom, and unrelated words that could plausibly complete the verb in the
verb phase. Participants had to point to the word that matched the idiomatic meaning of
the sentence. Participants were also asked to provide an oral definition of each idiom. In
the second experiment, the only difference from the first experiment was that 15
ambiguous and highly familiar idioms were used as stimuli. The findings indicated that
people with aphasia had difficulty in idiom comprehension and the degree ofdifficulty
varied depending on the type oftasks and type ofidioms used in the study. Their
difficulties possibly resulted from their general language difficulties combining with a
reduced inhibition mechanism ofthe literal interpretation. In addition, the literal
interpretation was not activated or easily suppressed for non-ambiguous idioms. However,
the literal interpretation remained activated and the inhibition was difficult and delayed
for ambiguous idioms.
A recent study done by Norswski (2009) on three participants with aphasia also
supported the findings that people with aphasia had more difficulties with interpreting
transparent idioms. The literal interpretations were more easily activated than opaque
ones, whose literal meanings were less easy to access. They used 60 Polish idioms
containing equal numbers ofdecomposable or non-decomposable and transparent or
opaque idioms in two multiple choice tasks, in which idioms were embedded in context.
43
In the first task, each of 30 idioms was paired with three target phrases matched in
lengths and syntactic forms, including an idiomatic paraphrase, a literal paraphrase, and
an unrelated foil phrase. Different 30 idioms were used in the second task, and the target
phrases for each idiom included one idiomatic paraphrase, one unrelated foil containing
the basic words from the idiom, and one filler phrase. It was revealed that transparent
idioms were more difficult than opaque ones for people with aphasia, and they had
tendency to interpret the idiom literally instead ofidiomatically. When the literal
interpretations were eliminated from the second task, participants performed better.
According to the studies reviewed above, it was evident that the left hemisphere
was involved in idiom comprehension process, which was in line with the theory that
language is processed mainly in the left hemisphere. However, the right hemisphere is
assumed to playa critical role in figurative language comprehension. Therefore, the
studies reviewed in the following section intended to explore the role ofthe right
hemisphere during the figurative language processing, especially idiom processing. How
people with right brain damage comprehend idioms was examined in the study ofMyers
and Linebaugh (1981). Six people with unilateral right hemisphere damage, six people
with unilateral left hemisphere damage, and six neurologically normal controls read five
two-sentence stories that each story delivered a common idiomatic expression. Each story
was accompanied with five pictures, including the correct context-correct interpretations,
correct context-literal interpretations, wrong context-correct interpretations, wrong
context-literal interpretations, and correct context-opposite interpretations. Participants
were asked to point to the picture that best described the story. It was found that people
44
with right hemisphere damage (RHD) tended to choose literal expressions and performed
worse in comprehending figurative language compared to normal controls even though
the supportive context was provided. In addition, people with RHD interpreted idioms by
breaking them down into constituent elements rather than understanding them as a whole.
Compared to people with RHD, people with left hemisphere damage used contextual
information better in idiom comprehension. Therefore, the findings ofthis study
suggested that people with RHD had difficulty comprehending the implication and
intention ofthe idiomatic phrases.
Kempler and Van Lancker (1993) used the Familiar and Novel Language
Comprehension (FANL-C) Test, a picture-pointing task requiring no verbal explanation,
to test 13 people with unilateral right hemisphere damage and 13 people with unilateral
left hemisphere damage in order to understand the loss of familiar language. The test
contained 20 familiar phrases, including proverbs, idioms, and contextually bound social
interaction formulas (e.g. I'll get back to you later) and 20 novel sentences (e.g. He s
racing a truck against a horse). Each phrase was expressed by four pictures with one
correct meaning, one related meaning to the familiar phrase, and two foils with concrete
interpretations for the familiar phrases. The findings revealed that familiar phrases were
processed in both hemispheres, while novel sentences were processed only in the left
hemisphere. Therefore, the right hemisphere might play an important role in certain
aspects oflanguage processing.
Kempler, Van Lancker, Marchman, and Bates (1999) used the same test, the
Familiar and Novel Comprehension Test (FANLC), to test participants' idiom
45
comprehension. Stimuli were 20 familiar idiomatic phrases and 20 literal sentences. Foils
for idiomatic phrases included two foils related to the individual words in the stimuli
sentence and the other foil was related to the figurative meaning ofthe idiomatic phrase.
Participants were asked to choose one picture that matched the sentence read by the
experimenter. In the first study, participants were 25 adults with left hemisphere lesions,
mean age at 58 years, 16 adults with right hemisphere lesions, mean age at 65 years, and
42 neurologically normal adults, age between 40 and 79 years. The results indicated that
the literal and idiomatic language was involved in different hemispheres in adults. It was
evident that people with RHD performed poorly on familiar nonliteral expressions, but
people with LHD were impaired in comprehension ofliteral expressions.
In order to understand how people with different regions ofbrain damage
processed idiomatic meanings during online spoken sentence comprehension, one patient
with Wernicke's aphasia, one patient with global aphasia, one patient with right
hemisphere damage, and one age-matched neurologically normal elder participated in the
Hillert study (2004). Stimuli were 16 common ambiguous German noun idioms, each
embedded in two types ofcontexts, including idiomatically-biasing and literally-biasing
contexts. Participants were instructed to listen to each sentence and were asked to press
the keys when they saw a word or a pseudo-word present on the monitor. Both patients
with Wernicke's aphasia and global aphasia performed as well as the normal control and
accessed both idiomatic and literal meanings with no signs ofimpairments. Moreover, it
was found that the patient with RHD in this study showed no difficulty accessing
46
idiomatic or literal meanings, which could be assumed that the posterior lesion ofthe
right hemisphere might not affect idiomatic processing.
It was still not clear that ifpeople with left hemisphere damage (LHD) were more
impaired in literal language than in figurative language and whether the level of
impairment in people with LHD was the same as in people with right hemisphere damage
(RHD). Papagno and colleagues (Papagno, Curti, Rizzo, Crippa, & Colombo, 2006)
compared how idioms were processed in 15 participants with RHD and 12 participants
with LHD/aphasia. Stimuli were forty unambiguous verbal idioms with a variety of
transparency. In the idiom comprehension task, participants were first asked to pick one
picture from three pictures that best matched the sentence read by the experimenter and
then to provide an oral definition ofeach item. In the literal sentence comprehension task,
40 sentenced were stimuli, and participants chose a picture that best described the
sentence read by the experimenter from two options. The only difference between the two
pictures was a single detail such as a boy is pushing a girl versus a boy is pushed by a
girl. The results indicated that participants with LHD were more seriously impaired in
both literal and idiomatic languages. The results ofthis study did not support the
prevailing role ofthe right hemisphere in idiom comprehension, but it provided precious
information regarding idiom processing. First, participants with LHD had the impaired
linguistic analysis which affected their idiom processing. Secondly, participants with the
frontal lobe lesion had difficulty inhibiting the literal meaning during the picture
matching task. Thirdly, the use of string-to-picture matching task might underestimate the
idiom comprehension ability of some people with brain damage.
47
In the Tompkins, Boada, and McGarry study (1992), twenty adults with RHD, 20
with LHD, and 20 neurologically normal controls performed an on-line word-monitoring
task and an off-line definition task. In the on-line word-monitoring task, participants
listened to the spoken sentences and pressed the button when the specified target words
were presented. The response time was recorded in three experimental conditions,
including idiomatic, literal, and control. In the off-line definition task, twelve highly
familiar idioms, six from the word-monitoring task, were presented auditorily and
visually. The findings from the online task indicated that three groups responded more
quickly in idiomatic phrases than the same nouns in the control context, and the idiomatic
phrases were processed automatically regardless ofthe context. The similar reaction time
among three groups suggested that the initial activation and retrieval offamiliar idioms
remained intact regardless ofthe site ofcerebral damage. In the offline definition task,
both groups with unilateral brain damage performed worse than the control group.
Tompkins and colleagues (1992) claimed that the idiom interpretation deficits in people
with unilateral brain damage did not show in the early activation and the retrieval stage
but rather in the later stage of information processing.
Cacciari and colleagues (Cacciari et aI., 2006) proposed three possible
explanations for the impairment ofidiom comprehension in people with aphasia,
including the impairment ofthe inhibition system, impairment ofthe recognition and
activation mechanisms, and impairment of linguistic processing. Papagno et aI's study
(2004) also supported that people suffering with aphasia had difficulty suppressing the
literal meaning during the idiom processing. The reasons that result in the impairment in
48
idiom comprehension for people with left hemisphere damage needs further research.
However, it should be noted that there were methodology concerns regarding the two
commonly used idiom tasks used to study people with impaired language, including
string-to-picture matching and the oral definition task. The string-to-picture matching
task might underestimate idiom comprehension since pictures which easily activated the
literal interpretation strongly interfered with the figurative meaning. This point ofview
was also supported in Papagno et aI's study (2006). Other than the string-to-picture
matching task, given limited speech output in people with left-brain-damage, the oral
definition task might also underestimate their actual comprehension (Cacciari et ai.,
2006).
Factors Involved in Idiom Comprehension
Nunberg, Sag, and Wasow (1994) suggested that the difficulties in analysis of
idioms resulted from a confusion ofthe key semantic properties associated with idiomatic
meanings, including conventionality, transparency and compositionality. Additionally,
idioms vary on the transparency, decomposability, ambiguousness, and these factors
affected how people interpret idioms (Cacciari, Reatio, Colombo, Padovani, Rizzo, &
Papagno, 2006).
The relationship between the literal and figurative use ofcommon idioms were
examined in the Popiel and McRae study (1988). Sixty idioms were selected as stimuli
and two questionnaires, each containing 30 idioms, were constructed. Each idiom was
followed by its literal and figurative definition. Participants were instructed to rate each
definition on a seven-point scale based on its frequency ofuse and familiarity. The
49
findings pointed out that the figurative interpretation ofidioms were more frequently
used and more familiar to participants compared to literal definitions. Moreover, subjects
tended to rate higher for the frequency ofuse and familiarity on figurative interpretations
than on literal interpretations, and the rating varied more widely on the literal ratings than
on the figurative ratings.
Schweigert (1986) examined the differences between the comprehension of
familiar and less familiar idioms on college students. Stimuli were eight idiomatic
sentences, eight literal sentences, eight unbiased sentences, and 24 control sentences
mixed with 76 filler sentences. Halfofthe idioms were familiar and the other halfwere
unfamiliar based on the ratings. Participants were asked to read the sentences showed on
the computer screen and to press a button ifthey understood the sentence. A recall task
was administered about every five sentences. The finding indicated that it took a longer
time for participants to comprehend a sentence containing a less familiar idiom than a
sentence containing a familiar idiom. Moreover, the familiarity effect was also found in
both idiomatic and literal sentences that were generated from the idioms. Therefore,
Schweigert (1986) suggested that the familiarity ofthe idiom needed to be taken into
account for the models ofidiom processing, and the less familiar idioms would take a
longer processing time compared to familiar idioms.
In another study ofSchweigert (1991), stimuli were 12 most familiar and 12 less
familiar idioms. One figurative paraphrase and one literal paraphrase were developed for
each idiom with the target sentence ending in each paraphrase. Undergraduates were
instructed to read out loud the sentence which was first presented only with one or two
.._-- ----------
50
words from the sentence and the presentation was continued until the participant read the
entire sentence correctly. The number ofpresentations needed to read each sentence
correctly was recorded, and a cued recall task was administered after the experiment. It
was found that the number ofpresentations was significantly fewer for control sentences
than the other sentences except the less familiar idioms used literally. Additionally,
familiar idioms were better recalled than less familiar ones. Schweigert (1991) suggested
that for less familiar idioms, the literal meaning might be processed initially as
nonfigurative sentences. For those highly familiar idioms, the figurative meanings were
processed first. The literal meanings were processed only ifnecessary. For idioms that
fell in the middle ofthe familiarity spectrum, the processing time for both literal and
figurative meanings might not be different.
An idiom can be classified based on the familiarity of its figurative meaning and
the likelihood of evoking its literal meaning. Cronk and Schweigert (1992) attempted to
determine the influence ofthe familiarity and literalness in idiom comprehension. Stimuli
were 40 idioms that were equally divided into four categories, including familiar idioms
with high literalness, less familiar idioms with high literalness, familiar idioms with low
literalness, and less familiar idiom with low literalness. Each idiom was embedded into a
sentence representing either the literal meaning or figurative meaning. Twenty control
sentences with no idioms were constructed as fillers. The sentences were represented on
the computer screen, and the reading time for each sentence was recorded. The findings
indicated that the reading time was related to both ofthe familiarity and literalness. The
familiar idioms were read faster than less familiar ones, and idioms with high literalness
- --_._------------
51
were read more rapidly than those with low literalness. It was worth noting that the effect
offamiliarity appeared only when the idiom was used figuratively, and sentences
containing less familiar idioms with lower literalness were the most difficult to
comprehend. Therefore, both ofthe familiarity and literalness affected reading times of
sentences containing idioms.
Titone and Connine (1994) examined the familiarity, compositionality,
predictability, and literality of 171 idioms rated by 226 college students. The familiarity
was rated based on the idiom's frequency of encounter and how well the participant knew
the meaning ofthe idiom on a seven-poing scale. Participants were instructed to
categorize idioms into three compositionality categories, including nondecomposability,
abnormal decomposability, and normal decomposability. The difference between
abnormal and normal decomposability was that ifthe idiom contained words that was
directly related to their figurative interpretations. In addition, participants had to rate if
the idiom had the plausible literal interpretation on a seven-point scale and to complete
the phrase with the first word that came to mind as a way to evaluate the predictability of
each idiom. The predictability was found to be positively correlated with the ratings of
familiarity, frequency, and meaningfulness. Titone and Connine (1994) pointed out that
36% ofthe idioms in the study were judged uncategorizable in the compositionality,
which suggested that the judgment ofsemantic analyzability was difficult.
How children and adolescents judged the idiom familiarity and idiom
transparency was explored in the study ofNippold and Taylor (2002). Fifty children,
mean age at 11;3 years, and 50 adolescents, mean age at 16;6 years, were given a
52
familiarity judgment task, an idiom comprehension task, and a transparency task. In the
familiarity judgment task, participants rated 20 idioms based on their familiarity on a
five-point scale. In the idiom comprehension task, participants were presented the same
20 idioms that embedded in a brief story context and then were asked to choose the best
answer from four interpretations ofthe idiom. Lastly, in the transparency task,
participants were instructed to indicate how closely the literal and figurative meanings
were related using a three-point scale. The findings indicated that children were less
familiar with idioms and performed worse in the comprehension task in comparison with
adolescents. However, there was no difference between two groups on the transparency
judgment task. Nippold and Taylor (2002) suggested that the transparency and familiarity
were strongly correlated to idiom comprehension in children and adolescents and needed
to be considered while examining idiom comprehension.
A more recent study done by Libben and Titone (2008) looked at the
multidimensional nature ofidiom processing. Two hundred and nineteen idiomatic
expressions and 30 non-idiomatic literal phrases were rated based on nine dimensions,
including the familiarity, meaningfulness, predictability, literal plausibility, global
decomposability, normal decomposability, abnormal decomposability, verb relatedness,
and noun relatedness. The global decomposability was defined as how each word in an
idiom contributed to the overall figurative meaning, and verb and noun relatedness was
viewed as to what extent that the constituent noun or verb was related to the overall
figurative meaning. The results indicated that the familiarity was positively correlated
with the meaningfulness and predictability but negatively correlated with the literal
53
plausibility. In addition, the meaningfulness was positively correlated with the verb and
noun relatedness. It was also concluded that the decomposability effects took place when
people were forced to pay attention to the semantic level, such as the meaningful
judgment, rather than to the lexical level, such as the predictability offinal words in the
phrase. Additionally, high-familiar idioms involved the direct retrieval but low-familiar
idioms required the decomposition for comprehension. The results ofthe study indicated
that increasing the familiarity, literal plausibility, and global decomposability facilitated
idiom comprehension. In general, idioms were represented and retrieved as units in
memory and could interact with the ongoing compositional analysis ofthe meaning ofthe
phrases. However, the decomposability could not be determined until the last word was
encountered. Therefore, deciding whether the idiom was decomposable or not occurred in
the very late stage ofthe processing.
Tabossi, Fanari, and Wolf (2009) asked 36 undergraduate students to read strings
and then to press a button as soon as they saw a meaningful Italian phrase. The test
materials were 16 decomposable idioms, 16 nondecomposable idioms, 16 cliches, and
control strings that were created by replacing the first one or two words ofthe
conventional strings. They found that all familiar verbal expressions were pressed faster
than the controls, but no significant difference was found among three different types of
conventional expressions. Additionally, in comparison to the predictability and
decomposability, the familiarity played a major role in explaining the idiom superiority
effect.
54
Summary ofIdiom Comprehension
Idioms are essential in everyday conversation; however, they are less studied and
have received less attention compared to other aspects oflanguage. The results ofhow
idioms are processed were in the debate depending on: the types oftasks used in the study,
different levels ofprocessing that the study measured, the types ofidioms used in the
study, and how researchers measured idiom comprehension. In addition to the difference
in the research method, idioms varied widely in the transparency, familiarity, and
compositionality that might impact how people interpret idioms. Although different
models ofidiom processing were proposed, how people interpreted idioms differently
from regular phrases still remained unclear and debatable. According to the studies
reviewed above, the right hemisphere was found to play an important role in idiom
comprehension although the results were inconclusive. Therefore, interpreting idioms or
familiar phrases may require the activation in both hemispheres. The clinical studies on
people with widespread brain damage and focal brain damage also provided valuable and
strong evidence for the role ofboth left and right hemispheres during idiom
comprehension even though the results were not consistent throughout.
Idiom comprehension was examined by different theories and models 1) to
understand the effect offamiliarity, transparency, and compositionality on
comprehending idioms, 2) to explore the role that the right or left hemisphere might have
during the idiom comprehension process, and 3) to discover how people with widespread
and focal brain damage processed idioms. However, how transparent and opaque idioms
are processed is rarely studied. The dual coding theory (Paivio, 1971, 1986) postulated
55
that concrete words and abstract words are processed differently since concrete words
activated both verbal and imaginal systems while abstract words are processed primarily
by the verbal system. Because ofthe activation in both systems, concrete words are better
retained and recalled compared to abstract words, which results in the concreteness effect.
Therefore, this present study attempted to determine ifthe dual coding theory was able to
explain how transparent and opaque idioms were processed differently and ifthe
concreteness effect was able to apply to the idiom comprehension process when
transparency ofthe idiom was taken into account.
Dual Coding Theory
Paivio (1971) claimed that the concrete language was retained better than abstract
language. Therefore, the dual coding theory, proposed by Paivio (1971, 1986), assumed
that there were two independent but interconnected systems, verbal and imaginal systems,
attributed to the different retrieval effect between concrete and abstract words. Concrete
words and phrases should be better recalled and retained than abstract words and phrases
because concrete words and phrases could be encoded both verbally and nonverbally.
Additionally, the dual coding theory suggested that abstract concepts were stored in the
language dominated hemisphere, the left hemisphere, but that concrete concepts were
processed in both hemispheres. Because ofthe two sets ofprocessing and both verbal and
imaginal representations for concrete words, concrete words were better recalled than
abstract words. It was assumed that concrete nouns were learned through the sensory
experience while the meanings ofabstract nouns were derived mainly from the networks
56
that were mostly made up of abstract words (Sabsevitz, Meddler, Seidenberg, & Binder,
2005).
Concreteness Effect
One hundred and one undergraduate students were recruited in the study of
Sadoski, Goetz, and Fritz (1993) in order to examine the effects of concreteness,
familiarity, comprehensibility, and interestingness in a sentence recall task. Two concrete
and two abstract sentences were written for each ten historical figures, and a total of40
sentences were used as the stimuli. In the first part ofthe study, 22 students rated 40
sentences based on the comprehensibility, concreteness, familiarity, and interestingness.
The rest ofthe 79 students read 20 sentences and then provided both immediate and
delayed written recalls. It was revealed that the concreteness had most influence on the
comprehensibility and on immediate and delayed recall tasks. The findings ofthis study
provided the evidence for the dual coding theory because the concreteness had a greater
effect on the comprehensibility and recall tasks compared to the content familiarity. In
addition, the concreteness showed much greater effects on both immediate and delayed
recalls than the comprehensibility, familiarity, and interestingness.
Sadoski (1995) further investigated whether the concreteness effects could be
extended beyond the sentence level to the text. Paragraphs about historical figures, one
highly familiar and one less familiar, were used as materials, and were further written in
either concrete or abstract language. College students were first asked to rate the
concreteness, content familiarity, and comprehensibility on a seven-point scale and then
to read either concrete or abstract passages. Participants were asked to write everything
57
they could remember for each historical figure within 10 minutes after they read the
passages. The results in the recall task showed that the difference between the concrete
and abstract paragraph recall was significant for highly familiar figures, but was not
significant for less familiar ones. In general, the concreteness effect in recalling concrete
paragraphs was found, and the advantage ofconcreteness in recall was also found in
concrete paragraphs with low familiarity content. In addition, the concrete content was
better recalled when the concrete and abstract information were equally familiar.
However, only when the abstract content was more familiar than the concrete content,
they were able to be equally recalled. Thus, this finding was consistent with the dual
coding theory.
Sadoski, Goetz, and Rodriguez (2000) investigated the concreteness effects on
comprehension, interestingness, and memorability in four types oftexts, including
persuasive, expository, literary stories, and narrative. Eighty undergraduate students
participated in the first study. Three concrete and three abstract texts for each text type
were used as materials. In order to facilitate recalls, one abstract and one concrete title
were written for each text. Half ofthe participants were presented 24 texts and asked to
rate each text on seven-point scales based on the familiarity, concreteness, interestingness,
and comprehensibility. The results ofthe rating indicated that the concreteness and
comprehensibility were highly correlated. In the second experiment, participants read 24
texts with either all concrete or all abstract titles, and the titles were served as recall cues
for participants to write what they remembered from the texts. The recalls were coded as
gist, elaboration, or distortion. The results indicated that the gist recall was better for
58
concrete texts and was greater for both persuasive and narrative texts but weaker for
literary texts. Overall, the concreteness was a strong predictor for the comprehensibility,
interestingness, and recall. In addition, concrete texts were more comprehensible,
interesting, and memorable than abstract texts.
Shibahara and Lucero-Wagoner (2002) studied how concrete words were
processed differently in both hemispheres, and whether abstract words were only
processed in the left hemisphere. Forty undergraduates were divided into two groups. The
stimuli were 20 concrete and abstract words that each paired with a semantically related
prime, such as apple-orange, a neutral prime composed of a row offour x's, such as
xxxx-orange, and a semantically unrelated prime, such as uncle-orange. Participants were
instructed to look at the prime without reading it aloud, but to read aloud for the target
words. The results indicated that the priming occurred in both hemispheres and was
activated greater in the right hemisphere than the left hemisphere for concrete words.
Overall, abstract nouns were mainly processed in the left hemisphere, and there were
hemispheric differences in processing concrete and abstract words.
Richardson (2003) investigated the concreteness effect through presenting
concrete and abstract words in different types ofsentences frames. This study tested the
interaction between the effects ofconcreteness and meaningfulness in a cued recall.
Twenty concrete and twenty abstract nouns with matched frequency were presented with
meaningful and anomalous sentence frames. While the nouns were read by the
experimenter, sixty undergraduates made judgments whether the word fitted into the
sentence frame to make a meaningful sentence. Participants were provided with 40
59
sentence frames and asked to write down the word that had been read by the
experimenter earlier but was considered against that sentence frame. It was found that the
concrete nouns were better recalled than the abstract nouns. The meaningful sentence
frames provided more effective cues than the meaningless sentence frames. Therefore,
the results supported the dual coding theory. In the second experiment, participants were
asked to perform a free recall ofthe words that had been read by the experimenter in
three minutes. It was again found that the performance in a free recall was a function of
the concreteness ofthe words and the meaningfulness ofthe sentence frames. Generally
speaking, performance was better in the cued recall than in the free recall when the nouns
were presented in meaningful sentence frames. However, there was no difference found
between the free recall and cued recall when the nouns were presented in anomalous
sentence frames. In addition, the increase in performance from the free recall to the cued
recall was greater in concrete nouns than in abstract nouns. Therefore, both results from
the free and cued recalls were consistent with the dual coding theory and supported that
the concreteness and meaningfulness were additive in their effects.
Kellogg, Olive, and Piolat (2007) asked 60 undergraduates to write definitions of
10 concrete or abstract nouns and to perform a verbal or visual working memory task
(WM) at the same time. The results showed that the interference ofWM task was shown
only when participants were writing the definitions ofthe concrete nouns but not the
definitions ofabstract nouns. Because participants did not need to generate the images
while writing the definitions ofabstract nouns, the interference did not show up. In
addition, participants responded more slowly in defining abstract nouns while performing
60
the verbal WM task but not the visual WM task. Therefore, their findings were also
consistent with the dual coding theory.
Different from studying on college students, Peters and Daum (2008) examined
whether age-related changes in verbal memory affected recollection ofthe concrete and
abstract words on a wide range of age group. Twenty-two young participants, aged 20-25
years, 20 middle-aged participants, age 33-51 years, and 23 older participants, age 52-72
years, saw 152 words equally drawn from four word groups, including concrete low
frequency, concrete high frequency, abstract low frequency, and abstract high frequency.
They were first asked to study words presented individually in an encoding trial. After 10
to 15 minutes, they were instructed to apply the remember/know techniques to recognize
words from a group ofwords, which 152 old words were intermixed with 76 distracters.
The findings showed that the age-related changes in recollection were affected by the
word concreteness. The recollection of concrete words decreased more than recollection
ofabstract words. The recollection ofconcrete words showed a steady decline with age,
but recollection ofabstract words reduced only from the young to middle-aged groups. In
addition, the word concreteness but not the familiarity had influence on age-related
changes in recollection.
Dual Coding Theory in Neuroimaging Studies
Kounios and Holcomb (1994) used the event-related potentials (ERPs) to test the
dual coding theory and the concreteness effect in the semantic processing. In the first
experiment, twelve participants, age between 19 to 30 years old, were presented 40
concrete, 40 abstract, and 40 pseudo words formed from abstract words. They were
61
instructed to rapidly decide whether the word was an actual English word or not. In the
second experiment, only concrete and abstract words from the first experiment were
presented to 12 different participants, and they were asked to judge ifthe target word
presented a concrete object or an abstract concept. The findings indicated that the
concrete words activated more semantic information in memory than the abstract words.
The concrete words also elicited greater activation in the right hemisphere than the left,
which was consistent with the dual coding theory. Additionally, the difference between
concrete and abstract words was larger in the right hemisphere and more anterior regions
ofboth hemispheres. Therefore, the findings ofthe study supported the dual coding
theory that different neurons were underlying the process during comprehension of
concrete and abstract words.
West and Holcomb (2000) also utilized ERPs to examine the processing of
concrete and abstract words under three levels ofprocessing: image generation, semantic
decision, and surface level evaluation. Thirty-six undergraduates were assigned to the
three experimental conditions. In the imagery condition, participants were asked to form
a mental image of each target word. In the semantic group, the participants were required
to process the meanings ofthe target word without generating any mental images. In the
surface condition, the participants only had to do a letter search, such as identifying ifthe
letter "n" was in the word "elephant/aptitude." Target words were 40 concrete and 40
abstract words embedded in eight types ofsentences with varied difficulties and
possibilities of generating mental images. The findings showed that the post-lexical
processing was necessary to elicit the concreteness effect since there was no difference in
.. - ------------
62
the reaction time and ERPs between the concrete and abstract words in the surface
condition. The reaction time was shorter for the concrete words than abstract words
especially in the imagery task. Therefore, this study supported the dual coding theory that
the use ofmental imagery and superior associative connections contributed to the
processing advantage of concrete words.
Another ERPs study was done by Nittono, Suehiro, and Hori (2002). Twelve
college students were asked to first rate the imageability ofthe words on a five-point
scale and to recall as many words as they could after the presentation ofthe stimuli. The
stimuli were composed ofhigh imagery words, low imagery words, and non-words. It
was revealed that the high imagery words were better recalled than the low imagery
words, and the high imagery words had denser and stronger associative interconnections
than the low imagery words. Moreover, the results indicated that high imagery words
activated additional imagery-related networks that were located in the right hemisphere
which were not activated by low imagery words.
Swaab, Baynes, and Knight (2002) also conducted an ERPs study and presented
320 word pairs auditorily to 12 adult participants. The stimuli were composed ofhigh
imageable and low imageable words, and the word pairs were either related or unrelated
in meaning. For example, pig-horse was high imageable and related in meaning, but
bile-sentence was low-imageable and unrelated in meaning. Participants were asked to
decide whether the words in pairs were semantically related or unrelated. It was found
that high imageable and low imageable words activated distinct areas, and the findings
63
supported the existence oftwo separate semantic systems, including verbal and
image-based systems.
The concreteness effect and the dual coding theory were also investigated through
the functional Magnetic Resonance Imaging (fMRI). Jessen and colleagues (Jessen, Heun,
Erb, Granath, Klose, Papassotiropoulos, & Grodd, 2000) investigated the cortical regions
that are responsible for the concreteness effect through the fMRI. Fourteen participants,
mean age at 31.5 years, participated in this study, and experimental materials were 120
concrete and 120 abstract words. Participants were presented 80 words and were asked to
encode as many words as possible for a later recognition task. During the recognition task,
stimuli were consisted of80 previously shown experimental words and 80 distract words,
and participants were asked to identify previously presented words. During encoding
phase, a stronger activation in the lower right parietal lobe was observed for the concrete
words, which supported that the right hemisphere participated in the processing of
concrete words. Moreover, the concrete words also activated areas in the left hemisphere
that were outside ofthe primary language areas. Conversely, the abstract words presented
a stronger activation in the left inferior frontal gyrus (Broca's area) and the right lateral
occipital gyrus. Therefore, Jessen and colleagues (2000) claimed that the superior
encoding ofconcrete words might result from the greater verbal context resources and the
additional activation ofa non-verbal, imagery-based system in the right parietal lobe.
Sabsevitz and colleagues (Sabsevitz, Meddler, Seidenberg, & Binder, 2005) also
conducted an fMRI study to determine ifthe difficulty ofthe task affected the
concreteness effect. The stimuli were composed of60 easy concrete, 60 hard concrete, 60
64
easy abstract, and 60 hard abstract nouns. The task difficulty varied depending on ifthe
meanings ofthe choice words were similar to the samples. Twenty-eight adults were
asked to select the word that was most similar to the meaning ofthe sample noun. The
results indicated that abstract nouns were mainly activated and processed in the left
hemisphere while concrete nouns were processed bilaterally. Therefore, this study
provided the evidence ofbilateral activation during the processing of concrete nouns,
which further supported the dual coding theory.
Another fMRI study done by Noppeney and Price (2004) asked 15 adults to
choose the word that had a similar meaning to the sample word. The words were drawn
from abstract concepts (intent, attempt), colors (pink, red), sounds (tone, melody), and
hand movements (squeeze, comb). The results showed that the abstract concepts activated
a left-Iateralized frontal temporal system that was usually involved in semantic
processing. Additionally, the abstract concepts and sensory-based semantics were
processed in different neural systems. Noppeney and Price (2004) suggested people
might generate a semantic context to obtain the meanings ofabstract concepts because
they could not be represented by real objects. Binder, Westbuy, McKiernan, Possing, and
Wedler (2005) also examined the concreteness effect using fMRI by requesting
participants to identify concrete and abstract words. Twenty-four participants, age
between 20 to 50 years old, were asked to make the lexical judgment whether each
stimulus was a word or a non-word on 200 stimuli, including 50 concrete words, 50
abstract words, and 100 non-words. It was evident that the concrete words elicited
stronger activation bilaterally in contrast to the abstract and non-words. Additionally, the
65
abstract words had stronger activation in the left posterior inferior frontal areas compared
to the concrete words and elicited mostly the left hemisphere in comparison to non-words.
Therefore, this study was in line with the previous studies that the processing ofconcrete
words was left-lateralized and the processing ofconcrete words was bilateral.
Fliessbach, Weis, Klaver, Elger, and Weber (2006) also examined the processing
ofconcrete and abstract words during encoding and retrieval using fMRI. Twenty-one
participants, age range from 19-43 years old, were presented 90 concrete and 90 abstract
simple nouns in the encoding phase. The participants were asked to remember the words
as well as possible. During the recognition phase, participants saw 180 previously
presented stimuli intermixed with 180 distract words, and they had to decide ifthe word
was definitely old, probably old, definitely new, or probably new. It was found that the
concrete words were better memorized than the abstract words, but there were no areas
activated more strongly by concrete words during the encoding phase. However, there
was a strong bilateral activation in the inferior parietal regions for concrete words during
the recognition phase. Contrast to the concrete words, the abstract words elicited a
stronger activation in the left interior-frontal cortex during both encoding and recognition
than the concrete words.
Other than the ERPs and fMRI, Mellet, Tzourio, Denis, and Mazoyer (1998)
investigated the activation of cortical areas through Positron Emission Tomography (PET)
when people listened to the definitions ofconcrete words. Eight participants, age between
20 to 25 years old, were presented to three experimental conditions, including generating
mental imagery while listening to the concrete word definition condition (CONC),
66
listening to the abstract word definition condition (ABST), and a rest condition (REST).
The participants in CONC andABST conditions were instructed to listen to 15 words and
their definitions and then recalled the words they had just heard. The findings indicated
that participants recalled more concrete words than abstract words, and bilateral
activations ofthe inferior temporal and the fusiform gyri were greater for concrete words
than abstract words. It was also found that listening to the abstract word definitions
elicited more intensive and extended activations in the language comprehension network
(inside Broca's Area) compared to listening to the concrete words. Similar results were
also found in a recent study conducted by Liu and colleagues (Liu, Xiang, Wang, Vannest,
Byars, & Rose, 2008). They used the whole head magnetoencephalography (MEG) to
examine how people processed concrete and abstract words. The findings also supported
that concrete and abstract words were processed differently in the brain, and the
difference existed not only in anatomical substrates, but also in the frequency ofneural
activations.
Literature reviewed above either fully or partially supported the dual coding
theory or the concreteness effect. However, the reverse effect or opposite findings were
revealed in other studies. Kiehl and colleagues (Kiehl, Liddle, Smith, Mendrek, Forster,
& Hare, 1999) asked six participants, age 22 to 26 years, to make a decision whether the
word presented was a real English word or not on a pool ofconcrete words, abstract
words, and pseudo words. The fMRI was used to analyze the neural pathways involved in
the processing of concrete and abstract words. The results showed that concrete words
were processed more efficiently and accurately than abstract words. However, the
67
analysis of fMRI indicated that both left and right hemispheres involved in the processing
ofconcrete and abstract words, and the processing of abstract words elicited a stronger
activation in the right hemisphere in comparison to the processing of concrete words. The
plausible explanation ofless accurate and slower responses for abstract words might
result from more extensive semantic processing. The findings ofthis study did not
support the dual coding theory. Moreover, Kiehl and colleagues (1999) claimed that the
right hemisphere was more heavily engaged in the processing ofabstract words.
The findings in Fiebach and Friederici's study (2003) were also inconsistent with
the hypothesis that the right hemisphere involvement was only found for concrete words.
They analyzed how people processed concrete and abstract words through fMRI and
asked participants to judge whether the stimulus was a German word or not. It was found
that the decision time yielded no difference between the concrete and abstract words. In
addition, the concrete words activated left basal temporal cortex, which usually involved
higher-level visual processing and mental imagery. Similarly, abstract words elicited
stronger activities in left inferior frontal region, which was related to retrieval of semantic
information. Thus, this study argued that both concrete and abstract words activated
comparable language networks.
Pexman and colleagues (Pexman, Hargreaves, Edwards, Henry, & Goodyear,
2007) used fMRI to examine 20 healthy adults, mean age at 26.5 years old, during a
semantic categorization task. They found that abstract words elicited more widespread
cortical activations than concrete words, and abstract words activated more strongly in
the network regions associated to semantic processing, including temporal, parietal, and
68
frontal regions. In other words, the findings were not consistent with both the dual coding
theory and context availability theory. It claimed that no cortical areas were activated
more strongly for concrete words than for abstract words. Pexman and colleagues (2007)
argued that the semantic categorization task required more extensive semantic processing
than the lexical decision task; therefore, the representation of abstract concepts could be
fully activated.
Three studies reviewed above mainly argued against the right hemisphere
lateralization in the processing ofconcrete words and claimed that abstract words indeed
activated similar or more widespread neural activities in comparison to concrete words.
Different from these studies, Papagno, Capasso, Zerboni, and Miceli (2007) examined the
concreteness effect on one participant with semantic dementia through a series oftests,
including the lexical decision, word-picture matching, verb naming, and naming objects
from pictures and definitions. It was found that the participant performed normally in
lexical decision, word-picture matching, and verb naming; however, participant was
impaired in naming objects from pictures and from definitions especially for concrete
stimuli. The participant and five neurologically normal controls were asked to point to the
word that was less familiar in meaning from two synonyms and one semantically related
word. They found again that the participant with semantic dementia performed worse
than controls on concrete words but not on abstract words. Since the participant had the
focal atrophy ofthe left temporal lobe, Papagno and colleagues (2007) argued that the
temporal lobe located in the left hemisphere did not involve in the processing ofabstract
words but the processing ofconcrete words.
69
Since the concrete content additionally activates the imaginal system, it was easier
retained and recalled than the abstract content. According to the studies reviewed
previously, the concreteness effect was found in words, sentences, and paragraphs, and
the right hemisphere was activated by the concrete information based on neuroimaging
studies. Although the results from few studies did not support the dual coding theory or
the concreteness effect, most ofthe studies showed that an advantage existed in recalling
and retaining the concrete information in contrast to the abstract information. The main
interest ofthis study was to explore the relationship between mental imagery and idiom
comprehension in adults across the lifespan; therefore, it was essential to review literature
that focused on mental imagery and figurative language, especially idioms.
Mental Imagery and Figurative Language
Up to date, a very limited number ofstudies examined the use ofmental imagery
in figurative language comprehension. Gibb and colleagues (Gibb, Strom, &
Spivey-Knowlton, 1997) attempted to provide the evidence that proverbs and their
figurative meanings were linked by conceptual metaphors and that mental imagery
motivated the figurative meanings ofproverbs. Seventy-two college students were
divided into three experimental conditions, including the proverb, literal alternative, and
figurative definition. Sixteen highly familiar proverbs were selected as the stimuli, and
the participants had to write down their mental images in response to four questions. The
proverbs were formulated for the literal alternative condition and were also modified to
present their figurative meanings for the figurative definition condition. Two yes/no
questions regarding the storability and intentionality and two open-ended questions,
70
causation and manner, were asked to instruct participants to write down detailed mental
images for each expression. The findings showed that the participants provided similar
mental images and consistent knowledge about the events they described in the proverb
condition in contrast to the literal alternative and figurative definition conditions. The
participants provided consistent general images for the proverbs and offered detailed and
consistent responses to questions regarding their mental images. It was proposed that the
conceptual metaphors, not the familiarity, motivated what the proverbs really meant
based on the high consistency ofparticipants' mental imagery and responses to the
questions. Gibbs and colleagues (1997) suggested that mental images revealed the
conceptual knowledge underlying the meaning oflinguistic expressions. Moreover,
figurative meanings ofproverbs were partly activated by conventional images and
conceptual metaphors, which were a significant part ofeveryday thought.
Duthie and colleagues (Duthie, Nippold, Billow, & Mansfield, 2008) also
examined the relationship between mental imagery and proverb understanding in
school-age children, adolescents, and adults. The mental imagery task and proverb
comprehension tasks were administered. Twenty concrete proverbs were used as the
stimuli in both tasks. Participants were asked to write down their images of each proverb
in the mental imagery task and were requested to choose a statement that best expressed
the proverb's meaning from four possible choices. In the comprehension task, a
four-sentence story was created for each proverb with a concluding proverb. The mental
imagery was scored based on a four-point scoring system, including irrelevant, literal
concrete-relevant, literal-metaphorical-relevant, and figurative-metaphorical-relevant.
71
The results indicated that mental imagery for concrete proverbs gradually increased with
age, but the use ofmental imagery was not significantly different between two young
groups. Although adults reported more metaphorical images compared to children or
adolescents, the literal concrete-relevant image was the most common ofthe mental
images reported by participants. The findings indicated that adults have a more
metaphorical understanding ofproverbs than children and adolescents. Although some of
the proverbs evoked stronger images than others, not all concrete proverbs were
imagineable. In addition, the relationship between mental imagery and comprehension
was significantly correlated for the children but not for the adolescents and adults.
According to the studies in proverb comprehension and mental imagery,
metaphorical images existed and were activated during the proverb comprehension
process. In addition to proverbs, mental image was also found related to idiom
comprehension. Forming the arbitrary link between idioms and their idiomatic meanings
was important in learning the meanings ofidioms. Therefore, Gibbs and O'Brien (1990)
investigated ifpeople have the tacit knowledge ofthe metaphorical basis about idioms. In
their first study, 24 undergraduates were asked to define 25 idioms and then to form a
mental image for the expression. Idioms were categorized into five different meanings,
including anger, exerting control, secretiveness, insanity, and revelation. Six probe
questions were asked to reveal participants' mental images, containing causation,
intentionality, consequence, negative consequence, and reversibility after they formed a
mental image. The findings indicated that participants had similar schemas underlying
their mental images with similar idiomatic interpretations. In addition, their responses to
72
the probe questions regarding the mental images were highly consistent and detailed. In
general, the highly consistent mental images ofidioms suggested the active role of
conceptual metaphors in idiom comprehension and that conceptual metaphors motivated
the link between idioms and their nonliteral meanings. Therefore, people know what the
idioms really mean because ofthe influence that conceptual metaphors have on their
mental images.
In their second study, the figurative definitions ofeach idiom were used as stimuli
and the same six probe questions were asked. It was revealed that participants' mental
images for the paraphrases ofthe idioms were inconsistent and varied compared to the
mental images formed in the first study solely for the idioms. Therefore, it was suggested
that mental images for idioms were not constructed solely on the basis ofthese phrases'
figurative meanings. Moreover, people's mental images ofidioms were constrained by
the conceptual metaphors that partially linked the idiom and its figurative meaning. In the
third study, Gibbs and O'Brien (1990) compared images formed for literal phrases to
those formed for idiomatic expressions. It was found that mental images and responses to
the probe questions were less consistent than those for idioms. It was suggested that the
difference in mental images between idioms and literal phrases resulted from the
constraining influence ofconceptual metaphors. These conceptual metaphors provided
tighter constraints in forming mental images for idioms and further resulted in consistent
knowledge oftheir mental images and figurative expressions. However, it was unlikely
that people would construct mental images for idioms during the normal processing.
Given that the rapid processing for idioms, the familiarity with idioms allowed people to
73
recognize the figurative meanings without awareness of conceptual metaphors motivating
the meanings ofthese expressions.
Cacciari and Glucksberg (1995) proposed two questions regarding the role of
conceptual metaphors in the Gibbs and O'Brien's 1990 study. First, it was confounded
when the idiom can be interpreted literally, such as kick the bucket. Visual images were
highly possible to reflect literal meanings rather than figurative meaning. Secondly,
according to the dual coding theory, concrete concepts would form mental images more
easily than abstract concepts. Therefore, it was highly unlikely to automatically form a
mental image for the idiomatic meaning which was more abstract than its literal meaning.
Thirdly, based on the Stroop Test, people automatically generated the meanings ofwords,
but the literal meaning ofwords was not inhibited. Therefore, Gibbs and O'Brien's
hypothesis was questionable while it was used to explain the automatic activation of
lexical meanings. In order to determine which interpretations of idioms, concrete-literal
or abstract-figurative, were reflected in people's mental images in processing idioms, 20
undergraduates participated in an imagery production experiment. Twenty Italian idioms
that varied in the transparency and familiarity were used as stimuli. Participants were
asked to provide a paraphrase ofeach idiom, to form mental images for the idiom, and to
describe it in details. It was found that the images reflected the concrete actions and
events that the idiom indicated. The findings were in line with the concreteness effect that
the concrete literal meaning of an idiom generated a mental image more easily than an
abstract-figurative meaning.
74
In the second study of Cacciari and Glucksberg (2005), the main research
question was whether mental images facilitated or interfered with idiom comprehension,
and 96 undergraduates participated in the second study. The materials were 20 idioms
used in the first study, 20 literal-concrete sentences, and 20 literal-abstract sentences.
Each sentence consisted ofone matching and one mismatching paraphrase. The
participants were divided into two groups, non-imagery instruction group and imagery
instruction group, and were asked to make a judgment ifthe paraphrase matched the
target sentence. The difference between two groups was that the imagery group had to
generate detailed mental images ofthe target sentence before the paraphrases showed on
the screen. The reading time for three types of sentences in the non-imagery group was
similar; however, abstract sentences took longer to respond in the imagery group.
Moreover, the image generation increased the verification time for more difficult
sentences and unfamiliar idioms. Regarding the error rates, participants made fewer
errors in verifying idioms than verifying concrete or abstract sentences. In general,
images that participants generated for idiom strings reflected literal rather than figurative
meanings. Additionally, mental images did not facilitate idiom comprehension but
prolonged the comprehension time. The participants also did not use mental imagery to
perform the paraphrase verification task no matter ifit was a concrete-literal,
abstract-figurative, or idiomatic. Therefore, Cacciari and Glucksberg (2005) argued that
images did not provide the evidence for conceptual metaphors in the idiom
comprehension process, which was inconsistent with the Gibbs and O'Brien 1990 study.
------ -----
75
In order to determine whether people access pre-existing conceptual metaphors
during the online idiom processing, 34 undergraduates participated in the study of Gibbs,
Bogdanovich, Sykes, and Barr (1997). Fifteen briefstories and 15 fillers were randomly
divided into six test lists. Each list contained five stories ending with idioms, five stories
ending with literal paraphrases, five stories ending with control phrase, and 15 filler
stories. Three ofthe five phrases in each list ending with related targets, unrelated targets,
or opposite types oftargets. The participants were instructed to make a lexical decision to
the letter strings, which appeared after the final phrase, as quickly as possible. The
[mdings showed that participants took the similar time to process idioms and literal
paraphrases, but longer to comprehend control sentences. Additionally, participants
responded to the related targets faster than the unrelated targets after reading the idioms,
but not after reading the literal paraphrases or control sentences. The findings supported
that conceptual metaphors were accessed during the idiom comprehension process.
Moreover, people accessed figurative meanings and literal meanings at the same time or
even before they processed the literal meanings. In their second study, 36 undergraduates
read 16 stories describing usual events. Each story ended in one ofthe two different
idiomatic phrases expressing the same figurative meaning, including the consistent idiom
motivated by the conceptual metaphor that reflected the idiom (anger is heatedfluid in a
container) and the inconsistent idiom motivated by a different conceptual metaphor
(anger is animal behavior). Similar to the first study, the final phrases ofthe stories were
followed by either related or unrelated targets. It was found that the participants
responded faster to the targets after reading consistent idioms than inconsistent ones.
76
Therefore, people did not access the same metaphorical knowledge when reading idioms.
It was concluded that the metaphorical knowledge may playa role in people's immediate
understanding ofcertain idioms.
Bortfeld (2002) conducted a series ofstudies to investigate mental imagery in
idiom comprehension in native and nonnative English speakers. In the first experiment,
25 native English speakers and 25 native Latvian speakers were recruited. Twenty five
idioms under five concepts, including anger, revelation, secretiveness, insanity, and
control, were used. Six probe questions were asked to reveal participants' mental images
including causation, intentionality, manner, consequence, negative consequence, and
reversibility. Participants were first asked to generate a sentence for each idiom using the
phrase figuratively and then to form a mental image based on the literal meaning ofthe
idiom. After the image generation, the participants were again asked to produce sentences
for the idioms using the phase figuratively. The findings showed that both native and
non-native speakers formed images reflecting the combination offigurative and literal
meanings and had difficulty separating the phrases' literal meanings from the figurative
meanings. Thirty undergraduate students rated the naturalness ofeach sentence that was
produced before and after the participants generated the mental image for each idiom.
The findings indicated that the naturalness ofsentences increased after participants
generated mental images.
In the second study, 25 native Chinese speakers were included instead ofLatvin
speakers. The stimuli were twenty-five idioms and questionnaires used in the fist
experiment. It was revealed that both English and Chinese speakers reported highly
77
consistent responses about their images ofEnglish idioms. Therefore, the analysis of
surface structures helped non-native speakers generate figurative meanings ofidioms.
General speaking, people were able to generate figurative meanings through the analysis
ofthe surface structure ofidiomatic phrases without additional supporting contexts.
Given that highly similar images generated from native and non-native speakers, some
specific conceptual structures might underlie some idiomatic phrases. It was suggested
that our understanding ofidioms must first map words to conceptual structures that are
meaningful to the speakers, and then the conceptual mapping would become well
established.
It was suggested that transparent idioms may be easier comprehend than opaque
ones because ofthe involvement ofmental imagery and the concreteness effect (Cacciari
& Glucksberg, 1995). Nippold and Duthie (2003) examined the use ofmental imagery in
idiom comprehension on 40 school-age children and 40 adults. Twenty high familiarity
idioms, including 10 transparent and 10 opaque, were used as the experimental items in
two tasks, mental imagery task and idiom comprehension task. The participants were
asked to write down their images when someone uses each ofidioms listed in the booklet.
After the mental imagery task, participants were required to interpret the 20 idioms by
choosing the correct answer from four possible responses. Each idiom was embedded in a
four-sentence story context with the final sentence containing the idiom used by a
character in the story to facilitate comprehension. The results supported that the
transparent idioms were easier to understand than the opaque ones, and adults
outperformed school-age children on the idiom comprehension task. However, even
78
adults did not master all the idioms used in this study. Regarding the mental imagery task,
it was revealed that for both groups, mental imagery was associated with the
comprehension ofboth transparent and opaque idioms. Moreover, both idiom
comprehension and mental imagery improved with increasing age. It was suggested that
mental imagery would be automatically created once the figurative meaning ofan idiom
was learned. Although images triggered by opaque idioms were less sophisticated than
the ones by transparent idioms, some images ofopaque idioms reflected the deep
figurative understanding. Therefore, mental imagery may serve as an indicator ofdeep
understanding offigurative meanings in idiom comprehension. Additionally, the
knowledge of idioms is expected to expand, and deepen and mental imagery is expected
to become more figurative with advancing age.
Summary ofDual Coding Theory
The dual coding theory was commonly used to examine the difference in
comprehension and recall between concrete and abstract words, sentences, concepts, and
content. It was found that concrete concepts and content were easily remembered,
recalled, and comprehended in comparison with abstract ones. This phenomenon was
often referred to the concreteness effect. The concrete concept or content elicited
additional right hemisphere activations, where the imaginal system was located.
Therefore, it provided the sensational experiences for the concrete information and
further resulted in better understanding and recall. The dual coding theory and the
concreteness effect were evident from behavioral and neuroimaging studies. Although not
all studies yielded the same results because ofdifferent stimuli and methods used across
79
the studies, the dual coding theory provided the strong ground knowledge for the
differences observed between concrete and abstract words, sentences, content, and
concepts.
Gibbs and O'Brien (1990) pointed out the active role ofconceptual metaphors in
idiom comprehension and further suggested that the conceptual metaphors motivated the
link between idioms and their nonliteral meanings. Additionally, researchers believed that
the conceptual metaphors motivated what the proverbs/idioms really meant according to
the high consistency ofpeople's mental imagery (Bortfeld, 2002; Gibb, Strom, &
Spivey-Knowlton, 1997). Although a few studies argued that mental imagery did not
facilitate the understanding but prolonged the processing time, the highly consistent
mental images reported by people implied some specific structures underlying the
understanding offigurative language. Duthie and colleagues (2008) argued that some
proverbs evoked stronger images than others, bur not all concrete proverbs were
imagineable. Therefore, how mental imagery was formed differently for transparent and
opaque idioms remained questionable.
Most ofthe studies reviewed in idiom comprehension and the dual coding theory
were limited to college student populations although a few ofthe studies examined
children and adolescents. However, figurative language was rarely studied in older people.
With the aging populations growing, it is important to understand the language
performance in aging populations since cognitive decline is commonly observed in the
elderly and their possible language decline may be observed. Since the focus ofthis study
80
is to examine idiom comprehension in adults across the lifespan, it is essential to have
comprehensive knowledge regarding language in aging populations.
Language and Aging
It is commonly assumed that once language is acquired, it never goes away except
ifthe language system is impaired due to brain damage such as stroke, traumatic brain
injury, or dementia. However, studies showed that language performance, both
production and comprehension, declined with advancing age due to decrement in
processing speed (KwongSee & Ryan, 1995; Waters & Caplan, 2005), working memory
(Carpenter, Miyake & Just, 1994; Gunter, Jackson, & Mulder, 1995; KwongSee & Ryan,
1995; Waters & Caplan, 2005), verbal working memory (DeDe, Caplan, Kemtes, &
Water, 2004), and inhibitory efficiency (Burke, 1997; Hasher & Zacks, 1988;KwongSee
& Ryan, 1995). Obler and Albert (1989) studied naming, discourse, comprehension,
automatic speech, and metalinguistic abilities on 150 adults, ages between 30-79 years.
They pointed out that naming ability decreased beginning in the 70s, but deficits in
comprehension started as early as the 50s. However, increased elaboration in discourse
was observed, and no change in metalinguistic tasks, such as word list generation and
interpretation ofproverbs and idioms, was found in the older group. Emery (1986) also
pointed out that there was a direct relationship between linguistic complexity and
performance deficits in elderly, and the semantic processing in the sentence level and the
syntactic processing were significantly impaired in elders. In addition, it was evident that
the greater the linguistic complexity, the more difficulty in language processing was
81
found in normal older people. Furthermore, Emery (1986) claimed that linguistic
declined in normal elders was not a random occurrence, but an orderly and predictable
sequence.
Although some researchers suggested that language deterioration was associated
to normal aging (Emery, 1986; Olber & Albert, 1989), not all aspects oflanguage
declined with advancing age. Understanding how age affected language performance in
normal aging was essential to further understand the relationship between more
complicated figurative language and normal aging. Since this study elicited both
receptive and expressive language performance, it was needed to review language
production and comprehension in normal aging populations.
Language Production andAging
It is suggested that there are different levels ofdeterioration in language while
getting old. It is evident that difficulties in naming, reduced syntactic complexity, and
decrement in verbal fluency are commonly observed during the age of70 (Ardila &
Rosselli, 1996). Elders might have difficulties accessing the appropriate lexical items,
producing syntactically complex sentences, and integrating all story elements in a
complex discourse structure (Juncos-Rabadan & Iglesias, 1994). Burke and MacKay
(1997) also stated that spoken and written language production consistently showed
age-related declines, and common deterioration observed in elders' language production
included word findingllexical retrieval difficulties, tip ofthe tongue (TOT), disfluency in
speech and ambiguous references. Ulatowska, Cannito, Hayashi, and Fleming (1985)
examined how information was told, retold, and summarized in different discourses such
82
as narrative, procedure discourse, and conversation. They found that there was an
age-related decrement for the complex processing ofdiscourse information in the later
years oflife, and the decrements might result from multiple causative factors rather than
just cognitive impairments with advancing age.
In order to investigate the linguistic patterns in normal elderly and the differences
oflinguistic function between normal elders and individuals with AD, twenty middle-age
adults, ages 30 to 42 years, twenty normal elders, ages 75 to 93 years, and 20 individuals
with AD participated in Emery's study (1986). The Token Test, The Emery Test for
Syntactic Complexity, The Chomsky Test ofSyntax, and The Boston Diagnostic Aphasia
Examination were used as linguistic measures. The results pointed out that there was a
direct relationship between language deficits and age and between language deficits and
linguistic complexity. In other words, the more complex the linguistic form was and the
later the development ofthe linguistic form, the quicker the deterioration ofthe linguistic
form. It was found that performance on the phonology tasks showed no difference
between middle-age adults and normal elders except the speed ofresponse. Additionally,
there was no difference found between middle-age adults and normal elders in
vocabulary ability. However, normal elders performed worse than middle-age control
group in morphological and syntactic level. When the level oflinguistic complexity
increased, the normal elders showed more pronounced decrement in language ability.
Therefore, normal elders performed as well as middle age adults did in lexicon, but
performed worse in the word-internal morphological processing, and did even worse on
the sentence level task since it was more abstract and complex.
83
Eight participants in each age category, 50-59, 60-69, 70-79, and 80-89, were
recruited in the Kynette and Kemper study (1986). Each participant provided a 20 minute
personal story about their lives. The number ofdifferent simple syntactic structures, the
number ofcomplex syntactic structures, the percentage ofsimple structures correctly
produced, the number of different verb tense, and the percentage ofverb tenses correctly
produced were measured in order to detect the difference in syntax, tense, form class,
lexical use, and disfluency across four age groups. The finding revealed that there was a
reduction in accuracy of syntactic structures, verb tense, and form classes from 50 to 90
years; however, disfluency measured by fillers and fragments did not increase with age.
Additionally, 50 and 60 year-olds used more complex structures with multiple
embeddings, but 70 and 80 year-olds made more errors in the use of simple syntactic
structures, correct past tense inflections, subject-verb number agreement, articles, and
possessive markers.
It was suggested that elders had difficulties producing syntactically complex
sentences because it would impose high memory demands. In order to further provide
evidence for this hypothesis, Kemper (1986) studied the imitation ofsentences on 16
healthy elders, ages between 70-89 years and 16 adults, ages between 30-49 years. A set
of 32 sentences varying from grammatical correctness, length, position, and type of
embedded clause were used as materials. The 32 sentences were read to the participants,
and then participants were asked to repeat each sentence. The findings indicated that
elders had impairment in syntactic processing and also had difficulties repeating long
sentences with embedded clauses, which might result from memory limitations.
84
A group ofnine normal middle-aged adults (50-60 years), a group of 11 normal
elders (80-90 years), and a group of 10 elders (80-90 years) with mild cognitive
impairments were compared on three discourse tasks including a retold narrative task, a
retold instructional task, and a personal narrative task in the Richardson study (1990).
The results revealed that normal elder adults and elderly adults with mild CI had more
difficulties in the retold instructional task than the retold narrative task. Additionally,
middle-aged adults included more essential procedure steps in the instructional task than
normal elders and elders with mild CI. A decrease in the percentage ofnouns used, an
increase in the use of ambiguous pronouns, and the lack of indefinite reference were
found in normal elders and elders with mild CI across three discourse tasks.
Juncos-Rabadan and Iglesias (1994) investigated ifthe reduced language abilities
were associated with impaired attention in older adults across 14 languages. Sixty normal
older adults, ages 50 to 91 years, in 14 languages, a total of 840 participants were further
divided into three age groups, 50-59 years, 60-69 years, and 70-91 years and three
educational levels, 0-4 years, 5-10 years, and beyond 11 years. Each subject was the
native speaker ofthe language and was tested on different linguistic levels including
morphology, syntax, lexicon, and language skills such as verbal fluency, comprehension,
reading, and writing. The results indicated that aging was associated with linguistic
deterioration in syntax, lexicon, morphology, phonology, and semantics. It seemed that
semantic and conceptual knowledge, organization of semantics, and phonological lexicon
were preserved; however, the deterioration was observed in synonyms, antonyms, and
semantic opposites. In addition, the difficulty shown in the definition, comprehension,
85
repeating complex sentences, judgment, and syntax, morphology, and discourse subtests
suggested that the decline resulted from the impairment in working memory and
executive function.
Thirty young adults, 18 to 28 years ofage, and thirty older adults, ages 70 to 80
years old, participated in Kemper and colleagues' study (Kemper, Herman, & Lian, 2003).
Fifty-four experimental trials containing 36 different two-word combinations, 36
three-word combinations, and 36 four-word combinations were administered. Participants
were asked to produce sentences by using the words presented on the computer screen.
The results suggested that memory load affected sentence planning in older adults since
the stimulus disappeared as soon as the participants began to speak. Moreover, older
adults had difficulties retaining the words and tended to produce less complex, shorter,
and less informative sentences when the numbers ofwords increased..
Kemper, Herman, and Liu (2004) attempted to examine sentence production by
young and older adults in controlled contexts, and twenty-four young adults, 18 to 28
years old, and 24 older adults, 70 to 80 years old, were recruited. Thirty-six left- or right 
branching stems were used as the stimulus, and the participants were asked to complete
the sentences with either right- or left-branching stems. In addition to the stems, 27 nouns
referring human characters and 27 nouns referring locations were provided for
participants to complete the sentences. Sentence length, grammatical complexity, and
propositional density were measured. The findings showed that left-branching which
required the participants to produce the main clauses in order to complete the sentences
imposed greater burden on working memory. Therefore, older adults made more errors,
86
nontluent responses, and fragments in contrast to younger adults while producing
responses to the left-branching stems compared to those to the right-branching stems.
Different from examining spoken language in older adults, Kemper, Greiner, Marquis,
Prenovost, and Mitzner (2001) analyzed autobiographies collected from 90 participants
who participated in the Nun study on grammatical complexity and idea density. The
grammatical complexity was measured based on the clauses embedded in one sentence,
and the idea density was assessed through the number ofpropositions per 10 words. It
was found that grammatical complexity and idea density declined with age. Additionally,
participants, especially above age of61, produced less complex sentences and used more
words to express the same idea. Furthermore, later autobiographies were vaguer and
consisted ofmore repetitions than early written ones.
Different from the studies reviewed above, Glosser and Deser (1992) attempted to
investigate the age changes in macrolinguistic and microlinguistic aspects ofdiscourse
production. They informally interviewed 14 middle-age adults, ages 43-61 years, and 13
normal elder, ages 67-88 years. Participants were interviewed individually for 10 to 20
minutes and were asked to describe their families and work experience. Microlinguistic
language including syntactic omissions, syntactically complete sentences, subordination,
verbal paraphasias, indefinite terms, and macrolinguistic language such as thematic
coherence and discourse cohesion were measured. It was found that there was no
difference on microlinguistic language measures between middle-aged adults and normal
elders. However, macrolinguistic language ability was found to be impaired in normal
elders since it required linguistic and nonlinguistic knowledge such as working memory,
87
long-term memory, and executive function. Normal elders in this study had difficulties
integrating and organizing information into a theme- or topic-coherent discourse. Glosser
and Deser (1992) claimed that normal elders' discourse was similar to individuals with
probable Alzheimer's disease who showed significant impairments in nonlinguistic
language, memory, and executive function.
Language Comprehension andAging
The goal of comprehension is to obtain an understanding of described states of
affairs and then the information needs to be stored for later use (Radvansky, 1999). Since
working memory, inhibitory efficiency, and processing speed appeared to be affected by
aging (Carpenter, Miyake, & Just, 1994; Gunter, Jackson, & Mulder, 1995; KwongSee &
Ryan, 1995; Waters & Caplan, 2005), comprehension ofspoken language might be also
influenced by the deficits ofthese cognitive functions. According to the study of DeDe,
Caplan, Kemtes, and Water (2004), an effect ofage mediated through verbal working
memory was the best-fit model for sentence and text comprehension. Obler, Fein,
Nicholas and Albert (1991) also pointed out three possible factors contributed to the
comprehension deficits observed in the normal aging populations including impairment
in syntactic processing, shift in strategy, and age-related deficits in extralinguistic
cognitive functions.
Cohen (1979) asked twenty participants in each group, highly educated old people
(OHE), highly educated young people (YHE), old people with low education level (OLE),
and young people with low education level (YLE), to listen to 16 short messages and to
answer two questions according to the content. Half ofthe messages were classified as
88
simple, and the other half were complex. The messages were spoken in two different
rates, slow and fast. Participants were asked one verbatim question that required
reproduction ofthe presented facts and one inference question that participants had to
draw inference from the presented facts. The results revealed that age-related deficits
existed in inferential stage but not for the verbatim questions. The plausible reasons
might be the decrease ofprocessing speed and memory loss. The same participants were
also asked to judge 16 messages which contained one anomaly and one making sense
version. The findings showed that older groups performed worse than younger groups,
and the mistakes were due to assessing incorrect prior knowledge. The third experiment
carried out to compare the recall ability between old and young groups with high and low
educational level. Same participants recalled a 300-word story, and the recalls were
scored based on the total number ofpropositions, summary propositions which
represented the gist ofthe story, and 24 modifiers including comparatives and quantifiers,
temporal modifier, locatives, and logical connectives. It was revealed that old groups
performed worse than young groups did because ofthe heavy memory load on the recall
task. In general, elders maintained surface comprehension and had deficits in language
comprehension involving integration, inference, and construction.
Sentence comprehension abilities in adults, ages 18 to 80 years, were compared in
the Feier and Gerstman study (1980). Fifteen participants in each age group, 18 to 25, 52
to 58, 63 to 69, and 74 to 80 years were tested on vocabulary and digit span and were
instructed to act out each stimulus sentence using small animals or human figures. Four
types ofsentences including self-embedded subject relative, self-embedded object
89
relative, right-branching subject relative, and right branching object relative, and three
conjoined sentences with no subordinate clause were auditorily presented. It was
suggested that the ability to comprehend sentences with subordinate clauses was stable
until 60s, but the decline was observed beginning in the 60s. Age difference distinguished
the performance ofthe two oldest groups from each other and from the two youngest
groups. The oldest group made errors more frequently in their enactments and also made
errors of a more serious nature compared to younger adults.
Davis and Ball (1989) investigated whether age-related changes occurred in
comprehension of complex sentences. Fifteen participants in each age group, 25 to 35
years, 40 to 50 years, 53 to 60 years, 61 to 70 years, and 71 to 79 years, were testedon
vocabulary and digit span, and were presented 24 sentences that emphasized either
semantic or syntactic constrains in interpretations. Three constraint conditions were
created. In the semantic condition, thematic roles ofnouns were consistent with real
world probabilities. In the syntax-I condition (implausible sentences), semantic cues
based on the world knowledge were either weakened or confused. In the syntax-PR
condition, sentences were plausible and reversible. Three types of questions were asked
to test the comprehension ofthematic roles in each sentence, and participants were asked
to choose the best answer from three choices after reading the question. The findings
indicated that normal aging affected comprehension ofcomplex sentences and the decline
began after age of60. Additionally, older adults did not process the syntactic component
effectively when sentences conveyed implausible information.
90
In order to detennine which factors contributed to comprehension difficulty due to
normal aging, 66 females were grouped into four age groups, 30-39 years, 50-59 years,
60-69 years, and 70-79 years and were presented 96 sentence-question pairs in the Obler
et al. study (Obler, Fein, Nicholas, & Albert, 1991). Six syntactic types ofsentences,
active, passive, single negative, double negative, double embedded, and comparative,
consisted ofsentence-question pairs. Halfofthe pairs in each type were semantically
plausible and the other half were semantically implausible. One question starting with
"was" to elicit yes/no answer was followed by each sentence. The results showed that
older participants took longer reaction time in the comprehension task, and the decline
rates and patterns of decline varied depending on syntactic structures. Moreover,
implausible and harder types ofsentences, such as double embedded, double negative,
and passive, elicited more errors in the elderly and showed more rapid decline.
Elders might have deficits in processing meaningful and implicit infonnation
without showing difficulty in processing the explicitly stated meaning. Therefore,
Belmore (1981) studied how younger and older adults processed implicit and explicit
meaning using a sentence verification task. Sixteen younger adults, ages 17 to 21 years,
and 16 older adults, ages 58 to 74 years, participated in the study. Thirty two short
paragraphs were used as stimuli. Each came with one sentence that paraphrased the
stimulus paragraph and one sentence that represented a plausible influence from the
paragraph. Halfofthe stimuli provided true paraphrase and inference, and the other half
had false paraphrase and inference. An unexpected 20 minute delayed verification test
without paragraphs was given. They found that participants processed inferences slower
91
than paraphrases on the immediate and delayed tests. The findings showed that older
adults comprehended the meaningful language slower and less accurately than younger
adults; however, younger adults outperformed older adults only on the delayed test not
the immediate test. Although the latency data indicated that older adults' language
processing speed was slow, it was not related to explicit or implicit meaning.
Similar to the language production tasks, the task types in language
comprehension also significantly influenced participants' performance. Burke and
MacKay (1997) argued that although the processing ofwords meaning maintained
constantly in elders when the test measured on-line processing such as semantic priming,
off-line processing might show different results because ofexamining ifpeople
remembered the sentences or paragraphs presented earlier. Therefore, when older people
required to encode new information, they might show age-linked deficits. Radvansky
(1999) also claimed that when the comprehension task taped on lower level ofprocessing,
such as remembering the information they encountered earlier, older people might show
signs ofdifficulties. However, ifthe tasks assessed higher level of comprehension, older
adults might do as well as young adults.
Factors Affect Language Performance in Normal Aging
Although studies reviewed in the previous section mainly focused on the language
deterioration with advancing age, people did not necessarily show language declines as
they became older. Using different types oftasks might result in inconclusive results.
Other factors, such as education and gender, might also contribute to different language
performance in aging populations. Cohen (1979) reported that elders with low education
92
levels performed worse on the comprehension tasks than did elders with higher education
levels; therefore, the decline could be compensated by the higher educational level and
continuing mental activity. Cohen (1979) stated that some ofthe tests commonly used to
detect language deterioration in aging such as definitions ofsingle words or word pairs
might not be sensitive enough to reveal the decline. In addition, Mackenzie (2000)
claimed that using specific linguistic tasks to examine language ability in older people
might not represent a true language reduction because ofthe unestablished relationship
between the tests and real life communication. Some factors such as anxiety, inability to
see and hear the stimuli clearly, and task unfamiliarity might contribute to the age-related
decrement in older adults' language ability. Therefore, it is critical to take those factors
into account in order to truly represent language performance in older people.
Kemper and colleagues (Kemper, Kynette, Rash, O'Brien, & Sprott, 1989)
investigated the relationship between individual differences such as education and
memory ability, and ifthe syntactic complexity would be affected by different genres and
modalities. Thirty young adults, ages 18 to 28 years, 37 adults, ages 60 to 69 years, 26
adults, ages 70-79 years, and 15 adults, ages 80-92 years provided one oral narrative
language sample, one oral expository language sample, and one written narrative
language sample during the interview. Language samples were analyzed on length, mean
length per utterance (MLU), mean clauses per utterance (MCU), clause structures, and
fluency. Additionally, 18 judges rated the language samples according to organization,
clarity, and interestingness ofthe statements. The findings showed that the use of
left-branching clauses showed age-related decrement for all three language samples.
93
Moreover, elderly adults produced fewer syntactical complex sentences to compensate
the loss ofmemory capacity. It was noteworthy that elders' essays were rated more
interesting and clearer than those ofthe young adults.
Mackenzie (2000) conducted a 10-minute conversation and a picture description
task on 60 participants in each ofthe age group, ages 40 to 59 years, 60 t074 years, and
75 to 88 years. Each group was further divided into three subgroups according to
participant's educational level. Conversational initiation, turn-taking, verbosity, topic
maintenance, and referencing were measured. The number ofrelevant content units, the
number ofwords, efficiency ofimparting content, and occurrences of extraneous
materials were examined in the "cookie theft" picture description task. The results
showed that older participants failed to maintain topic, and their conversation consisted
ofpoor turn-taking, unclear referencing, and verbosity. In the picture description task,
older adults were less efficient and took longer time to transmit information than younger
people. In addition, education rather than age appeared to be a significant factor in the
picture description task, since the two-minute picture description task required less
cognitive demand compared to conversation.
Petros and colleagues (Petros, Norgaard, Olson, & Tabor, 1989) attempted to
examine ifindividuals' verbal ability and text genre difference had an influence on age
difference in prose memory. Eleven younger adults with high verbal ability, ages 18 to 27
years, 19 younger adults with low-verbal ability, ages 18 to 30 years, 16 older adults with
high-verbal ability, ages 60 to 79 years, and 13 older adults with lower-verbal ability,
ages 65 to 84 years were asked to listen to three narratives and three expository stories.
94
Stories were presented at three different rates and were categorized into three levels
according to the importance to the theme ofthe passage. The findings showed that older
adults recalled fewer details than younger adults, and individuals with high-verbal ability
recalled more information than those with low-verbal ability. Additionally, narratives
were better remembered than expository passages. Moreover, participants recalled fewer
details when the passage was presented at a faster rate and recalled more when the
importance level increased.
In the study ofJuncos-Rabadan and Iglesias (1994), they pointed out that
individuals with high educational level performed better on 30 language subtests
including morphology, syntax, lexicon, and language skills such as verbal fluency,
comprehension, reading, and writing than those with low education level regardless of
age. Juncos-Rabadan (1996) examined the effect ofage, gender, education, and culture on
telling narratives in normal elders. The total of94 middle-age adults, mean age of53.93
and 90 normal elders, ages 70 to 91 years, were asked to tell a story based on a
six-picture "Nest story" from the Bilingual Aphasia Test (BAT). The findings pointed out
that normal elders included fewer sense units, used simpler sentence structures, fewer
numbers ofcohesion links, but more descriptive and tangential sentences and place deixis
than did middle-aged adults, which was a sign ofdecrement in narrative competence. In
addition, individuals with higher education performed better in the story telling task and
used the story structure better than those with lower education.
Aging affected differently on different language functions. In order to find out
how education and gender playa role in language during normal aging, Ardila and
95
Rosselli (1996) conducted a study on 108 normal participants, ages 16 to 65 years, with
different education levels. Participants were divided based on age, 16-30 years, 31-50
years, and 51-65 years, and education levels, 3-7 years, 8-12 years, and beyond 12 years.
The cookie theft task from the Boston Diagnostic Aphasia Examination (BDAE) was
used as the stimuli for the spontaneous oral language production. Language samples were
analyzed based on nouns, verbs, adjectives, and grammatical connectors, such as
conjunctions and pronouns. The findings showed that there was a reduction in words used
in the spontaneous production across age groups, especially in the ages 31 to 50 years. In
addition, spontaneous language was correlated to the education level and gender. People
with higher education level used more words to describe the picture than did the people
with lower education level. As regard to gender, females used greater number ofwords
than did males even at the ages 51-65 years.
Harris, Rogers, and Qualls (1998) attempted to examine the influence oftext
genre, cognitive processing requirement, and repeated reading on written language
comprehension in younger and older adults. Twenty-seven younger adults, ages 18 to 27
years, and 27 older adults, ages 64 to 80 years, were assessed on working memory,
reading comprehension, and exposure to print. They were presented three expository
passages, three narratives, and three procedural passages that each was followed by 12
verification statements. The results pointed out that older adults performed as well as
younger adults did on the discourse comprehension task. It was worth noting that the
age-related decline in language processing was alleviated in older adults with higher level
of education and verbal ability. In addition, the text genre had an influence on reading
96
time; expository passages were read faster than narratives and procedural passages.
Therefore, the findings ofthis study indicated that certain text processing skills were
persevered across the adult lifespan.
Language Difficulties in People with Dementia
In addition to examining language performance in the normal aging populations,
findings drawn from people with Dementia and Alzheimer's disease (AD) provided
valuable insight ofhow and what aspects of language decline along with brain
degeneration. Kemper and Zelinski (1994) compared the language changes in normal
elders to those with dementia through word finding, syntax, and discourse. They found
that people with normal aging had mild word finding difficulties due to decreased lexical
retrieval ability. However, word finding was a prominent symptom in people with
dementia, which occurred in every sentence. In terms ofsyntax, syntactic knowledge was
preserved in the normal aging populations and people with dementia, but the performance
declined because ofthe decrement in attention and memory. Normal elders were able to
structure conversations and stories coherently and interestingly; however, discourses
produced by people with dementia were confusing and incoherent.
Twenty-eight normal elders, mean age of 73 and 28 individuals with dementia,
mean age of 80, took the cognition, memory/learning, and language/speech tasks in the
Bayles study (1982). Five language tasks including naming, verbal description, sentence
correction, sentence disambiguation, and story-retelling were used to evaluate
participant's language ability. It was found that individuals with dementia did not
recognize or correct the errors they made during the tasks. Additionally, individuals with
97
senile dementia often produced semantically inappropriate sentences. The results pointed
out that syntax and phonology were not disrupted as semantics was, and there was no
evidence showing that there was a naming deficit in individuals with dementia. However,
individuals with dementia often produced empty and irrelevant utterances, and the
frequency ofproducing irrelevant utterances increased as the severity ofdementia
progressed. It was evident that senile dementia affected the language ability profoundly
with uneven rates in different components ofthe language system.
Lyons and colleagues (Lyons, Kemper, LaBarge, Ferraro, Balota, & Storandt,
1994) studied the relationship between the severity ofdementia and syntactic complexity
on 117 individuals with different severity ofdementia. Participants' characteristics such
as education, cognitive, and linguistic abilities were obtained, and a language sample was
collected individually during a 2-hour interview. It was found that sentence length,
grammatical complexity, propositional content, and verbal fluency were reduced with the
increased severity of dementia. Individuals with mild dementia produced simpler, shorter
and more incomplete sentences compared to individuals without dementia. Overall, there
was a decline in the complexity and length of sentence produced by people with dementia
possibly resulting from AD. Because ofthe cognitive decline in people with AD, they
were less able to produce longer, complex, and multi-clause embedded sentences than did
people without dementia.
It is difficult to identify the language changes due to normal aging or early stage
ofAlzheimer's disease; therefore, Chapman and colleagues (Chapman, Ulatowska, King,
& Johnson, 1995) investigated discourse difference on 12 individuals with early
98
Alzheimer's disease (AD), ages 51 to 76 years, 12 normal elderly adults (OE), ages 80 to
92 years, and 12 normal adults with matched ages to the AD group as the control group
(NC). Three pictures containing familiar life situations such as a son going to a college
were used as stimulus to elicit language samples. Participants were asked to create a story
about the picture but not to describe the picture. The findings pointed out that individuals
with early AD did significantly worse in discourse coherence. Additionally, participants
with early AD had difficulties transforming the description information into a narrative
form while participants in the OE and NC groups had little difficulties with it. In addition,
discourse coherence was impaired in individuals with early AD, although the results were
not consistent for all individuals with early AD. Furthermore, the ability to provide
information according to the frame and to transform information into a narrative form
was preserved in normal aging individuals while it was impaired in people with early AD.
Kemper and colleagues (Kemper, Thompson, & Marquis, 2001) compared the
grammatical complexity (D-level) and propositional content (P-level) in oral language
samples produced by 30 normal elders, ages 65 to 75 years, and 30 elders with dementia.
Language samples were collected annually up to 15 years for normal older adults and
were collected every 6-month for up to 2.5 years for individuals with dementia.
Participants were asked to response to one ofthe elicited questions such as "describe the
person who most influences your life." It was found that linguistic ability declined in
both grammatical complexity and propositional content in normal elders' spoken
narratives between ages of 74 to 78 years. The results also showed that dementia
accelerated the decline in linguistic abilities, and age was a major predictor in linguistic
99
decline for people with dementia. However, although linguistic abilities diminished
sharply as dementia progressed, individuals with dementia were still able to express basic
information. It was worth noting that linguistic ability in normal elders was found to
decline between ages of 65 and 80 years.
In order to detect the early signs ofAD without confounding with the language
changes observed in normal aging, Chapman and colleagues (Chapman, Zientz, Weiner,
Rosenberg, Frawley, & Bums, 2002) attempted to analyze discourse differences based on
the gist and detail-level processing. The gist-based processing acquired information
reconstruction, integration, and different cognitive and language functions than the
detailed-level processing. Different from the gist-based processing, the detailed-level
processing asked people to recall the details with little or no transformation ofthe content.
Twenty-four individuals with mild AD, 20 individuals with mild cognitive impairments
(MCl), and 25 normal elders participated in this study. A biographical narrative was read
aloud to the participants while a written version ofthe narrative was provided.
Participants were asked to provide a summary, a main idea, and a lesson as a measure of
the gist-level processing, and detailed information about the narrative as the measure of
the detail-level processing. The findings supported that people with MCI and AD were
impaired in the gist-level and detail-level processing while the gist-level processing was
preserved in normal elders.
Five individuals with AD and 27 neurologically normal elders were asked to
provide a story based on a bank robbery picture in the study ofDuong, Tardif, and Ska
(2003). The narratives were collected every 6 month for five consecutive times. The ratio
- --- --------
100
ofmodalizing discourse to referential discourse (M/R ratio) and the repetition ofexpected
ideas were calculated. Modalizing discourse was defined as participants' comments,
judgments, or uncertainty while delivering their narratives such as "I think." The findings
showed that individuals with AD used more modalizing discourse than normal elders did;
however, the MIR ratio decreased from the first to fifth assessment session in most ofthe
individuals with AD. Therefore, pragmatic abilities ofindividuals with AD were well
preserved at the early stage since they produced large amount ofmodalizing discourse in
order to maintain the conversation.
Fleming and Harris (2008) intended to determine ifdiscourses provided definite
information to distinguish normal elders and people with mild cognitive impairments
(MCI). Eight adults with mild cognitive impairments and eight age-matched normal
elders were asked to provide a spoken discourse sample about a trip to New York. It was
revealed that normal elderly adults used more words and provided more detailed
information than did people with MCI. The inability ofretrieving words and supplying
more detailed information was detected from the discourse analysis. Additionally, the
inability to provide core elements in people with MCI reflected the reduced planning,
organizing, and cognitive flexibility skills. As regard to the syntactic complexity, there
was no difference between the normal elderly group and the MCI group. Therefore, the
syntactic seemed spared in people with MCI.
Since the focus ofthe present study was idiom comprehension in adults across the
lifespan, it was essential to review the literature about how elderly processed figurative
language especially idioms. The literature offigurative language in normal aging
101
populations was limited. In the following section, how figurative language was acquired
in children was reviewed first as ground knowledge, and then studies focusing on how
normal elderly processed figurative language were reviewed to provide a comprehensive
understanding oflanguage ability in normal aging populations.
Figurative Language in Normal Aging
Figurative language is commonly used in everyday conversation and speech, but
is acquired relatively late in comparison with regular sentences because of its multiple
meaning and abstract thinking process. Kempler and Van Lancker (1993) used The
Familiar and Novel Language Comprehension (FANL-C) protocol to test 175 participants,
ages 3 to 18 years, in order to understand the acquisition offamiliar language. The test
contained 20 familiar phrases including proverbs, idioms, and contextually bound social
interaction formulas, and 20 novel sentences. Each phrase was expressed by four pictures
with one correct meaning, one related meanings to the familiar phrase, and two foils with
concrete interpretations ofthe familiar phrases. Participants were asked to point to the
picture that best described the phrase. It was found that adult level comprehension was
reached by 8 years of age on novel sentences but not until 12 years ofage for familiar
phrases.
How children and adolescents developed the understanding of literal and
figurative meanings ofidioms was investigated in the Kempler et al.'s study (Kempler,
Van Lancker, Marchman, & Bates, 1999). Participants were 250 normal children,
adolescents, and young adults, age three to nineteen years, and they were divided into 17
groups based on their age. It was found that the comprehension of idiomatic and literal
102
meaning followed two different developmental trajectories with literal interpretations
developing faster and reaching adult level around age ofnine or ten, but idiomatic
interpretations reaching adult levels at the age of 15. Therefore, it was clear that literal
meanings ofidioms were acquired earlier than idiomatic meanings, and idiomatic
interpretations improved only after the children's understanding of literal meanings
reached ceiling. The dissociation appeared because the two types ofidiom interpretations
required different cognitive and neurological mechanism, and idiomatic expressions were
more difficult comprehend than the matched literal expressions.
Figurative language was rarely studied in neurologically normal older adults.
According to literature reviewed in the previous sections regarding language
comprehension and production in normal aging populations, several nonlinguistic
cognitive mediators such as declines in attention, memory, processing speed, and
executive function affected language ability in neurologically normal elders. Although
people with high educational level and strong verbal ability maintained their language
ability even at their 70s, findings were inconclusive across different tasks, measures, and
individual differences. Figurative language is commonly used in everyday speech, has
special functions, and is processed differently from regular sentences. Therefore, it is
necessary to review how neurologically normal older people processed figurative
language.
An early study done by Boswell (1979) studied metaphoric processing on 30
retired adults, ages 62 to 86 years, and 31 high-school students, ages 17 to 19 years. They
were asked to construct a story or to invent an explanation for four metaphors such as "a
--------------
103
nation is a warm ocean" and then to rate the difficulty of generating explanations for each
metaphor on a five-point scale. The analysis ofjudges' ratings revealed that high school
students' interpretations were more literal and analytic; however, adults' explanations
were more poetic and synthesizing. Therefore, the quality ofmetaphor interpretation
clearly distinguished the two age groups; moreover, education was strongly associated to
interpretation scores for two age groups.
The quality ofthe metaphor interpretations in young and old adults was examined
through professional and nonprofessional judgments in the Szuchman and Erber study
(1990). Thirty young adults, ages 20 to 30 years, and 30 older adults, ages 62 to 74 years,
were instructed to make up a story or a situation using eight metaphors. Five English
teachers served as professional raters and rated the interpretations on a seven-point scale
ranging from very literal to very poetic. Sixteen younger adults and sixteen older adults
who served as nonprofessional judges rated the same interpretations on a seven-point
scale ranging from excellent to poor instead ofpoetic to literal. The findings indicated
that the metaphor interpretation skills were well maintained through the later adult years.
Gregory and Waggoner (1996) studied how older and younger adults interpreted
metaphors that described four emotions: happiness, anger, sorrow, and fear. Twenty-four
college students and 24 older adults, mean age at 69.8 years, were presented 12
metaphors, three metaphors for each emotion. They were asked to select an emotion that
best described each sentence and further provided an explanation for each metaphor. It
was found that performance was similar on the forced-choice task but was different on
the explanation task. Older adults tended to make up a story, but were less likely to
104
specify relevant attributes and connections between metaphor and emotion compared to
younger adults. In general, older and younger adults used different explanation style and
explained the basis ofthe meaning differently.
Nippold and colleagues (Nippold, Uhden, & Schwarz, 1997) conducted a proverb
explanation task on 353 participants, ages 13 to 79 years, to examine the relationship
between educational level and proverb understanding in adults. Eight groups, 13-14 years,
16-17 years, 20-29 years, 30-39 years, 40-49 years, 50-59 years, 60-69 years and 70-79
years, were asked to provide short written explanation for 24 proverbs. All proverbs were
low familiarity with half ofthe proverbs concrete and the other half abstract. Each
proverb was embedded in a four-sentence story with the proverb always occurring at the
end ofthe story. The findings showed that proverb explanation reached peak around the
age 20s and remained stable during the age 30s, 40s and 50s, but slightly declined in the
age 60s, and the decline reached significant around the age 70s. Additionally, concrete
proverbs were easier than abstract ones for the three younger groups; however, there was
no significant difference between concrete and abstract proverbs explanation beyond the
age 30s. Regarding the relationship between the education level and proverb explanation,
adults with more years offormal education performed better than those with less
education. It was suggested that lexical retrieval and storage might contribute to the
poorer responses observed in older participants due to the common use ofvague and
unspecified terms. In addition, lifestyle factors, such as vocation, should be taken into
account in interpretation ofthe proverb explanation task.
105
Qualls and Harris (2003) attempted to determine how age, working memory,
figurative language type, and reading ability affected the comprehension offigurative
language. Participants were 40 young African American adults, age between 17 and 31,
and 40 older African American adults, age between 54 and 73. Participants were tested on
working memory, reading comprehension, vocabulary, and figurative language
comprehension which contained 20 idioms, 20 metaphors, and 20 metonyms. Each
phrase was followed by four choices including one correct figurative interpretation, one
correct literal interpretation, one incorrect opposite foil, and one incorrect elaborated foiL
It was found that working memory and reading comprehension had significant influence
on adults' comprehension offigurative language, and different cognitive processing
supported the comprehension ofdifferent types offigurative language. No age-related
decline was found in figurative language comprehension in older adults when a selection
ofresponse choices was provided. In addition, older adults performed better in idioms
than metaphors and showed the greatest difficulty in metonyms.
Thirty-five young adults, 35 middle-age adults and 35 older adults were recruited
to discover the relationship between proverb comprehension and executive function in
the Uekeermann, Thoma and Daurn study (2008). Participants were first asked to rate the
familiarity for each proverb and then to choose the figurative meanings ofeach proverb
from four possible choices, abstract-meaningful, abstract meaningless,
concretistic-meaningful, and concretistic-meaningless. Executive function was measured
by inhibition, set shifting, short-term, and working memory. The findings showed that
older people had impaired proverb comprehension and made more errors in choosing
106
abstract-meaningless, concretistic-meaningless, and concretistic-meaningful. Familiarity
ofproverbs did not contribute to the deficits since proverbs were rated more familiar in
older people compared to younger adults. Additionally, it was revealed that older people
tended to choose literal meanings (concretistic-meaningful) rather than figurative ones. It
was suggested that plausible executive function impairment was related to the proverb
comprehension deficits because ofolder people's poorer performance in working
memory and inhibition compared to younger adults. Therefore, older people who suffered
executive function impairments had difficulty suppressing alternative literal
interpretations of figurative expressions.
Summary ofLanguage andAging
Age had great impact on attention, memory, and executive function when people
gradually approached to their 70s and 80s, and the decline found in overall cognition
further negatively impacted language performance in older populations. Studies reviewed
above showed that the greater the linguistic complexity, the more difficulty in language
processing was found in normal older people (Emery, 1986). Additionally, difficulties in
naming, reduced syntactic complexity, and decrement in verbal fluency are commonly
observed during the age of70 (Ardila & Rosselli, 1996). Moreover, older people tended
to produce less complex sentences compared to younger people (Kemper, Greiner,
Marquis, Prenovost, & Mitzner, 2001; Kemper, Herman, & Lian, 2003; Kemper, Herman,
& Liu, 2004) to avoid high cognitive demands. As regard to comprehension, elders
maintained surface comprehension but had deficits in language comprehension involving
integration, inference, and construction (Cohen, 1979). Researchers pointed out that
107
normal aging affected comprehension ofcomplex sentences, and the decline began after
age of60 (Davis & Ball, 1989; Feier & Gerstman, 1980; Obler et aI., 1991).
However, variables such as education, verbal ability, and gender needed to be
taken into account since studies showed the impact ofthese variables on language ability
in the normal aging populations (Ardila & Rosselli, 1996; Harris, Rogers, & Qualls, 1998;
Juncos-Rabadan & Iglesias, 1994; Juncos-Rabadan, 1996; Mackenzie, 2000; Petros,
Norgaard, Olson, & Tabor, 1989). Additionally, different tasks might attribute to different
performance observed in older people (Burke & MacKay, 1997; Glosser & Deser, 1992;
Radvansky, 1999). For example, ifthe participant required encoding new information, it
might show age-linked deficits. In addition, low demanding tasks, such as interview,
showed no signs ofdecline compared to high demanding tasks, such as experimental
tasks. No sign ofdecline was observed ifparticipants were asked to perform an on-line
task, such as semantic priming. However, language deficits were shown on the off-line
tasks, such as information recall.
The study results in figurative language in the normal aging population were
inconclusive. No change in the metalinguistic tasks such as interpretation ofproverb and
idioms was found in older people in the Obler and Albert study (1989). Similarly,
Szuchman and Erber (1990) pointed out that metaphor interpretation performance was
well maintained through the later adult years. However, declines in figurative language
ability in normal elderly were reported in Nippold, Uhden, and Schwarz (1997) and
Uekeermann, Thomas, and Daum (2008). Different tasks and measurements yielded
different results in how normal older people processed figurative language. In addition,
108
Qualls and Harris (2003) claimed that different cognitive-linguistic abilities were
required for comprehension ofdifferent types offigurative languages, and working
memory and reading comprehension abilities were associated with figurative language
comprehension.
Hypotheses ofthe Current Study
This study was expected to find that younger and older adults would perform
equally well on the comprehension task because Qualls and Harris (2003) suggested that
no age-related decline was found in figurative language comprehension in older adults
when a selection ofresponse choices was provided. Secondly, this study predicted that
older adults would perform poorer than younger adults on the explanation task. Nippold,
Uhden, and Schwarz (1997) indicated that proverb explanation reached a peak around the
age 20s and remained stable during the 30s, 40s, and 50s, but showed a slight decline in
the age 60s with the decline reaching significance during the 70s. Therefore, this study
expected similar findings. Thirdly, according to Nippold and Duthie (2003), the
knowledge of an idiom was expected to be expanded and deepened, and mental imagery
was expected to become more figurative with advancing age. Therefore, mental imagery
ofidioms was expected to become more figurative with advancing age in the present
study.
Fourthly, participants were expected to perform better on highly familiar idioms
than on less familiar idioms across four age groups since familiar idioms were recalled
better, read faster, and comprehended better than less familiar idioms (Cronk &
109
Schweigert, 1992; Libben & Titone, 2008; Nippo1d & Taylor, 2002; Schweigert, 1986;
1991). Lastly, according to Cacciari and Glucksberg (1995), concrete literal meanings
generated mental images more easily than abstract-figurative meanings. Additionally, the
dual coding theory (Paivio, 1971, 1986) suggested that concrete concepts would more
easily form mental images than abstract concepts. Therefore, this study expected to find
differences in mental images between transparent and opaque idioms, with better
performance on transparent idioms than opaque ones. The hypotheses ofthis study were
summarized as following:
1. Younger and older adults would perform equally well on the comprehension task.
2. Older adults would perform poorer than younger adults on the explanation task.
3. Mental imagery ofidioms was expected to become more figurative with advancing
age.
4. Participants were expected to perform better on highly familiar idioms than on less
familiar idioms.
5. Participants were expected to perform better on transparent idioms than opaque idioms.
110
CHAPTER III
METHOD
Participants
Recruiting and Sampling
The participants in the present study were all neurologically healthy adults with
no known brain damage, neurological disorders, or cognitive impairments, and had not
been diagnosed with attention deficit disorders, anxiety, depression, schizophrenia, or any
other psychiatric conditions. All participants were not under any psychiatric medications
for attention deficit disorders I attention deficit hyperactivity disorders (ADD/ADHD),
depression, or anxiety while taking the tasks. In addition, all participants were native
English speakers according to participants' selfreports. The tasks used in the present
study might create disadvantages for English as second language individuals because of
their English proficiency rather than idiom understanding. All participants were recruited
from the Los Angeles-Long Beach metropolitan area in Southern California.
Participants who met the following exclusion criteria were excluded from this
study even though they are neurologically healthy adults and within qualified age ranges.
Exclusion criteria used in this study were as follows:
111
1. Participants whose first language was not English were excluded from this study.
2. Participants who were not able to read and write with corrected vision and
appropriate motor writing skills were excluded from this study. Tasks used in this
study required reading and writing; therefore, data collected from participants who
were unable to read and write would be misrepresented.
3. Participants older than 60 year-old were given the Mini-Mental State Examination
(MMSE) to evaluate mental state/cognitive functioning in order to exclude possible
dementia (Folstein, Folstein & McHugh, 1975). Participants who did not meet the
cutoff score, 24 out of30 points ofthe MMSE task (Tombaugh & McIntyre, 1992),
were still given the tasks, but the data points were excluded from this study.
Initially, a total of 134 individuals, 34 adults, ages 20-29 years (20s Group), 32
adults, ages 40-49 years (40s Group), 33 adults, ages 60-69 years (60s Group), and 35
adults, ages 80-89 years (80s Group), were recruited in the study. However, four adults in
20s Group were excluded because they did not complete all four tasks. Two adults in 40s
Group were also excluded because ofthe incompletion ofthe tasks and the other one's
first language was not English. Three adults in 60s Group were excluded because two of
them did not complete the tasks and the other one's handwriting was not legible due to
the poor hand coordination. Five adults in 80s Group were excluded because three of
them did not complete the tasks and the other two did not meet the MMSE cutoffscore
requirement. Therefore, the total number ofparticipants for the present study was limited
to 120.
112
Participants' Characteristics
The participants (n =120) included in the present study were thirty adults, ages
between 20 and 29 years old (M=23. 93, SD =3.11), thirty adults, ages between 40 and 49
years old (M=44.07, SD =3.52), thirty adults, ages between 60 and 69 years old (M=64.77,
SD =3.17), and thirty adults, ages between 80 and 89 years old (M=82.1 0, SD =2.73).
There were 6 males and 24 females in 20s Group, 17 males and 13 females in 40s Group,
7 males and 23 females in 60s Group, and 8 males and 22 females in 80s Group. The total
gender ratio was 31.7% male and 68.3% female.
Participants' highest education was converted into total years of education
adapted from the study ofNippold, Uhden, and Schwarz (1997). The converting method
for total education years was listed in Table 3.1. The distribution ofeducation years was
summarized in Table 3.2 and Figure 3.1.
The mean education length was 15.13 years (SD=2.00) for 20s Group, 15.50
(SD=2.43) for 40s Group, 15.47 (SD=2.15) for 60s Group, and 15.30 (SD=2.22) for 80s
Group. A one-way, between-subjects analysis ofvariance (ANOVA) was conducted to
analyze the effect ofage on total education years. There was no significant main effects
for the age difference, F(3, 116) = 0.18,p = .91, "1
2
= .01. There were no significant
differences among the four groups' total years of education.
Table 3.1. The Converting Method for Total Education Years
Highest Education
High school diploma
Some college
Completed trade school
Associate ofArts (AA) degree
Bachelor ofArts (B.A.) or Bachelor ofSciences (B.S.),
Some grad school
Master's degree
Law degree
Ph.D
Education Years
12 years
13 years
14 years
15 years
16 years
17 years
18 years
19 years
20 years
113
Table 3.2. The Education Distributions for Each Age Group
20s Group 40s Group 60s Group 80s Group
High School 6 4 2 5
Some College 2 6 6 5
AA 6 3 3 5
BAIBS 12 8 12 7
MA/JD/PhD 4 9 7 8
Total 30 30 30 30
114
1020s
__I'~40S
o 60s
,B80s
o
2
14 --"------------------------------
I
I
"I
10 +----------------------
I
8 +1-------------------
High School Some College M BAiBS MAlJD/PhD
Figure 3.1. The Education Distributions for Each Age Group
The Mini-Mental State Examination (MMSE) was given only to 60s Group and
80 Groups to identify possible cognitive declines. All participants met the cutoff score, 24
out of30 points ofthe MMSE task (Tombaugh & McIntyre, 1992), The mean MMSE
score was 29.80 (SD=0.48, range=28-30) for 60s Group and 28.90 (SD=0.96,
range=27-30) for 80s Group.
Materials
Stimuli
Twenty idioms, 10 transparent and 10 opaque idioms used in the Nippold and
Duthie study (2003), were used as the stimuli in the present study. The complete list of
115
the idioms was provided in Appendix A. Each expression was a verb phrase with three to
five words, such as ''put one sfoot down" or "have a soft spotfor something." The
transparency rating was drawn from the study ofNippold and Rudzinski (1993). In their
study, twenty adults living in western Oregon were asked to rate 100 idioms for the
familiarity and transparency. These participants were asked to rate how closely the literal
and figurative meanings ofeach idiom compared using a three-point scale (1 = closely
related, 2 = somewhat related, 3 = not related). For the 10 transparent idioms used in the
present study, the mean transparency rating was 1.63 (range = 1.15 - 1.90) and the mean
transparency rating was 2.80 (range = 2.65 - 3.00) for the 10 opaque idioms. In addition,
according to Nippold and Duthie (2003), the difference between the 10 transparent and
10 opaque idioms reached the statistically significant level.
Background Questionnaire
Participants' contact information, including both phone numbers and email
addresses, age, gender, date ofbirth, date oftest, primary language spoken at home,
length of speaking English, length of living in the United States, current or previous
occupation, and highest level of formal education achieved were collected from the
background questionnaire. The background questionnaire was provided in Appendix B.
Familiarity Rating
The familiar rating task asked participants to rate each idiom and to circle the
number that best described how familiar he or she was with the idiom. Two practice
examples were provided to insure that the participants understood the purpose ofthe
idiom familiarity task. A statement "1 have heard or read this idiom" was asked after each
116
idiom was presented. Five responses, including never, once, a few times, several times,
and many times, were provided. The participant was asked to circle only one response
that best described his or her experience with the idiom. The familiarity rating task was
provided in Appendix C.
Idiom Explanation/Mental Imagery Task
The same twenty idioms were used in Idiom Explanation/Mental Imagery Task
with different orders to avoid the order effect. In each group, 15 participants were given A
form and the other 15 were given B Form. The only difference between Form A and B
was the order ofidioms. Participants first were asked to write what they thought each
idiom meant and then were asked to describe a situation where someone would use that
idiom. Two practice examples were provided to instruct the participants to complete this
task. This task was untimed, and participants were given a certain amount ofspace to
write. In addition, no minimum or maximum number of sentences was required. The
idiom explanation and mental imagery task were provided in Appendix D.
Idiom Comprehension Task
The same twenty idioms were used in Idiom Comprehension Task; this task was a
forced-choice task. No context was provided for each idiom because the context may
provide sufficient cues for participates to figure out the meaning ofthe idiom, which
would create the ceiling effect in the present study. Participants were asked to circle the
best interpretation for each idiom from four responses, and were encouraged to guess if
they did not know the answer. Two practice examples were provided to instruct the
117
participants to complete this task. There was no time limit to complete the task. The
idiom comprehension task was provided in Appendix E.
Procedures
The participants were tested either individually or in small groups ofthree to five
people. Each participant was given the tasks in the same order: one-page background
questionnaire, Familiarity Rating of20 idioms, Idiom Explanation/Mental Imagery Task,
and then Idiom Comprehension Task. Only one task was given at a time. Instructions for
taking each task and examples were provided prior to the task. Participants older than 60
years-old were additionally given the Mini-Mental State Examination (MMSE) to
evaluate their mental state/cognitive functioning after filling out the one-page background
questionnaire.
After filling out the questionnaire, participants were asked to rate the familiarity
of20 idioms on a five-point scale based on their experience. After rating each idiom on
familiarity, participants were asked to provide a written explanation for each idiom and
then to describe a situation where someone would use that idiom. Participants then were
asked to circle the best interpretation of each idiom from four possible choices. The idiom
explanation and mental imagery tasks were submitted to the participants before the
comprehension task was presented. This prevented participants from using the answer
choices on the comprehension task to assist in explaining the idioms. In sum, the order of
the tasks was consistent across all age groups as following: Background Questionnaire~
118
MMSE (if older than 60 year old)~Familiarity Rating~Idiom Explanation! Mental
Imagery Task~Idiom Comprehension Task.
All tasks used in this study were untimed since older groups might be
disadvantaged by their slow processing speed. In addition, there was no minimum
number ofsentences for Idiom Explanation and Mental Imagery Task in order to avoid
taxing the older participants' motor skills. The font was enlarged to improve the visibility
for older adults.
Coding System
Idiom Comprehension Task
Participants' answer on Idiom Comprehension Task was scored either 0 for the
incorrect response or 1 for the correct response. The total points ofthis task were 20
points.
Idiom Explanation and Mental Imagery Scoring
The coding systems adapted from the previous studies (Nippold & Duthie, 2003;
Duthie et aI., 2008) were used to score participants' explanations ofthe idioms on the
Idiom Explanation and Mental Imagery Task. The points were assigned as follows.
119
Table 3.3. Idiom Explanation Scoring System
Points
o
1
2
Scoring Criteria
The response is inaccurate, a restatement, literal, or absent
The response is related but vague or incomplete
The response is accurate, clear, and complete
For example, responses to the idiom "have a soft spot/or something," which
means to be kind to others or fond of something, were marked as follows (selected from
participants'responses):
o= A weak spot
1 = Someone is sensitive toward some aspect ofobject or person
2 = You like it more than other things
Participant's mental images were scored using the following scoring system.
Table 3.4. Mental Imagery Scoring System
Points
o
1
2
Scoring criteria
The response is irrelevant, a restatement, or absent
The response is relevant but vague or incomplete
The response is relevant, clear, and complete
120
For example, responses to the idiom "have a soft spotjor something," which
means to be kind to others or fond of something, were marked as follows (selected from
participants'responses):
o:=: Someone is asking for money in front ofthe store, your spouse gets upset
and you tell them have a soft spot.
I:=: I have a soft spot when it comes to an older person needing help with doors,
or wheelchairs. I always try to assist them.
2:=: Usually he didn't like Irish music but he had a soft spot for certain lines in
Danny Boy, so he always gave money to any street musician who played it.
Please see Appendix F for more examples ofparticipants' responses for Idiom
Explanation and Mental Imagery Task, and Appendix G for examples ofgood mental
imagery responses.
Interrater Reliability
Another investigator, who was trained and familiar with the scoring system,
scored both Idiom Explanation Task and Mental Imagery Task. The level ofagreement
between two investigators was 96% for 20s Group, 93% for 40s Group, 83% for 60s
Group, and 87% for 80s Group. The disagreements were discussed, and a second round
of interrater reliability was conducted. The second round inter-rater agreement was 99%
for all four groups. The disagreements were further discussed and then conducted a third
round ofinterrater reliability check. All disagreements in scoring were resolved through
discussion and reached 100% agreement.
121
CHAPTERIY
RESULTS
The dependent variables ofthis study were scores of Idiom Explanation Task,
Idiom Mental Imagery Task, and Idiom Comprehension Task. The independent variable
was age group. The data was analyzed by one-way analysis of covariance (ANOYA). Eta
Square (11
2
) was also calculated to report effect size. The significance level was set at
p< .05. The familiarity rating ofidioms and education years were further examined to
discover their relationships with the idiom explanation, idiom comprehension, and mental
imagery performance.
Age versus Idiom Comprehension
Hypothesis #1: Younger and older adults wouldperform equally well on the idiom
comprehension task.
A one-way, between-subjects analysis ofvariance (ANOYA) was conducted to
analyze the effects ofage on the idiom comprehension task. The independent variable
was the age group, with four levels (20s Group, 40s Group, 60s Group, and 80s Group).
For 20s Group, the mean ofthe comprehension task scores was 17.50 (SD = 1.61). For
40s Group, the mean ofthe compression task scores was 18.20 (SD = 1.71). For 60s
122
Group, the mean ofthe compression task scores was 18.27 (SD = 1.64). For 80s Group,
the mean ofthe compression task scores was 18.03 (SD = 1.40). The performance on
Idiom Comprehension Task was summarized in Table 4.1.
Table 4.1. Performance on Idiom Comprehension Task (n=30 per group)
20s 40s 60s 80s
M 17.50 18.20 18.27 18.03
SD 1.61 1.71 1.64 1.40
Range 14-20 13-20 15-20 15-20
(0-20)
There was no significant main effect ofage on idiom comprehension, F(3, 116) =
1.42,p = .24, r/ = .04. In other words, there was no significant difference among the four
groups' comprehension task scores. Measures ofthe idiom comprehension task scores in
each age group were reported in Table 4.2.
Table 4.2. One-Way, Between Subjects Analysis ofVariance on Idiom Comprehension
Source df F
11
2
P
Age 3 1.42 .04 .24
Error Between 116
Total 120
*p < .05. ** P < .01.
123
Age versus Idiom Explanation
Hypothesis #2: Older adults were expected to perform worse than younger adults on the
explanation task.
A one-way, between-subjects analysis ofvariance (ANOVA) was conducted to
analyze the effects ofage on Idiom Explanation Task. The independent variable was the
age group, with four levels (20s, 40s, 60s, and 80s Group). For 20s Group, the mean of
the explanation task scores was 28.23 (SD = 4.18). For 40s Group, the mean ofthe
explanation task scores was 30.30 (SD = 4.90). For 60s Group, the mean ofthe
explanation task scores was 32.17 (SD = 5.72). For 80s Group, the mean ofthe
explanation task scores was 30.50 (SD = 6.20). The performance on Idiom Explanation
Task across four groups was summarized in Table 4.3.
Table 4.3. The Performance on Idiom Explanation Task (n=30 per group)
20s 40s 60s 80s
M 28.23 30.30 32.17 30.50
SD 4.18 4.90 5.72 6.20
Range 22-35 21-38 18-39 16-40
(0-40)
There was a significant main effect ofage on idiom explanation, F(3, 116) = 2.77,
P =.04, '7
2
= .07. Follow-up tests were conducted to evaluate pairwise differences among
the means using a Tukey HSD test (atp < .05). There was a significant difference
obtained only between the comparison of20s Group and 60s Group. The participants in
60s Group obtained significantly higher scores than 20s Group did on Idiom Explanation
124
Task. No statistically significant differences obtained among other group comparisons.
Measures ofthe explanation task scores in each age group were reported in Table 4.4.
Table 4.4. One-Way Between Subjects Analysis ofYariance on Idiom Explanation Task
Source 4f F
1]2
P
Age 3 2.77* .07 .04
Error Between 116
Total 120
*p < .05. ** p < .01.
Age versus Mental Imagery
Hypothesis #3: Mental imagery ofidioms was expected to become more figurative with
advancing age.
A one-way, between-subjects analysis ofvariance (ANOYA) was conducted to
analyze the effects ofage on the mental imagery task. The independent variable was the
same age groups. For 20s Group, the mean ofthe mental imagery task scores was 28.63
(SD = 4.68). For 40s Group, the mean ofthe mental imagery task scores was 30.93
(SD=5.00). For 60s Group, the mean ofthe mental imagery task scores was 30.23
(SD=8.72). For 80s Group, the mean ofthe mental imagery task scores was 27.60
(SD=8.69). The performance on Mental Imagery Task across four groups was
summarized in Table 4.5.
125
Table 4.5. The Performance on Mental Imagery Task (n=30 per group)
20s 40s 60s 80s
M 28.63 30.93 30.23 27.60
SD 4.68 5.00 8.72 8.69
Range 20-36 18-38 8-38 9-40
(0-40)
There was no significant main effect ofage on mental imagery performance,
F(3,116)= 1.38,p = .25, 1]2 = .04. There was no significant difference among the four
groups' mental imagery task scores. Measures ofthe mental imagery task scores in each
age group were reported in Table 4.6.
Table 4.6. One-Way Between Subjects Analysis ofVariance on Mental Imagery Task
Source df F
'Il
2
p
Age 3 1.38 .04 .25
Error Between 116
Total 120
*p < .05. ** p < .01.
126
Familiarity Rating versus Idiom Comprehension, Explanation, and Mental Imagery
Hypothesis #4: Participants were expected to perform better on highlyfamiliar idioms
than less familiar idioms.
Pearson product-moment correlation coefficients were computed to assess the
relationship between the familiarity rating and 1) idiom comprehension task scores, 2)
idiom explanation task scores, and 3) mental imagery task scores. There was a significant
correlation between the familiarity rating and idiom comprehension task scores,
r(120)= .34, p < .01. There was also a significant correlation between the familiarity
rating and idiom explanation task scores, r (120) = .32, p < .01. In addition, there was a
significant correlation between the familiarity rating and mental imagery task scores,
r(120) = .30,p < .01. The results ofthe correlation analysis were reported in Table 4.7.
Table 4.7. Pearson Correlation Coefficients among Familiarity Rating and Idiom
Explanation, Mental Imagery, and Idiom Comprehension Task
Idiom Idiom Explanation Mental Imagery
Comprehension
Familiarity
*p < .05. ** p < .01.
.34** .32** .30**
Age versus Familiarity Rating
A one-way, between-subjects analysis ofvariance (ANOVA) was conducted to
analyze the effects ofage on the idiom familiarity rating. The independent variable was
127
the same age groups. The dependent variable was the raw scores ofthe familiarity rating.
For 20s Group, the mean ofthe familiarity rating scores was 87.03 (SD = 7.95). For 40s
Group, the mean ofthe familiarity rating scores was 91.13 (SD = 7.59). For 60s Group,
the mean ofthe familiarity rating scores was 93.90 (SD = 7.17). For 80s Group, the mean
ofthe familiarity rating scores was 90.33 (SD = 10.26). The performance on Idiom
Familiarity Rating across four groups was summarized in Table 4.8.
Table 4.8. The Idiom Familiarity Rating across Four Age Groups (n=30 per group)
20s 40s 60s 80s
M 87.03 91.13 93.90 90.33
SD 7.95 7.59 7.17 10.26
Range 72-98 71-100 76-100 61-100
(0-100)
There was a significant main effect for the age difference, F(3, 116) = 3.46,
p= .02, 11
2
= .08. Follow-up tests were conducted to evaluate pairwise differences among
the means using a Tukey HSD test (atp < .05). There was a significant difference
obtained only between the comparison of20s Group and 60s Group. The participants in
60s Group were more familiar with the idioms used in this study than the participants in
20s. No statistical significant differences obtained among other group comparisons.
Measures ofthe explanation task scores in each age group were reported in Table 4.9.
128
Table 4.9. One-Way Between Subjects Analysis ofYariance on Familiarity Rating Scores
Source df F
11
2
P
Age 3 3.46* .08 .02
Error Between 116
Total 120
*p < .05. ** p < .01.
Transparent Idioms versus Opaque Idioms
Hypothesis #5: Participants were expected to perform better on transparent idioms than
opaque idioms
A one-way, within-subjects analysis ofvariance (ANOYA) was conducted to
analyze the effects ofidiom transparency on idiom mental imagery. The independent
variable was the idiom transparency, with two levels (transparent idioms and opaque
idioms). For the transparent idioms, the mean ofMental Imagery Task scores was 14.54
(SD = 4.01). For the opaque idioms, the mean ofMental Imagery Task scores was 14.79
(SD = 3.83).
There was no significant effect ofidiom transparency on mental imagery
performance, F(l, 238) = 0.24,p = .62, 11
2
< .01. There was no significant difference
between the transparent and opaque idioms on mental imagery task scores. Measures of
the mental imagery task scores in each idiom type were reported in Table 4.10.
129
Table 4.10. One-Way, Within-Subjects Analysis ofVariance on Mental Imagery Task
Source df F
1]2
P
Age 0.24 <.01 .62
Error Within 238
Total 240
*p < .05. ** P < .01.
Education Years versus Idiom Comprehension, Explanation, and Mental Imagery
Pearson product-moment correlation coefficients were computed to assess the
relationships between education years and 1) Idiom Comprehension Task scores, 2)
Idiom Explanation Task scores, 3) Mental Imagery Task scores, and 4) Idiom Familiarity
Rating Task scores. There was no significant correlation between education years and
idiom comprehension task scores, r (120) = .01,p = .96, and between education years and
idiom familiarity rating scores, r (120) = .1 0, P = .29. There was a significant correlation
between education years and idiom explanation task scores, r (120) = .27, P < .01. There
was also a significant correlation between education years and mental imagery task
scores, r (120) = .30,p < .01. The results ofanalysis were reported in Table 4.11.
130
Table 4.11. Pearson Correlation Coefficients among Education Years and Idiom
Comprehension, Idiom Explanation, Mental Imagery, and Idiom Familiarity
Education
Years
*p < .05. ** p < .01
Idiom
Comprehension
.01
Idiom
Explanation
.27**
Mental
Imagery
.30**
Idiom
Familiarity
.10
Relationships among Idiom Familiarity, Idiom Comprehension, Idiom Explanation, and
Mental Imagery
Pearson product-moment correlation coefficients were computed to analyze the
relationships among four tasks administrated in the present study, including 1) Idiom
Comprehension Task scores, 2) Idiom Explanation Task scores, 3) Mental Imagery Task
scores, and 4) Idiom Familiarity Rating scores for each age group.
For 20s Group, the analyses yielded statistically significant and strong
correlations between Idiom Comprehension and Idiom Explanation (r(30)=.65,p<.01),
between Idiom Comprehension and Mental Imagery (r(30)=.50,p<.01), and between
Idiom Explanation and Mental Imagery (r(30)=.82,p<.01). However, the results were not
statistically significant between Idiom Familiarity and Idiom Comprehension (r(30)=.25,
p=.19), between Idiom Familiarity and Idiom Explanation (r(30)=.23,p=.21), and
between Idiom Familiarity and Mental Imagery (r(30)=.33,p=.08). The results of
analysis for 20s Group were reported in Table 4.12.
131
Table 4.12. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for 20s Group
Compo Exp. Img. Familiarity
Compo 1 .65** .50** .25
Exp. ..65** 1 .82** .23
Img. .50** .82** 1 .33
Familiarity .25 .23 .33 1
Note. Comp.= Idiom Comprehension, Exp.= Idiom Explanation, Img.=Mental Imagery
*p<.05. **p<.OI
For 40s Group, the result also showed statistically significant and strong
correlations between Idiom Comprehension and Idiom Explanation (r(30)=.45,p=.01)
and between Idiom Explanation and Mental Imagery (r(30)=.86,p<.01), and between
Idiom Familiarity and Mental Imagery (r(30)=.50, p<.O 1). It also showed statistically
significant, but moderate correlations between Idiom Comprehension and Mental
Imagery (r(30)=.37,p=.04), and between Idiom Familiarity and Idiom Explanation
(r(30)=.41,p=.02). There was no statistically significant correlation between Idiom
Familiarity and Idiom Comprehension (r(30)=.17,p=,37). The results of analysis for 40s
Group were reported in Table 4.13.
132
Table 4.13. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for 40s Group
Compo Exp. Img. Familiarity
Compo 1 .45* .37* .17
Exp. .45* 1 .86** .41 *
Img. .37* .86** 1 .50**
Familiarity .17 .41 * .50** 1
Note. Comp.= Idiom Comprehension, Exp.= Idiom Explanation, Img.=Mental Imagery
*p<.05. **p<.OI
For 60s Group, no statistically significant correlation was obtained between Idiom
Comprehension and Idiom Explanation (r(30)=.28,p=.14), between Idiom Familiarity
and Idiom Comprehension (r(30)=.27,p=.I5), between Idiom Familiarity and Idiom
Explanation(r(30)=.29,p=.12), and between Idiom Familiarity and Mental Imagery
(r(30)=.24, p=.20). A statistically significant and moderate correlation was found between
Idiom Comprehension and Mental Imagery (r(30)=.40, p=.03), and a strong correlation
was found between Idiom Explanation and Mental Imagery (r(30)=.87,p<.OI). The
results ofanalysis for 60s Group were reported in Table 4.14.
133
Table 4.14. Pearson Correlation among Idiom Comprehension, Idiom Explanation,
Mental Imagery, and Idiom Familiarity Rating Task for 60s Group
Compo Exp. Img. Familiarity
Compo 1 .28 .40* .27
Exp. .28 1 .87** .29
Img. .40* .87** 1 .24
Familiarity .27 .29 .24 1
Note. Comp.= Idiom Comprehension, Exp.= Idiom Explanation, Img.=Mental Imagery
*p<.05. **p<.OI
For 80s Group, the analyses yielded statistically significant and strong
correlations between Idiom Comprehension and Idiom Explanation (r(30)=.60,p<.01),
between Idiom Comprehension and Mental Imagery (r(30)=.54,p<.01), between Idiom
Explanation and Mental Imagery (r(30)=.88,p<.01), but a moderate correlation between
Idiom Familiarity and Idiom Comprehension (r(30)=.46, p=.O 1). The results ofanalysis
were reported in Table 4.15. However, no statistically significant correlation was found
between Idiom Familiarity and Idiom Explanation (r(30)=.l8,p=.36), and between Idiom
Familiarity and Mental Imagery (r(30)=.29,p=.l2).
134
Table 4.15. Pearson Correlation Coefficients among Idiom Comprehension, Idiom
Explanation, Mental Imagery, and Idiom Familiarity Rating Task for 80s Group
Compo Exp. Img. Familiarity
Compo 1 .60** .54** .46*
Exp. .60** 1 .88**
.18
Img. .54** .88** 1 .29
Familiarity .46* .18 .29 1
Note. Comp.= Idiom Comprehension, Exp.= Idiom Explanation, Img.=Mental Imagery
*p<.05. **p<.OI
Additionally, the accuracy of Idiom Comprehension, Idiom Explanation, and
Mental Imagery Tasks for each idiom across four age groups was reported in Appendix H
and Appendix J. Additionally, the frequency offamiliarity rating for each idiom across
four age groups was reported in Appendix 1. Interestingly, all participants in 20s Group
and 40s Group were unable to provide accurate and relevant responses for Idiom #20 "go
against the grain" in Idiom Explanation and Mental Images Task. For Idiom #12 "get the
lead out, " participants in 20s Group appeared to have less accurate and relevant responses
on both Idiom Explanation and Mental Imagery Tasks, and also made more errors on
Idiom Comprehension Task. Moreover, participants in 20s Group and 40s Group also
showed more difficulties with Idiom #17 "bring the house down." Otherwise, no specific
trend was observed in the visual analysis.
135
CHAPTER V
DISCUSSION
The purpose ofthe present study was to examine how neurologically healthy
adults, ages between 20-29,40-49,60-69, and 80-89 years old, comprehended idioms
using the definition explanation task, mental imagery task, and forced-choice
comprehension task. Moreover, the effects offamiliarity and transparency ofthe idioms
were also examined in this study to provide a comprehensive understanding ofidiom
comprehension. The first aim ofthe study was to evaluate whether younger and older
adults performed equally well on the forced-choice idiom comprehension task. The
second aim was to discover ifyounger adults perfoffiled better than older adults on Idiom
Explanation Task, but provided less figurative mental images on Mental Imagery Task.
The present study also discovered whether transparency and familiarity ofthe idiom
affected idiom understanding.
136
Conclusions
Age Differences in Idiom Comprehension
The results generated from the comparisons ofthe four age groups showed that
there was no significant difference on the forced-choice comprehension task across four
age groups. The results supported the study of Qualls and Harris (2003) that no
age-related decline was found in figurative language comprehension in older adults when
a selection ofresponse choices was provided. The results were also consistent with
Gregory and Waggoner's findings (1996) that the performance on metaphors was similar
on the forced-choice task but was different on the explanation task between younger and
older people. Although no context was given for the comprehension task, the
forced-choice task was still relatively easier than the explanation task. Therefore,
different comprehension measures, direct versus indirect measures, and differences in
inferential demand yielded different performance (Gregory & Waggoner, 1996).
Regarding the explanation task, it was predicted that older adults would perform
worse than younger adults. According to Nippold, Uhden, and Schwarz (1997), proverb
explanation reached a peak around the age 20s and remained stable during the age 30s,
40s and 50s, but slightly declined in the age 60s, and the decline reached significance
around the age 70s. The present study focused on idiom comprehension with the age
range expanded to 80s, and mixed findings were presented. The participants in 60s Group
performed significantly better than 20s Group, but there was no significant difference
between other age groups. One possible explanation ofbetter performance in 60s Group
was that the familiarity of idioms was rated significantly higher in 60s than in 20s. It was
137
critical to consider the familiarity ofidioms since a significant correlation was found
between the familiarity rating and the idiom explanation task scores.
In addition, participants in 20s Group performed the worst on Idiom Explanation
Task among the four groups, and 40s Group and 80s Group performed equally well
although the performance on the explanation task varied widely in 80s than in 20s. The
results were also consistent with the familiarity rating that 20s rated lowest on the
familiarity rating than other age groups, whereas 40s and 80s' ratings were similar.
Participants in 80s Group did worse on the explanation task than 60s Group; however, it
did not reach statistical significance, which was similar to Nippold and colleagues'
finings (1997) that idiom explanation improved with advancing age from 20 years to 60
years old, but declined around 70s.
The results were also in line with Obler and Albert study (1989) that no changes
were shown in the metalinguistic tasks, such as interpretation ofproverbs and idioms in
normal older people. Additionally, the metaphor interpretation performance was also
found to be well maintained through the later adult years (Szuchman & Erber, 1990).
Although participants in 80s Group in the current study performed slightly worse on
Idiom Explanation Task, it did not reach the statistical significance. Therefore, it was
concluded that no significant change was found in idiom explanation in older populations.
It was worthy to note that 20s Group had difficulties on certain idioms, such as "get the
lead out" and "bring down the house"; on the other hand, the errors that 60s and 80s
made were heterogeneous and varied widely.
138
Researchers pointed out that difficulties in naming, reduced syntactic complexity,
and decrement in verbal fluency were commonly observed during the age of 70 years old
(Ardila & Rosselli, 1996). In addition, older people tended to produce less complex
sentences (Juncos-Rabadan & Iglesias, 1994; Kemper, Greiner, Marquis, Prenovost, &
Mitzner, 2001; Kemper, Herman, & Lian, 2003; Kemper, Herman, & Liu, 2004), had
difficulties accessing appropriate lexical items, integrating story elements in a complex
discourse structure (Juncos-Rabadan & Iglesias, 1994), and processing complex
discourse information (Ulatowska, Cannito, Hayashi, & Fleming, 1985). However, the
present study did not examine the sentence complexity, lexicon use, and information
integration. Therefore, the decline commonly observed in research was not shown in this
study since the focus ofthis study was idiom comprehension and processing, not
producing complex discourse, naming, and integrating information.
The interpretation of idioms was viewed as a metalinguistic task, which was a
higher level oflinguistic and cognitive process. According to Emery's study (1986), when
the level of linguistic complexity increased, the normal elders showed more pronounced
decrement in language ability. Normal elders performed as well as middle aged adults in
lexicon; performed worse during the word-internal morphological processing, and did
even worse on the sentence level task since it was more abstract and complex. However,
the interpretation of figurative language, such as idioms, metaphors, and proverbs, could
yield different results in comparison to the results commonly observed in language
decline with advancing age because ofthe way that idioms are stored and used. Idioms
are viewed as part ofthe mental lexicon because they can not be decomposed but restored
139
and accessed as a whole like mental lexicon (Nippold, 2007; Swinney & Cutler, 1979).
Although the definitions provided by 80s Group were less accurate in comparison with
40s and 60s Groups, it could not be overgeneralized to any other difficulties in the
syntactic complexity and word finding ability in elderly reported in other studies since
the syntactic complexity and use of lexicon were not measured in the present study.
Regarding language comprehension, research showed that elderly maintained
surface comprehension, but had deficits in language comprehension involving integration,
inference, and construction (Cohen, 1979). Moreover, normal aging affected the
comprehension of complex sentences, and the decline began after age 60 years old (Davis
& Ball, 1989; Feier & Gerstman, 1980; Obler et ai., 1991). Although the present study
tested idiom understanding using both receptive and expressive methods, idioms were
viewed more like giant lexical units rather than regular phrases. The current study did not
measure the comprehension of complex sentences, information integration, and inference;
therefore, whether language comprehension declined with advancing age was out ofthe
scope ofthe present study. According to the findings from the forced-choice
comprehension task, it was concluded that idiom comprehension was maintained with
advancing age when options were provided, which supported Qualls and Harris's findings
(2003).
Based on the results ofthe idiom comprehension and explanation task, 80s Group
did not perform significantly worse than the younger groups, and 60s Group
outperfonned other age groups on Idiom Explanation Task and Idiom Comprehension
Task although the significant difference was only found between 20s and 60s Group on
140
the explanation task. No significant difference was shown in the comprehension task.
Four age groups reached the ceiling while the forced-choice task was used to test the
idiom comprehension. It was noteworthy that the results yielded from the explanation
task followed a similar pattern shown in the familiarity rating. The 60s Group rated
highest on the familiarity rating and scored highest on Idiom Explanation Task, followed
by 40s and 80s Group with considerably similar ratings and scores. Therefore, the
familiarity ofthe idiom seemed to playa critical role while asked to provide a definition
or explanation for an idiom. The idioms used in the present study were more familiar to
40s, 60s, and 80s Groups; therefore, they performed better on the explanation and
comprehension task. The generation difference might also exist in idiom comprehension
across different age groups since 20s Group consistently had difficulties with certain
idioms such as "go against the grain, " "sing a different tune," and "get the lead out. "
The results ofthe present study indicated that idiom comprehension improved
with age from 20s to 60s and well maintained into 80s iftheir cognitive functions did not
decline. The results also showed that adults even in their early adulthood did not master
all idioms. Idiom comprehension is a life long learning process. How people
comprehended and understood idioms mainly depended on their exposure to and
familiarity with the idiom. Some idioms, such as "go against the grain" and "get the lead
out," were used more frequently among older people than younger populations. But,
some idioms, such as "get into one shair" and "go by the book," were widely known in
all age groups. Therefore, it was critical to take all possible factors into account while
exploring idiom comprehension in different ages.
141
It was noteworthy that some idioms were misinterpreted more frequently than
others. For example, "go against the grain" was often misinterpreted as "sail against the
wind." "Hold your head up" was frequently misinterpreted as "keep your chin up, " while
"sing a different tune" was easily misinterpreted as "march to the beat ofa different
drum." "Cast the first stone" was often misinterpreted as "break the ice." "Bring down
the house" was easily misinterpreted as ''paint the town red." "Blow oJ!steam" was
frequently mistaken as "blow your top."
The different results drawn from the forced-choice comprehension task and the
explanation task also provided valuable information. Gregory and Waggoner (1996)
claimed that different comprehension measures, direct versus indirect measures, and
differences in inferential demand yielded different performance. The explanation task, a
highly linguistic and cognitive demanding task, was more sensitive and more
straightforward to detect possible language impairments or declines. Participants in 80s
Group performed equally well as 40s and 60s Group did on the comprehension task, but
received lower scores on the idiom explanation than 60s Group. Although the difference
between the groups did not reach the statistical significant level, it provided different
insights while examining idiom comprehension through different tasks.
Dual Coding Theory and Idioms
The results ofMental Imagery Task indicated that no significant difference in
generating mental images between transparent and opaque idioms was found. Mental
images generated from the idioms used in the present study tended to be figurative rather
than literal for both transparent and opaque idioms. Although participants in the present
142
study had difficulty generating mental images for some idioms such as "go against the
grain," "get the lead out," and "sing a different tune, " no trend for a certain type of
idioms was observed. Ifthe participant was not familiar with the idiom or was not able to
provide a correct definition, he or she also had difficulty generating appropriate mental
images for the idiom. Cacciari and Glucksberg (1995) suggested that transparent idioms
may be easier to comprehend than opaque ones because ofthe involvement ofmental
imagery and the concreteness effect. However, the effect ofidiom familiarity was not
included in their study. According to the dual coding theory (Paivio, 1971, 1986),
concrete concepts would more easily form mental images than abstract concepts.
However, Nippold, Uhden, and Schwarz (1997) indicated that although some ofthe
proverbs evoked stronger images than others, not all concrete proverbs were imageable.
In addition, the relationship between imagery and comprehension was significantly
correlated for the children but not for the adolescents and adults. When the idiom was
highly familiar to people, the effect ofidiom transparency reduced, and generating mental
imagery was considerably easier for both transparent and opaque idioms. Given that all
participants in the present study were adults, and all idioms were rated as highly familiar,
it was not surprising that no significant difference in generating mental images was found
between transparent and opaque idioms in the current study.
Nippold and Duthie (2003) proposed that the knowledge ofan idiom was
expected to expand and deepen, and mental imagery was expected to become more
figurative with advancing age. Duthie and colleagues (Duthie, Nippold, Billow, &
Mansfield, 2008) also indicated that adults had a more metaphorical understanding of
143
proverbs than children and adolescents. When mental images were compared between
children and adults, the difference was more prominent since children were still in the
learning process through understanding the connection and relationship between literal
and figurative meanings. However, as soon as idioms were learned and stored as lexical
units, mental images reported by adults tended to be figurative rather than literal.
It was also noteworthy that 80s Group received the lowest score on Mental
Imagery Task compared to the other groups; however, it did not reach the statistical
significant level. Moreover, 40s and 60s Group performed equally well on the imagery
task and did better than 20s Group. Therefore, mental imagery became more figurative
with advancing age from 20 years to 69 year old, but people in 80s Group provided less
accurate mental images than the other groups did. Irrelevant mental images or wrong
interpretation ofthe idioms were observed more frequently in 80s Group than in other
age groups. Although 20s Group also received lower scores on the mental imagery task
compared to 40s and 60s Group, the lower scores mainly resulted from not knowing the
idioms, which was consistent with the familiarity rating. The 20s Group rated lowest on
the familiarity rating in comparison to other age groups. Without the knowledge to
provide an accurate definition for the idiom, it was not possible to form an appropriate
and relevant mental image for the idiom.
Forming the arbitrary link between idioms and their idiomatic meanings was a
critical process to learn idioms. According to the highly consistent mental images
provided by participants, conceptual metaphors played the key role that motivated what
the proverbs or idioms really meant (Bortfeld, 2002; Gibb, Storm, & Spivey-Knowlton,
144
1997). Gibbs and O'Brien (1990) suggested that conceptual metaphors motivated the link
between idioms and their nonliteral meanings, and conceptual metaphors were accessed
during the idiom comprehension process (Gibbs, Bogdanovich, Sykes, & Barr, 1997).
Participants in the present study formed similar mental images for the idioms they were
familiar with, which further supported the existence of conceptual metaphors. For
example, a great performance for "bring down the house/' a celebration event for "paint
the town red)" and following rules strictly like soldiers and police for "go by the booR'
were provided consistently across the four age groups.
Nippold and Duthie (2003) suggested that mental imagery would be automatically
created once the figurative meaning ofan idiom was learned. The present study showed
that participants tended to leave the idiom definition in blank instead of generating from
its literal meaning ifthey did not know the meaning ofthe idiom. This finding was
different from what has been observed from children and adolescents. With the exposure
experience and the maturity ofunderstanding and using language, adults provided highly
consistent mental images for familiar idioms; however, they did not generate figurative
meanings through analyzing literal meanings or syntactical/semantic parts for unfamiliar
ones. Thus, mental images provided a unique approach to access how conceptual
metaphors work in the process ofidiom comprehension, and to examine ifboth literal
and figurative meanings had to be accessed during the idiom comprehension process. The
results ofthe present study were consistent with Nippold and Duthie's findings (2003)
that mental imagery may serve as an indicator of deep understanding offigurative
meanings in idiom comprehension once the meaning ofan idiom has been learned.
145
Factors Involved in Idiom Comprehension
Nunberg, Sag, and Wasow (1994) suggested that difficulties in the analysis of
idioms resulted from a confusion ofthe key semantic properties associated with idiomatic
meanings, including conventionality, transparency, and compositionality. Additionally,
idioms varied on transparency, decomposability, and ambiguousness, which affected how
people interpreted idioms (Cacciari, Reatio, Colombo, Padovani, Rizzo & Papagno,
2006). Therefore, the present study also examined the impact offamiliarity and
transparency during the idiom comprehension process.
Familiarity
According to Schweigert (1986), the familiarity ofthe idiom needed to be taken
into account in the model ofidiom processing. Nippold and Taylor (2002) also suggested
that the transparency and familiarity were strongly correlated to idiom comprehension in
children and adolescents, and needed to be considered while examining idiom
comprehension. Participants in the present study rated all 20 idioms as highly familiar
idioms. Although the familiarity rating increased from 20s to 60s, the rating surprisingly
decreased slightly in 80s Group.
The results ofthe present study also supported the importance oftaking idiom
familiarity into account while examining the idiom comprehension process. It was found
that the familiarity was significantly correlated to the scores on the explanation task,
mental imagery task, and comprehension task. Therefore, when people were familiar with
the idiom, people was able to provide an accurate definition, a relevant mental image, and
also selected the best explanation from options even though no context was provided for
146
the idiom. It was also interesting to discover that people had difficulties providing
definitions and images for some idioms even though the idioms were rated familiar. Thus,
people might have heard ofor were exposed to the idiom several times, but they did not
understand the meaning ofthe idiom, or they grasped the wrong meaning ofthe idiom.
For example, participants in the present study commonly misinterpreted "go against the
grain" as being unique from rest ofthe people. Since familiarity was subjective and
heavily depended on the personal experience and feelings, it was difficult to detect the
influence of familiarity in idiom comprehension based only on the rating scale.
Literal interpretation was rarely shown across four age groups. Based on Graded
Salience Hypothesis (GSH) (Giora, 1997), comprehension offigurative and literal
language should be depending on its salient meanings. The salience ofa word or an
utterance is a function of its conventionality, familiarity, and given context. According to
Giora (1997), in idiom comprehension, the idiomatic meaning is salient because it can
not be decomposed but restored as a whole like mental lexicon. Schweigert and Moates
(1988) made a similar statement that familiar idioms were more likely processed as
lexical units than less familiar idioms. Therefore, it was not surprising to find that only
figurative meanings were reported on the explanation task, and the mental image
generated for each idiom tended to be idiomatic instead ofliteral. Since idioms were
viewed and stored as mental lexicon, the figurative meaning ofthe idiom was activated
first. Unlike children, who learned the idiom meaning through the context and the
relationship between literal and figurative meanings ofthe idiom, adults tended to leave it
in blank on the explanation and mental imagery task instead ofguessing the meaning
147
through syntactic or semantic analysis ofthe idiom. In addition, the present study did not
provide context for each idiom; therefore, it was difficult for participants to generate the
meaning ofthe idiom using context cues.
Given that all the idioms used in the present study were rated as highly familiar,
figurative meanings activated first and immediately without assessing and analyzing their
semantic and syntax structures because the figurative meaning was more salient than its
literal meaning (Giora, 1997). Moreover, mental images were more metaphorical than
literal because ofthe easiness ofactivating conceptual metaphors which motivated the
link between idioms and their nonliteral meanings and the salience ofits figurative
meaning rather than literal meaning.
Transparency
The transparent and opaque idioms were only compared on the mental imagery
task in order to examine the difference of generating mental images between two types of
idioms. The purpose ofexamining transparency in idiom understanding was to find out if
the concreteness effect existed in idiom comprehension and ifthe dual coding theory was
applicable to the idiom comprehension process. According to the dual coding theory
(Pavio, 1976), mental images were more easily generated when people processed
concrete concepts or information than abstract information. Cacciari and Glucksberg
(1995) also suggested that transparent idioms may be easier to comprehend than opaque
ones because ofthe involvement ofmental imagery and the concreteness effect.
Therefore, one ofthe research hypotheses was that participants would score better on
transparent idioms than opaque ones since people were expected to generate mental
148
images for transparent idioms more easily. However, the difference on the mental
imagery task between these two types ofidioms was not found in the present study.
Participants in the present study generated mental images associated to the figurative
meanings for both transparent and opaque idioms ifthey were familiar with the idiom.
Therefore, in the present study, familiarity seemed to playa stronger role than
transparency while forming mental images for idioms. In other words, once an idiom was
learned and stored as mental lexicon, people applied idiomatic meanings and generated
figurative mental image immediately without accessing its literal meanings and mental
images.
The transparency rating used in the present study was adapted from Nippold and
Rudzinski (1993) in which 20 adults living in western Oregon were asked to rate 100
idioms for familiarity and transparency; therefore, participants in the present study did
not rate the transparency for each idiom. However, similar to familiarity, the transparency
rating was subjective and difficult to be applied to different studies and participants.
Therefore, it was not surprising to discover that no significant difference on the mental
imagery task was found between transparent and opaque idioms.
Education
The total years ofparticipant's education was significantly related to the
definition task and mental imagery task, but not to the comprehension task. It was not
surprising to find the correlation since these two tasks were highly linguistic and
cognitive demanding tasks and required sufficient understanding and knowledge of
language that developed relatively late. Therefore, when people were exposed to
149
education longer, it was expected that they had more opportunities to deepen and expand
their language knowledge. It was not unexpected to find that no significant correlation
was found between education years and the comprehension task since the forced-choice
task format was relativity easier than the tasks that required integration, inference, and
organization.
Limitations and Future Research
Limitations
Some limitations were noted in the present study. First, idioms used in the study
were highly familiar to the participants according to the familiarity rating. Therefore, it
would be difficult to detect the different interpretation oftransparent and opaque idioms
since familiarity played a stronger role than transparency in idiom comprehension in the
present study. Secondly, the subjects' definitions ofidioms varied among the different age
groups, and people tended to interpret their figurative meanings slightly differently.
According to Gregory and Waggoner (1996), older and younger adults used different
explanation styles and explained the basis ofthe meaning differently while they
interpreted metaphors. Thus, the difference ofinterpreting idioms may not only create
difficulties on scoring but also underestimate the comprehension ofidioms. Thirdly,
individual differences existed in all four age groups; however, the results represented
only the performance ofeach age group in general. In addition, this study required
participants to write down the definitions and mental images, which created great
disadvantages for older participants. Poor control ofhand movement and taking up to two
150
to three hours to finish all tasks were commonly observed in 80s Group. Although all
tasks were untimed in the present study, writing was highly demanding and remained as a
major difficulty for older participants to complete the tasks.
Another limitation was drawn from the lack ofunderstanding ofparticipants'
language and cognitive performance. Although Mini-Mental State Examination (MMSE)
was given to 60s and 80s Groups to rule out the possibility of cognitive decline due to
aging, idiom comprehension was also associated with people's language ability. No
standardized language test was administered in the present study, which might limit the
scope ofunderstanding idiom comprehension and overlook the connection between
overall language performance and idiom understanding. Rassiga and colleagues (Rassiga
et aI., 2009) hypothesized that elderly with good executive function would perform well
on the idiom comprehension task; however, elderly with lower executive function would
perform at a lower level. The present study was not able to verify this hypothesis because
no standardized cognitive test was given to establish the baseline ofparticipants,
executive function.
Moreover, the present study adapted Nippold's (2007) idiom transparency rating.
However, transparency might differ from person to person depending on their
interpretation and exposures to idioms. The concreteness effect might not be reflected on
the mental imagery task because ofthe selection oftransparent and opaque idioms based
on a different study. Moreover, the high familiarity rating did not mean that people had
complete understanding ofthe idiom. Using the idiom in an accurate and appropriate
context is at a different processing level from only having heard ofthe idiom. Therefore,
151
the disconnection between familiarity and idiom comprehension was a major missing
puzzle in the present study.
Future Research
Given limited research in idiom comprehension in normal aging populations and
the lack ofknowledge ofthe role ofmental imagery in idiom comprehension, more
research is warranted. Future research could use oral explanations instead ofwritten
explanations to avoid creating handwriting difficulties for elderly and to obtain more
detailed information about their mental images. In addition, asking participants to rate the
easiness of generating a mental image and to form images by providing specific prompts
could offer more insights about creating mental images during idiom comprehension,
since it was difficult to detect the process ofproducing mental images using behavioral
tasks. In addition, future research should ask participants to rate not only familiarly but
also transparency to collect complete information of people's idiom understanding.
Familiarity and transparency are subjective to people depending on their life experience,
understanding, and exposures. Therefore, adapting the transparency rating from other
studies may not be sensitive enough to reflect idiom comprehension in different age
groups.
The present study did not include low familiarity idioms; thus, it was difficult to
identify how people interpreted their mental images differently between low familiar and
highly familiar idioms. According to Nippold (2007), children learn idioms through
analyzing literal meaning ofthe idiom and using context clues to figure out the figurative
meaning. Therefore, it would be interesting to compare how people generate mental
152
images differently for low and highly familiar idioms without context, and how they
interpret these idioms in different ways.
Understanding figurative language is an important way to appreciate daily
language; however, it is rarely studied and explored in the research in children,
adolescents, or adults. In order to indentify the subtle language changes resulting from
mild Alzheimer's disease or mild cognitive impairments, it is necessary to have the
fundamental knowledge ofhow people use language in a more practical and natural way.
In addition, research beyond the scope ofexamining word finding function, syntactic
structures, and information integration is needed to detect the subtle changes. People with
widespread brain damage have often lost the ability of abstract thinking and frequently
substituted concrete interpretation (Kempler & Van Lancker, 1988). Obler and Albert
(1984) also pointed out that people with dementia had difficulties interpreting idioms and
proverbs, and their interpretations were fairly concrete. Future research could expand the
scope from the normal aging population to people with mild cognitive impairments and
early stage ofAlzheimer's disease to discover the possible changes in idiom
comprehension. Researchers also could examine different types offigurative language,
such as metaphors and proverbs, in the normal aging population to widen and deepen the
understanding offigurative language, which is an important piece in later language
development.
APPENDIX A
THE LIST OF IDIOMS USED IN THE PRESENT STUDY
(NIPPOLD & DUTHIE, 2003)
Transparent Idioms
1. Hold one's head up
2. Go by the book
3. Take someone under one's wing
4. Turn the other cheek
5. Run out ofsteam
6. Blow off stearn
7. Skate on thin ice
8. Sing a different tune
9. Cast the first stone
10. Go through the motions
Opaque Idioms
1. Get in someone's hair
2. Get the lead out
3. Keep one's shirt on
4. Blow one's own horn
5. Put one's foot down
6. Beat around the bush
7. Bring the house down
8. Paint the town red
9. Have a soft spot in one's heart
10. Go against the grain
153
APPENDIXB
BACKGROUND QUESTIONNAIRE
Name: (please Print) _
154
(first) (last)
female
Local phone: _
Email (if any): _
Age: _
Birth date:
---------------
Today's Date: _
Primary language spoken at home: _
(e.g., English, Spanish, Chinese, etc.)
How long have you spoken English?~
City and state ofresidence: _
How long have you lived in the United States? _
Gender: (circle one) male
Current or previous occupation(s):
Highest level of formal education achieved: _
(e.g., high school diploma, Associate ofArts, BA, BS, MA, etc.)
155
APPENDIXC
FAMILIARITY RATING
IDIOM FAMILIARITY TASK
Name:
--------------------------
(first) (last)
Many people use idioms when they talk or write. Examples include throw in the towel
and go around in circles. Some idioms are common and are used all the time. Others
are rare, and are hardly ever used. And still others are somewhere in between 
sometimes used but not very often.
I would like you to read a list ofidioms and tell me how common or rare you think they
are. Read each idiom and circle the number that best tells how familiar you are with that
expression. There are no right or wrong answers. Just circle the number that seems to be
the best answer for you.
Let's practice:
A. Throw in the towel
I have heard or read this idiom:
1. never
2. once
3. a few times
4. several times
5. many times
156
B. Go around in circles
I have heard or read this idiom:
1. never
2. once
3. a few times
4. several times
5. many times
There are 20 idioms in this booklet. Please answer each item. Draw a circle around
the number ofyour answer choice.
1. Bring the house down
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
2. Turn the other cheek
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
3. Go through the motions
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
157
4. Hold your head up
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
5. Blow one's own hom
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
6. Get the lead out
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
7. Blow off steam
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
8. Have a soft spot for something
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
158
9. Get in someone's hair
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
10. Take someone under one's wing
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
11. Put one's foot down
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
12. Keep one's shirt on
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
159
13. Paint the town red
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
14. Skate on thin ice
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
15. Go by the book
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
16. Run out of steam
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
160
17. Beat around the bush
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
18. Cast the first stone
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
19. Sing a different tune
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
20. Go against the grain
I have heard or read this idiom:
1. never before
2. once before
3. a few times before
4. several times before
5. many times before
161
APPENDIXD
IDIOM EXPLANATION TASK
Name:
------------------
(first) (last)
Your booklet contains a list of20 idioms. Please explain what you think each idiom
means. Then, describe a situation where someone would use that idiom. You may use
a pencil or a pen to write your answers.
Look at the two examples below:
1. What does it mean to get offthe hook?
It means you try to get out ofsomething that you don't
want to do.
Describe a situation where someone would use that idiom.
A girl was lookingforward to a party on Saturday night.
But then she remembered that she had agreed to baby sit
the neighbor's child that same night. She didn't want
to miss the party. She asked her dad, "How can I get
offthe hook? "
2. What does it mean to turn back the clock?
It means you want to be young again or to go back in
time.
Describe a situation where someone would use that idiom.
A man was watching his granddaughter play basketball
one evening. He remembered a time when he could run
quickly, shoot baskets, and dribble just like his
granddaughter. He felt a little sad. He told his
friend, "I wish I could turn back the clock. "
162
Please answer the following questions by yourself.
1. What does it mean to have a soft spotfor something?
Describe a situation where someone would use that idiom:
2. What does it mean to cast the first stone?
Describe a situation where someone would use that idiom:
163
3. What does it mean to get in someone 's hair?
Describe a situation where someone would use that idiom:
4. What does it mean to skate on thin ice?
Describe a situation where someone would use that idiom:
5. What does it mean to paint the town red?
Describe a situation where someone would use that idiom:
164
6. What does it mean to holdyour head up?
Describe a situation where someone would use that idiom:
7. What does it mean to put one 'sfoot down?
Describe a situation where someone would use that idiom:
8. What does it mean to blow offsteam?
Describe a situation where someone would use that idiom:
165
9. What does it mean to go by the book?
Describe a situation where someone would use that idiom:
10. What does it mean to go through the motions?
Describe a situation where someone would use that idiom:
11. What does it mean for something to go against the grain?
Describe a situation where someone would use that idiom:
166
12. What does it mean to bring the house down?
Describe a situation where someone would use that idiom:
13. What does it mean to get the lead out?
Describe a situation where someone would use that idiom:
14. What does it mean to blow one's own horn?
Describe a situation where someone would use that idiom:
167
15. What does it mean to beat around the bush?
Describe a situation where someone would use that idiom:
16. What does it mean to run out ofsteam?
Describe a situation where someone would use that idiom:
17. What does it mean to keep one's shirt on?
Describe a situation where someone would use that idiom:
168
18. What does it mean to sing a different tune?
Describe a situation where someone would use that idiom:
19. What does it mean to take someone under one's wing?
Describe a situation where someone would use that idiom:
20. What does it mean to turn the other cheek?
Describe a situation where someone would use that idiom:
169
Name:
APPENDIXE
IDIOM COMPREHENSION TASK
170
Introduction:
(first) (last)
Your booklet contains a set ofidioms. Please select the answer that best
explains the meaning ofeach idiom. Draw a circle around the letter of
your choice - A, B, C, or D.
Let's try some for practice.
Practice Problems:
1. What does it mean to get offthe hook?
A. to do many different things
B. to think carefully about a problem
C. to help other people when needed
D. to get out ofa situation
2. What does it mean to turn back the clock?
A. to think about others
B. to be young again
C. to enjoy sporting events
D. to understand a situation
171
Directions:
Now you are ready to answer the rest ofthe questions by yourself. Read the four answer
choices carefully. Ifyou aren't sure ofthe best answer, take a guess. Pick the one that you
think is best. Make sure you don't skip any problems.
1. What does it mean to skate on thin ice?
A. to almost miss something
B. to make a bad decision
C. to be in a dangerous situation
D. to make someone angry
2. What does it mean to go by the book?
A. to ignore suggestions
B. to follow the rules
C. to take a long time
D. to be uncooperative
3. What does it mean to sing a different tune?
A. to change one's mind
B. to act selfishly
C. to request special treatment
D. to argue with others
4. What does it mean to holdyour head up?
A. to be honest
B. to be friendly
C. to be proud
D. to be kind
5. What does it mean to have a soft spotJor something?
A. to feel sorrow
B. to feel affection
C. to feel weakness
D. to feel freedom
6. What does it mean to keep one's shirt on?
A. to think about others
B. to act unselfishly
C. to stay calm
D. to do the right thing
7. What does it mean to run out ofsteam?
A. to feel bored
B. to feel annoyed
C. to feel discouraged
D. to feel tired
8. What does it mean to put one 'sfoot down?
A. to insist on something
B. to act cautiously
C. to be overly concerned
D. to worry about others
9. What does it mean to paint the town red?
A. to relax
B. to celebrate
C. to make plans
D. to spend lots ofmoney
10. What does it mean to bring the house down?
A. to make people listen
B. to make people watch
C. to make people proud
D. to make people applaud
11. What does it mean to beat around the bush?
A. to try to impress someone
B. to show pride in one's work
C. to avoid discussing a topic
D. to talk about one's self
172
12. What does it mean to get in someone 's hair?
A. to be playful
B. to be naughty
C. to be energetic
D. to be annoying
13. What does it mean to get the lead out?
A. to organize others
B. to work faster
C. to complain loudly
D. to make a promise
14. What does it mean to blow one's own horn?
A. to brag too much
B. to be very busy
C. to act unselfishly
D. to try to lead others
15. What does it mean to blow offsteam?
A. to do something fun
B. to feel frustrated
C. to forgive someone
D. to get rid of anger
16. What does it mean for something to go against the grain?
A. to cause discomfort
B. to cause happiness
C. to cause jealousy
D. to cause arguments
17. What does it mean to turn the other cheek?
A. to solve a problem
B. to ignore something
C. to change one's behavior
D. to listen to others
173
18. What does it mean to go through the motions?
A. to be frustrated
B. to day dream
C. to work without interest
D. to behave badly
19. What does it mean to take someone under one's wing?
A. to offer guidance to someone
B. to make a statement about someone
C. to wish someone good luck
D. to show appreciation to someone
20. What does it mean to cast the first stone?
A. to lie about others
B. to be overly helpful
c. to ignore a problem
D. to criticize another
174
APPENDIXF
SCORING EXAMPLES OF IDIOM EXPLANATION TASK REPORTED
BY PARTICIPANTS ACROSS FOUR AGE GROUPS
The following examples are explanations of20 idioms reported by participants across
four age groups.
Score 0 = The response is inaccurate, a restatement, literal, or absent
Score 1 = The response is related but vague or incomplete
Score 2 = The response is accurate, clear, and complete
Idiom Explanation Task-Transparent Idioms
1. Hold one's head up (to be proud, brave, courageous)
0= To stay strong
1 = To maintain your dignity
2 = Be proud ofyourself
2. Go by the book (to follow rules and regulations)
o= To go by the numbers
1 = Too strict not being able to bend
2 = Follow the rules or stated procedure exactly
3. Take someone under one's wing (to offer protection or guidance)
0= No irrelevant or inaccurate responses
1 = Help them out
2 = Take care of someone
4. Turn the other cheek (turn awayfrom anger and ignore it rather than retaliate)
o= Look the other way
1 = Take the punishment that you don't deserve
2 = Not be angry. Do not take revenge
5. Run out ofsteam (to run out ofenergy)
0= Giving the same advice over and over
1 = When you keep working and can't stop because you want to finish
2 = Run out of speed or energy
175
6. Blow offsteam (to get rid ofanger)
o== A person gets angry about a situation
1 == Togo ahead and vent about something that is bothering you
2 == Release anger
7. Skate on thin ice (to take a big risk)
o== This is one ofthe last warning; losing patience
1 == Take precaution
2 == Enter a danger situation
8. Sing a different tune (to change one's mind, opinion, or point ofview)
o== You go against the norm; change your mind
1 == To see things with s different and possibly better point ofview
2 == To change your mind/opinion, usually without waiting to or expecting to
9. Cast the first stone (to criticize another)
0== To be the first to initiate something
1 == You are judging someone
2 == To give the first insult
10. Go through the motions (to pretend to do something by acting as ifone was really
doing it)
o== To go through the routine or procedure
1 == Doing things repeatedly; proceed through tedious, boring strategies.
2 == Do something when you heart is not in it
Idiom Explanation Task- Opaque Idioms
11. Get in someone 's hair (to bother someone)
o== Being too opinionated
1 == Someone not mind their own business and getting unnecessarily personal
2 == To annoy someone
12. Get the lead out (to start moving or move more rapidly)
o== Getting into activity; Remove something or someone who is blocking others
from improving
1 == To get started on something
2 == To hurry
13. Keep one's shirt on (to stay calm, to be patient)
o== Suddenly be upset about a situation
1 == Don't get too excited
2 == To stay calm
14. Blow one's own horn (to brag)
o== To be considered
1 == To commend yourself
2 == Praise yourself
176
15. Put one'sfoot down (to assert something strongly)
0= To try to get something done that's hard to you
1 = To meant what you
2 = Make a final decision, don't change your mind
16. Beat around the bush (to avoid a topic, stallfor time)
0= To procrastinate
1 = Not stating the facts about a situation
2 = Speak about something indirectly
17. Bring the house down (to make others applaud a spectacular performance)
o= Have a fun time
1 = To excite and deliver words ofinspiration to a crowd or audience
2 = Do so well that the audience approves wildly
18. Paint the town red (to go out and celebrate)
o= Going crowed showing offmy accomplishments
1 = Togo out and carefree
2 = To celebrate
19. Have a soft spot in one's heart (to be kind to others or fond ofsomething)
o= A weak spot
1 = Someone is sensitive toward some aspect ofobject or person
2 = You like it more than other things
20. Go against the grain (to do something that is the opposite ofwhat is usually done)
o= A person who does what they want
1 = To rub the wrong way
2 = Something to which you are not accustomed to doing
177
178
APPENDIXG
GOOD MENTAL IMAGES REPORTED BY PARTICIPANTS
ACROSS FOUR AGE GROUPS
The following examples are mental images of20 idioms reported by participants across
four age groups. Mental images that received full score (score = 2) were selected as
examples. The code numbers followed each example indicated the participant's age group
and anonymous code. For example, 20-01 meant that the participant was in 20s Group
and the subject code was 01.
Transparent Idioms
1. Hold one's head up (to be proud, brave, courageous)
A. Jessica did the right thing by not joining in with the cool kids and using pot. She
was embarrassed to see them he next day at school because they had laughed at
her the day before. Her mother said "I am so proud ofyou. When you go to school
today you should hold your head up and be proud." (20-13)
B. Despite falling from the balance beam during the gymnastics finals, Cindy got up
and back on the beam to finish her routine. She held her head up that day. (40-20)
C. The young child suffered from rheumatoid arthritis and was reluctant to start
middle school, afraid that her peers would notice her bump and make fun ofher.
Her mother told her to hold her head up and walk straight and tall into the new
classroom. (60-30)
D. Mother go daughter at school conference with teacher, "Dear, I want you to walk
in there and hold your head up and the proud ofyour heritage" (80-4)
2. Go by the book (to follow rules and regulations)
A. A cop is accused oftaking bribes, but all his co workers and bosses testify on his
behalf. They state on the witness stand, "Bob would never take a bribe. He does
everything by the book" (20-5)
B. You are doing your taxes, tempted to cheat, but decide you better go by the book.
(40-12)
C. Policemen don't have a grey area when it comes to giving parking tickets because
they go by the book (60-17)
D. Two policemen were arguing about a felon they just brought in. One cop wants to
charge him on one count, while the other says he wants to go by the book, "in
which case I will charge him for two counts" (80-4)
179
3. Take someone under one's wing (to offer protection or guidance)
A. A person starting an internship may be taken under the wing oftheir supervisor
who would sow them around, answer questions, make sure they're comfortable.
(20-28)
B. He was new at work so I took him under my wing to show him the right way to
get the job done. (40-3)
C. Claudine had been in her job for 20 years. When a new employee arrived, she
took her under her wing. (60-14)
D. A reasoned and very successful actor decided to take the young aspiring actress
under his wing and funded a scholarship for her at a famous actor's college. (80-9)
4. Turn the other cheek (turn awayfrom anger and ignore it rather than retaliate)
A. An employer heard an employee spread rumors about him. His first impulse was
to write up a citation. He thought better ofit and told his wife at home "I decided
to ignore it and turn the other cheek." (20-5)
B. Sometimes my sister says really sarcastic and rude things and I just turn the other
cheek instead ofarguing with her and I don't say anything, so we don't have a
fight. (40-6)
C. Joey was being called names at school. His mom told him not to call names or
fight, but to treat the other with respect. (60-29)
D. The drunk was insulting the minister but the minister kept on beseeching the man
to stop drinking. (80-16)
5. Run out ofsteam (to run out ofenergy)
A. Students studying for finals decide to stop for the night, expressing that they have
run out ofsteam. (20-22)
B. I was a mile away from the finish line in the marathon then I ran out ofsteam.
(40-3)
C. After getting ready for the garage sale for house, I ran out of steam by the time
buyers started to arrive. (60-14)
D. I said I'd participate in this idiom explanation task. But now it seems I have run
out ofsteam or used up all my energy. (80-20)
6. Blow offsteam (to get rid ofanger)
A. The teenager leaves and complaints to her friend that her mom is not letting her
take the car. She says to her friend "I need to blow off some steam. Let's go
running." (20-5)
B. Because work is so stressful, you tell a coworker that you are going to the gym to
work out and blow off steam. (40-12)
C. After the argument with his partner, the man ran three miles to blow offsome
steam and keep her blood pressure down. (60-30)
D. Mother to son who has just come home in a looking anxious and up tight: "son, I
want you to go to your room and stay there until you blow off steam and clam
down." (80-4)
180
7. Skate on thin ice (to take a big risk)
A. A mother and father set a curfew for their teenage son. He has come home late
three times in one week. His mom or dad may tell him, "You are skating on thin
ice." (20-7)
B. An employee is constantly late to work and was warned 2 times by her boss to call
and let him know but also it is not acceptable. The employer said to the employee
"you're skating on thin ice." (40-4)
C. The young woman was skating on this ice by walking home down a dark alley
rather than following the busy public street. There had a few cases ofwomen
being mugged in the alley. (60-30)
D. In 2006, many investment banks were skating on thin ice as they loaned money to
those who could only pay back ifhousing process skyrocketed. (80-16)
8. Sing a different tune (to change one's mind opinion, or point ofview)
A. Alex cheated on a test and passed. He bragged about it to his friends who told his
teacher. The teacher said 'He's happy about it now, but when he finds out he's
failing the class, he'll be singing a different tune." (20-28)
B. "He thinks I'm too strict with him" said the father, "but when he has children of
his own, he'll sing a different tune." (40-16)
C. The woman said she would never meet a man through online dating sites-but sang
a different tune after her divorce. (60-4)
D. If a president has indicated that he favors universal health care and then changes
his viewpoint to allow it only to the rich, he is singing a different tune. (80-20)
9. Cast the first stone (to criticize another)
A. The judge said to a group ofattorneys "which one ofyou is going to cast the first
stone against this man?" As far as the judge was concerned, the suspect appeared
innocent. (20-6)
B. An example would be ofNewt Gingrich's efforts against Bill Clinton following
his denial ofhaving an inappropriate relationship with an intern while failing to
mention his own marital affair. (40-23)
C. I don't want to be in the position ofcasting the first stone-I would prefer the other
person imitate the confrontation. (60-28)
D. Before you accuse her ofmisbehaving be sure you are free ofmisbehavior before
you cast the first stone. (80-15)
10. Go through the motions (to pretend to do something by acting as ifone was really
doing it)
A. If a person breaks up with her boyfriend, she may go to work the next day and
Xerox papers, make phone calls, go to the meeting, but the whole time she's
thinking ofher boyfriend and she's not really "present" at work. (20-13)
B. A man goes to his office ob the last day ofhis job. He is still required to complete
his job responsibilities for the day, but he is really just going through the motions.
(40-30)
C. The employ pretended to enjoy the group seminar and went through the motions
ofbeing a team player. However, all she could think about was her upcoming trip
to Hawaii. (60-30)
181
D. He really didn't agree with all the rules and regulations, but decided it would be
easier to just go through the motions. (80-14)
Opaque Idioms
11. Get in someone 's hair (to bother someone)
A. The older sister stated that her younger sister was in her hair. The older sister was
trying to study for an exam and her little sister was constantly interrupting her.
(20-6)
B. Whenever George has a disagreement with his wife, his mother always gets in his
hair about how to make things better. (40-15)
C. I could say to my grandchildren who were running all over the kitchen while I was
cooking. "Go outside for awhile. You're getting in my hair." (60-6)
D. Ifa messy person makes a mess in the room of a tidy person, he gets in the hair of
the neat person. (80-20)
12. Get the lead out (to start moving or move more rapidly)
A. In a meeting, the employees were encouraged to work faster and get the lead out.
(20-6)
B. Nancy was taking a very long time with the dinner dishes. Her mother told her,
"Get the lead out! You've been washing that same plate for five minutes already.
" (40-16)
C. The bus is coming and you are about to miss it. Run fast. (60-24)
D. During the race, John started lagging and could not keep up with the others. His
coach urged him to get the lead out. (80-22)
13. Keep one's shirt on (to stay calm, to be patient)
A. A man at a bar takes offense at something he thought someone else said. He
beings yelling at the man. The other man replies "Keep your shirt on. I was not
talking to you." (20-5)
B. An 8 yr old boy is excited for his birthday party and his mom tells him to be
patient and keep his shirt on. (40-14)
C. When he kept pressing me for an answer, I told him to keep his shirt on. I'd get to
it as soon as I could manage. (60-13)
D. The husband waits for his wife to get dressed. He is told to wait with no
complaints. (80-10)
14. Blow one's own horn (to brag)
A. The girl said "I don't want to blow my own horn but I think I got an A on my
math test." (20-9)
B. The proud mom blew her own horn when she mentioned her children won the
school spelling bee. (40-7)
C. The young man irritated his date because be continually bragged and blew his own
horn. He was not interested in hearing about her accomplishments. (60-30)
D. I tell people how beautiful I am, how skilled a person I am. I do this over and over
again. (80-6)
182
15. Put one'sfoot down (to assert something strongly)
A. Ifa child asks for a cookie and his mother says "no," she might say "I'm putting
my foot down." (20-28)
B. My husband and I were not agreeing with how to handle a problem and I felt my
solution was the better one. So I decided to put my foot down and insist we do
what I suggested. (40-6)
C. When children misbehave, someone must put their foot down to show them rules
must be obeyed. (60-4)
D. Your 10 year old is watching TV after 8:00pm. You say the rule in this house is
no TV after 8pm. I'm putting my foot down-turn offTV. (80-29)
16. Beat around the bush (to avoid a topic, stallfor time)
A. When the police questioned Tommy the whereabouts ofMisti, a Las Vegas
showgirl, Tommy beat around the bush without alluding to the fact she was hiding
in the truck ofhis Mazda. (20-30)
B. You want to tell someone that you do not love them, so you try to say a lot ofnice
things first before you get it out. (40-1)
C. Afraid ofrejection, Tom beat around the bush in asking Carol for a date. (60-14)
D. When I asked Sally why the car had a dent in the fender, what she told me had
confusing statements so I couldn't figure it out. (80-19)
17. Bring the house down (to make others applaud a spectacular performance)
A. The pianist's amazing performance brought the house down. (20-27)
B. The band was planning to "Bring down the house" when playing at the concert.
(40-13)
C. The audience was so thrilled by the party that it brought down the house. (60-27)
D. John made a speech about the finer aspects ofpoetry, and brought the house down
with his eloquence and humor. (80-25)
18. Paint the town red (to go out and celebrate)
A. The girls painted the town red when they got all dressed up and went to dinner
and then to a dance club. They were celebrating their friend's engagement. (20-6)
B. I got the job today so I went out to paint the town red to celebrate. (40-3)
C. The friends ofthe bride-to-be rented a limousine and rode her club to club,
painting the town red with their drinking and dancing. (60-30)
D. After winning a close soccer game with a traditional rival, the whole team went
out and painted the town red. (80-16)
19. Have a soft spot in one's heart (to be kind to others orfond ofsomething)
A. A person finds a stray dog or cat that will go to the pound but kept it as a pet
because they have a soft spot for animals. (20-12)
B. A person sitting in a park sees an ice cream truck approach. Remembering
childhood. The person has a soft spot for snow cones. (40-2)
C. I have a soft spot for chocolate. (60-26)
D. Usually he didn't like Irish music but he had a soft spot for certain lines in Danny
boy, so he always gave money to any street musician who played it. (80-16)
183
20. Go against the grain (to do something that is the opposite ofwhat is usually done)
A. No relevant, clear, and complete response for 20s Group
B. No relevant, clear, and complete response for 40s Group
C. It goes against my grain to insult someone in front ofothers. (60-28)
D. Someone doesn't like flattering the boss's wife at a party but does it anyway to get
promoted. (80-16)
APPENDIXH
THE ACCURACY OF EACH IDIOM IN IDIOM COMPREHENSION TASK
ACROSS FOUR AGE GROUPS
A=Accurate; I=Inaccurate
Transparcnt Idioms
20s 40s 60s 80s
(n=30 per group)
Hold one's A 30 (100%) 29 (97%) 30 (100%) 28 (93%)
head up I 0(0%) 1 (3%) 0(0%) 2 (7%)
Go by the A 30 (100%) 30(100%) 30 (100%) 29 (97%)
book I 0(0%) 0(0%) 0(0%) 1 (3%)
Take A 30 (100%) 30 (100%) 30 (100%) 30 (100%)
someone I 0(0%) 0(0%) 0(0%) 0(0%)
under one's
wing
Turn the A 29 (97%) 25 (83%) 24 (80%) 17 (57%)
other cheek I 1 (3%) 5 (17%) 6 (20%) 13 (43%)
Run out of A 28 (93%) 30 (100%) 29 (97%) 27 (90%)
steam I 2 (7%) 0(0%) 1 (3%) 3 (10%)
Blow off A 26 (87%) 23 (77%) 28 (93%)
27 (90%)
stcam I 4 (13%) 7 (23%) 2 (7%)
3 (10%)
Skate on A 21 (70%) 27 (90%)
25 (83%) 24 (80%)
thin ice I 9 (30%) 3 (10%)
5 (17%) 6 (20%)
Sing a A 22 (73%)
24 (80%) 21 (70%) 23 (77%)
different I 8 (27%) 6 (20%)
9 (30%) 7 (23%)
tune
184
185
Cast the A 26 (87%) 29 (97%) 27 (90%)
23 (77%)
first stone I 4 (13%) 1 (3%) 3 (10%)
7 (23%)
Go through A 29 (97%) 29 (97%) 28 (93%) 29 (97%)
the motions I 1 (3%) 1 (3%) 2 (7%) 1 (3%)
Opaque Idioms
Get in A 30 (100%) 30 (100%) 30 (100%) 30 (100%)
someone's I 0(0%) 0(0%) 0(0%) 0(0%)
hair
Get the lead A 15 (50%) 27 (90%) 27 (90%) 29 (97%)
out I 15 (50%) 3 (10%) 3 (10%) 1 (3%)
Keep one's A 27 (90%) 29 (97%) 29 (97%) 29 (97%)
shirt on I 3 (10%) 1 (3%) 1 (3%) 1 (3%)
Blow one's A 30 (100%) 30 (100%) 30 (100%) 30 (100%)
own horn I 0(0%) 0(0%) 0(0%) 0(0%)
Put one's A 30 (100%) 30 (100%) 30 (100%) 30 (100%)
foot down I 0(0%) 0(0%) 0(0%) 0(0%)
Beat around A 30 (100%) 30 (100%) 29 (97%) 29 (97%)
the bush I 0(0%) 0(0%) 1 (3%) 1 (3%)
Bring the A 26 (87%) 27 (90%) 29 (97%) 30 (100%)
house down I 4 (13%) 3 (10%) 1 (3%) 0(0%)
Paint the A 30 (100%) 28 (93%) 30 (100%) 30 (100%)
town red I 0(0%) 2 (7%)
0(0%) 0(0%)
Have a soft A 22 (73%) 19 (63%)
25 (83%) 26 (87%)
spot I 8 (27$) 11 (37%)
5 (17%) 4 (13%)
Go against A 14 (47%) 20 (67%)
18 (60%) 21 (70%)
the grain I 16 (53%)
10 (33%) 12 (40%) 9 (30%)
,..------------- ._----
186
APPENDIX I
THE FREQUENCY OF EACH IDIOM IN IDIOM FAMILIARITY RATING TASK
ACROSS FOUR AGE GROUPS
5=many times; 4=several times; 3=a few times; 2-=once; 1= never
Transparent Idioms
20s 40s 60s 80s
(n=30 per group)
Hold one's 5 25 (84%) 23 (77%) 23 (77%) 15 (50%)
head up 4 4 (13%) 6 (20%) 6 (20%) 12 (40%)
3
1 (3%) 1 (3%) 1 (3%) 3 (10%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Go by the 5 23 (77%) 24 (80%) 25 (83%) 22 (73%)
book 4 7 (23%) 4 (13%) 5 (17%) 5 (17%)
3
0(0%) 2 (7%) 0(0%) 3 (10%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Take 5 25 (84%) 26 (87%) 25 (83%) 17 (57%)
someone 4 4 (13%) 4 (13%) 5 (17%) 8 (27%)
under one's 3
1 (3%) 0(0%) 0(0%) 4 (13%)
wmg 2
0(0%) 0(0%) 0(0%) 1 (3%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Turn the 5 25 (84%) 27 (90%) 27 (90%)
25 (83%)
other cheek 4 4 (13%) 3 (10%) 2 (7%) 5 (17%)
3
1 (3%) 0(0%) 1 (3%) 0(0%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Run out of 5 17 (57%) 20 (67%)
27 (90%) 22 (73%)
steam 4 10 (33%) 8 (26%)
3 (10%) 2 (7%)
3 1 (3%) 2 (7%) 0(0%) 6 (20%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
2 (7%) 0(0%) 0(0%)
0(0%)
187
Blow off 5 27 (90%) 24 (80%)
24 (80%) 23 (77%)
steam 4 2 (7%) 5 (17%)
4 (13%) 5 (17%)
3
1 (3%) 1 (3%) 2 (7%) 1 (3%)
2
0(0%) 0(0%) 0(0%) 1 (3%)
1
0(0%) 0(0%) 0(0%)
0(0%)
Skate on 5 19 (63%) 19 (64%) 23 (77%)
17 (57%)
thin ice 4
6 (20%) 9 (30%) 7 (23%) 10 (33%)
3
5 (17%) 1 (3%) 0(0%) 3 (10%)
2
0(0%) 1 (3%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Sing a 5 13 (43%) 18 (60%) 19 (64%) 15 (50%)
different 4 8 (27%) 6 (20%)
6 (20%) 12 (40%)
tune 3
8 (27%) 5 (17%) 4 (13%) 3 (10%)
2
0(0%) 0(0%) 1 (3%) 0(0%)
1
1 (3%) 1 (3%) 0(0%) 0(0%)
Cast the 5 13 (43%) 22 (74%) 20 (66%) 16 (54%)
first stone 4 7 (23%) 4 (13%)
8 (27%) 10(33%)
3
8 (27%) 3 (10%) 2 (7%) 4 (13%)
2
2 (7%) 1 (3%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Go through 5 20 (68%) 25 (83%) 26 (87%) 19 (64%)
the motions 4 7 (23%) 5 (17%) 4 (13%) 10 (33%)
3
1 (3%) 0(0%) 0(0%) 1 (3%)
2
1 (3%) 0(0%) 0(0%) 0(0%)
1
1 (3%) 0(0%) 0(0%) 0(0%)
Opaque Idioms
Get in 5 12 (40%) 15 (50%) 20 (67%) 13 (44%)
someone's 4 14 (47%) 9 (30%) 7 (23%) 10 (33%)
hair 3
4 (13%) 4 (13%) 2 (7%) 7 (23%)
2
0(0%) 2 (7%) 0(0%) 0(0%)
1
0(0%) 0(0%) 1 (3%) 0(0%)
Get the lead 5 7 (23%) 16 (53%) 23 (77%) 20 (67%)
out 4 4 (13%) 3 (10%) 3 (10%) 8 (27%)
3 7 (23%) 6 (20%) 0(0%) 1 (3%)
2
1 (3%) 2 (7%) 0(0%) 0(0%)
1
11 (38%) 3 (10%) 4 (13%) 1 (3%)
Keep one's 5 10 (34%) 15 (50%) 23 (77%)
22 (74%)
shirt on 4
8 (27%) 5 (17%) 2 (7%) 4 (13%)
3
4 (13%) 9 (30%) 4 (13%) 3 (10%)
2
1 (3%) 1 (3%) 0(0%) 1 (3%)
1
7 (23%) 0(0%) 1 (3%) 0(0%)
188
Blow one's 5 17 (57%) 16(53%) 23 (77%) 23 (77%)
own horn 4 8 (27%)
11 (37%) 4 (13%) 6 (20%)
3
4 (13%) 2 (7%) 3 (10%) 1 (3%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
1 (3%) 1 (3%) 0(0%) 0(0%)
Put one's 5 27 (90%) 28 (93%) 26 (86%) 20 (67%)
foot down 4 3 (10%) 2(7%) 2 (7%) 6 (20%)
3 0(0%) 0(0%) 2 (7%)
4 (13%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Beat around 5 27 (90%) 28 (94%) 28 (93%) 25 (84%)
the bush 4 3 (10%) 1 (3%) 2 (7%) 4 (13%)
3
0(0%) 1 (3%) 0(0%) 1 (3%)
2
0(0%) 0(0%) 0(0%) 0(0%)
1
0(0%) 0(0%) 0(0%) 0(0%)
Bring the 5 12 (40%) 15 (50%) 17 (57%) 16 (54%)
house down 4 11 (37%) 11 (37%) 8 (27%) 9 (30%)
3
5 (17%) 3 (10%) 4 (13%) 4 (13%)
2
2 (6%) 1 (3%) 1 (3%) 0(0%)
1
0(0%) 0(0%) 0(0%) 1 (3%)
Paint the 5 12 (40%) 17 (56%) 25 (84%) 17 (57%)
town red 4 9 (30%) 7 (23%) 4 (13%) 12 (40%)
3 7 (24%) 2 (7%) 1 (3%) 1 (3%)
2
1 (3%) 2 (7%) 0(0%) 0(0%)
1
1 (3%) 2 (7%) 0(0%) 0(0%)
Have a soft 5 28 (93%) 21 (70%) 23 (77%) 17 (57%)
spot 4 0(0%) 7 (23%) 4 (13%) 8 (27%)
3
2 (7%) 2 (7%) 2 (7%) 4 (13%)
2
0(0%) 0(0%) 0(0%) 1 (3%)
1
0(0%) 0(0%) 1 (3%) 0(0%)
Go against 5 15 (50%) 23 (77%)
24 (80%) 18 (60%)
the grain 4 5 (17%) 6 (20%) 5 (17%) 8 (27%)
3 6 (20%) 0(0%) 1 (3%) 4 (13%)
2
3 (10%) 0(0%) 0(0%)
0(0%)
1
1 (3%) 1 (3%) 0(0%) 0(0%)
189
APPENDIXJ
THE ACCURACY OF EACH IDIOM Il\T IDIOM EXPLANATION AND MENTAL
IMAGERY TASK ACROSS FOUR AGE GROUPS
(A=Accurate; V=Vague or Incomplete; I=Inaccurate)
-- -
Idiom Explanation Task Mental Imagery Task
Transparent 20s 40s 60s 80s 20s 40s 60s 80s
Idiom (n=30 per group) (n=30 per group)
Hold one's A 9 (30%) 14 (47%) 20 (67%) 13 (43%) 8 (27%) 17 (57%) 16 (53%) 16 (53%)
head up V 1 (3%) 2 (6%) 3 (10%) 5 (17%) 2 (6%) 0(0%) 6 (20%) 2 (7%)
I 20 (67%) 14 (47%) 7 (23%) 12 (40%) 20 (67&) 13 (43%) 8 (27%) 12 (40%)
Go by the A 28 (94%) 29 (97%) 27 (90%) 29 (97%) 30 (100%) 27 (90%) 26 (86%) 23 (77%)
book V 1 (3%) 1 (3%) 3 (10%) 1 (3%) 0(0%) 1 (3%) 2 (7%) 5 (16%)
I 1 (3%) 0(0%) 0(0%) 0(0%) 0(0%) 2 (7%) 2 (7%) 2 (7%)
Take someone A 28 (93%) 29 (97%) 29 (97%) 29 (97%) 27 (90%) 29 (97%) 27 (90%) 28 (94%)
under one's V 2 (7%) 1 (3%) 1 (3%) 1 (3%) 1 (3%) 0(0%) 3 (10%) 2 (6%)
wmg I 0(0%) 0(0%) 0(0%) 0(0%) 2 (7%) 1 (3%) 0(0%) 0(0%)
Turn the other A 22 (73%) 23 (77%) 25 (83%) 20 (67%) 18 (60%) 25 (83%) 23 (77%) 18 (60%)
cheek V 5 (17%) 4 (13%) 3 (10%) 4 (13%) 3 (10%) 2 (7%) 3 (10%) 1 (3%)
I 3 (10%) 3 (10%) 2 (7%) 6 (20%) 9 (30%) 3 (10%) 4 (13%) 11 (37%)
Run out of A 29 (97%) 30 (100%) 28 (94%) 29 (97%) 28 (94%) 29 (97%) 26 (86%) 26 (86%)
stearn V 0(0%) 0(0%) 1 (3%) 1 (3%) 1 (3%) 0(0%) 2 (7%) 2 (7%)
I 1 (3%) 0(0%) 1 (3%) 0(0%) 1 (3%) 1 (3%) 2 (7%) 2 (7%)
Blow off A 28 (94%) 24 (80%) 17 (57%) 9 (30%) 28 (94%) 25 (83%) 18 (60%) 6 (20%)
stearn V 1 (3%) 2 (7%) 3 (10%) 4 (13%) 1 (3%) 2 (7%) 1 (3%) 1 (3%)
I 1 (3%) 4 (13%) 10(33%) 17 (57%) 1 (3%) 3 (10%) 11 (37%) 23 (77%)
,........
'-0
o
Idiom Explanation Task Mental Imagery Task
Transparent 20s 40s 60s 80s 20s 40s 60s 80s
Idiom (n=30 per group) (n=30 per group)
Skate on thin A 21 (70%) 23 (77%) 18 (60%) 18 (60%) 28 (94%) 23 (77%) 17 (57%) 15 (50%)
Ice V 8 (27%) 4 (13%) 6 (20%) 6 (20%) 1 (3%) 3 (10%) 6 (20%) 2 (7%)
I 1 (3%) 3 (10%) 6 (20%) 6 (20%) 1 (3%) 4 (13%) 7 (23%) 13 (43%)
Sing a A 13 (43%) 16 (53%) 17 (57%) 21 (70%) 13 (43%) 16 (53%) 18 (60%) 17 (57%)
different tune V 2 (7%) 2 (7%) 0(0%) 0(0%) 2 (7%) 2 (7%) 0(0%) 2 (7%)
I 15 (50%) 12 (40%) 13 (43%) 9 (30%) 15 (50%) 12 (40%) 12 (40%) 11 (36%)
Cast the first A 13 (43%) 14 (47%) 15 (50%) 14 (47%) 17 (57%) 14 (47%) 10 (33%) 13 (43%)
stone V 6 (20%) 7 (23%) 6 (20%) 6 (20%) 4 (13%) 5 (16%) 11 (37%) 6 (20%)
I 11 (37%) 9 (30%) 9 (30%) 10 (33%) 9 (30%) 11 (37%) 9 (30%) 11 (37%)
Go through A 18 (60%) 22 (73%) 23 (77%) 17 (57%) 18 (60%) 24 (80%) 22 (73%) 15 (50%)
the motions V 4 (13%) 5 (17%) 3 (10%) 5 (16%) 3 (10%) 2 (7%) 3 (10%) 4 (13%)
I 8 (27%) 3 (10%) 4 (13%) 8 (27%) 9 (30%) 4 (13%) 5 (17%) 11 (37%)
Opaque Idiom
Get in A 29 (97%) 28 (93%) 25 (84%) 29 (97%) 29 (97%) 26 (87%) 21 (70%) 23 (77%)
someone's V 1 (3%) 2 (7%) 4 (13%) o ( 0%) 1 ( 3%) 4 (13%) 5 (17%) 4 (13%)
hair I 0(0%) 0(0%) 1 ( 3%) 1 ( 3%) o ( 0%) o ( 0%) 4 (13%) 3 (10%)
Get the lead A 6 (20%) 20 (67%) 25 (83%) 20 (67%) 6 (20%) 23 (77%) 21 (70%) 16 (53%)
out V 1 (3%) 3 (10%) o ( 0%) 3 (10%) 2 ( 7%) o ( 0%) 2 ( 7%) 6 (20%)
I 23 (77%) 7 (23%) 5 (17%) 7 (23%) 22 (73%) 7 (23%) 7 (23%) 8 (27%)
Keep one's A 21 (70%) 22 (73%) 25 (84%) 21 (70%) 21 (70%) 23 (77%) 20 (66%) 19 (63%)
shirt on V 2 (7%) 5 (17%) 1 ( 3%) 7 (23%) 2 ( 7%) 3 (10%) 5 (17%) 8 (27%)
I 7 (23%) 3 (10%) _~(13%) _~(7~o)_ IJ23%) _ 4 (13%) 5 (17%) 3 (10%)
........
\0
........
Idiom Explanation Task Mental Imagery Task
Opaque 20s 40s 60s 80s 20s 40s 60s 80s
Idiom (n=30 per group) (n=30 per group)
Blow one's A 25 (83%) 27 (90%) 28 (94%) 27 (90%) 27 (90%) 24 (80%) 27 (90%) 23 (77%)
own hom V 3 (10%) 2 ( 7%) 1 ( 3%) 2 ( 7%) 1 ( 3%) 4 (13%) 1 ( 3%) 3 (10%)
I 2 (7%) 1 ( 3%) 1 ( 3%) 1 ( 3%) 2 ( 7%) 2 ( 7%) 2 ( 7%) 4 (13%)
Put one's foot A 26 (87%) 25 (83%) 27 (90%) 25 (83%) 28 (93%) 27 (90%) 25 (83%) 22 (73%)
down V 4 (13%) 3 (10%) 3 (10%) 5 (17%) 2 ( 7%) 1 ( 3%) 2 ( 7%) 5 (17%)
I 0(0%) 2 ( 7%) o ( 0%) o ( 0%) o ( 0%) 2 ( 7%) 3 (10%) 3 (10%)
Beat around A 29 (97%) 25 (83%) 27 (90%) 27 (90%) 27 (90%) 25 (83%) 26 (86%) 17 (57%)
the bush V 0(0%) 3 (10%) 1 ( 3%) 1 ( 3%) 2 ( 7%) 2 ( 7%) 2 ( 7%) 5 (16%)
I 1 (3%) 2 ( 7%) 2 ( 7%) 2 ( 7%) 1 ( 3%) 3 (10%) 2 ( 7%) 8 (27%)
Bring the A 5 (17%) 8 (27%) 18 (60%) 16 (53%) 10 (33%) 18 (60%) 22 (74%) 26 (87%)
house down V 7 (23%) 10 (33%) 6 (20%) 11 (37%) 2 ( 7%) 1 ( 3%) 4 (13% 1 ( 3%)
I 18 (60%) 12 (40%) 6 (20%) 3 (10%) 18 (60%) 11 (37%) 4 (13%) 3 (10%)
Paint the town A 22 (74%) 24 (80%) 27 (90%) 26 (87%) 24 (80%) 26 (86%) 24 (80%) 26 (86%)
red V 4 (13%) 4 (13%) 2 ( 7%) 4 (13%) 2 ( 7%) 2 ( 7%) 3 (10%) 2 ( 7%)
I 4 (13%) 2 ( 7%) 1 ( 3%) o ( 0%) 4 (13%) 2 ( 7%) 3 (10%) 2 ( 7%)
Have a soft A 24 (80%) 19 (64%) 22 (73%) 20 (67%) 25 (83%) 25 (83%) 21 (70%) 22 (73%)
spot V 3 (10%) 7 (23%) 8 (27%) 7 (23%) 3 (10%) 2 ( 7%) 4 (13%) 2 ( 7%)
I 3 (10%) 4 (13%) o ( 0%) 3 (10%) 2 ( 7%) 3 (10%) 5 (17%) 6 (20%)
Go against the A 0(0%) o( 0%) 11 (37%) 9 (30%) o ( 0%) o( 0%) 11 (37%) 10 (33%)
gram V 0(0%) o( 0%) 2 ( 6%) 4 (13%) o ( 0%) o( 0%) o ( 0%) 3 (10%)
I 30 (100%) 30 (100%) 17 (57%) 17 (57%) 30 (100%) 30 (100%) 19 (63%) 17 (57%)
......
\0
tv
193
BIBLIOGRAPHY
Abkarian, 0., Jones, A, & West, 0. (1992). Young children's idiom comprehension:
Trying to get the picture. Journal o/Speech and Dearing Research, 35, 580-587
Anaki, D., Faust, M., & Kravetz, S. (1998). Cerebral hemispheric asymmetries in
processing lexical metaphors. Neuropsychologia, 36(7), 691-700.
Amanzio, M., Geminiani, 0., Leotta, D., & Cappa, S. (2008). Metaphor comprehension in
Alzheimer's disease: Novelty matters. Brain and Language, 107(1), 1-10.
Ardila, A, & Rosselli, M. (1996). Spontaneous language production and aging: Sex and
educational effects. International Journal ofNeuroscience, 87, 71-78.
Atchley, R., Burgess, c., Audet, C., & Arambel, S. (1996). Timecourse, context effects,
and the processing oflexical ambiguity in the cerebral hemispheres. Brain and
Cognition, 30,277-280.
Bayles, K. (1982). Language function in senile dementia. Brain and Language, 16
265-280.
Beeman, M. (1993). Semantic processing in the right hemisphere may contribute to
drawing inferences from discourse. Brain and Language, 44, 80-120.
Belanger, H., & Cimino, C. (2002). The lateralized Stroop: A meta-analysis and its
implications for models ofsemantic processing. Brain & Language, 83(3),
384-402.
Belmore, S. (1981). Age-related changes in processing explicit and implicit language.
Journal ofGerontology, 36(3), 316-322.
Binder, 1., Westbuy, C., McKiernan, K., Possing, E., & Wedler, D. (2005). Distinct brain
system for processing concrete and abstract concepts. Journal ofCognitive
neuroscience, 17(6), 905-917.
Bobrow, S., & Bell, S. (1973). On catching on to idiomatic expressions. Memory and
Cognition, 1, 343-346.
194
Bortfeld, H. (2002). What native and non-native speakers' images for idioms tell us about
figurative language. In R. Heredia & J. Altarriba (Eds.), Advances in psychology:
Bilingual sentence processing (pp. 275-295). North Holland: Elsevier Press.
Boswell, D. (1979). Metaphoric processing in the mature years. Human Development, 22,
373-384.
Bottini, 0., Corcoran, R, Sterzi, R, Paulesu, E., Schenone, P., Scarpa, P., Frackowiak, R,
& Frith, C. (1994). The role ofthe right hemisphere in the interpretation of
figurative aspects of language: A positron emission tomography activation study.
Brain, 117, 1241-1253.
Brundage, S., (1996). Comparison ofproverb interpretations provided by right
hemisphere-damaged adults and adults with probable dementia ofAlzheimer type.
ClinicalAphasiology, 24, 215-231.
Brundage, S., & Brookshire, R (1995). A system for scoring proverb interpretations
provided by non-brain-damaged adults and aphasic adults. Clinical Aphasiology,
23, 165-177.
Burgess, c., & Chiarello, C. (1996). Neurocognitive mechanisms underlying metaphor
comprehension and other figurative language. Metaphor and Symbolic Activity, II,
67-84.
Burgess, C. & Simpson, 0. (1988). Cerebral hemispheric mechanisms in the retrieval of
ambiguous word meanings. Brain and Language, 33,86-103.
Burke, D. (1997). Language, aging and inhibitory deficits: Evaluation ofa theory.
Journal o/Gerontology: Psychological Sciences, 52B(6), 254-264.
Burke, D., & MacKay, D. (1997). Memory, language, and ageing. Philosophical
Transaction ofthe Royal Society, 352, 1845-1856.
Bush, P., & Drummond, S. (1985). Comprehension and production ofidioms in dysphasia.
Archives ofPhysical Medicine and Rehabilitation, 66, 697-700.
Cacciari, C., & Glucksberg S. (1995). Understanding idioms: Do visual images reflect
figurative meaning? European Journal o/Cognitive Psychology, 7(3),283-305.
Cacciari, c., Reatio, E, Colombo, M., Padovani, R., Rizzo, S., & Papagno, C. (2006).
The comprehension of ambiguous idioms in aphasia patients. Neuropsychologia,
44, 1305-1314.
195
Cacciari, C., & P. Tabossi ( 1988). The comprehension of idioms. Journal ofMemory and
Language, 27, 668-683.
Carpenter, P., Miyake, A & Just, M. (1994). Working memory constraints in
comprehension: Evidence from individual differences, aphasia, and aging. In M.
Gemsbacher (Ed.), Handbook ofPsycholinguistics (pp. 1075-1122). San Diego,
CA: Academic Press.
Chapman, S., Ulatowska, H., Franklin, L., Shobe, A, & Thompson, J. (1997). Proverb
interpretation in fluent aphasia and Alzheimer's disease: Implications beyond
abstract thinking. Aphasiology, 11(4), 337-350
Chapman, S., Ulatowska, H., King, K, Johnson, J., & McIntire, D. (1995). Discourse in
early Alzheimer's disease versus normal advanced aging. American Journal of
Speech-Language Pathology, 4, 124-129.
Chapman, S., Zientz, J., Weiner, M., Rosenberg, R, Frawley, W., & Bums, M. (2002).
Discourse changes in early Alzheimer disease, mild cognitive impairment, and
normal aging. Alzheimer Disease andAssociated Disorders, 16(3),177-186.
Chiarello, C., Burgess, C., Richards, L., & Pollock, A (1990). Semantic and associative
priming in the cerebral hemispheres: Some words do, some words don't...
sometimes, some places. Brain andLanguage, 38, 75-104.
Cohen, G. (1979). Language comprehension in old age. Cognitive Psychology, 11,
412-429.
Coulson, S., & Williams, R. (2005). Hemispheric asymmetries and joke
comprehension. Neuropsychologia, 43(1), 128-141.
Cronk, B., & Schweigert, W. (1992). The comprehension ofidioms: The effects of
Familiarity, literalness and usage. Applied P:,ycholinguistics, 13, 131-146.
Duong, A, Tardif, A, & Ska, B. (2003). Discourse about discourse: what is it and how
does it progress in Alzheimer's disease? Brain and cognition, 53(2),177~80.
Duthie, J., Nippold, M., Billow, J., & Mansfield, T. (2008). Mental imagery of concrete
proverbs: A developmental study ofchildren, adolescents, and adults. Applied
P:,ycholinguistics, 29,151-173.
Davis, A, & Ball, H. (1989). Effects ofage on comprehension ofcomplex sentences in
adulthood. Journal ofSpeech andHearing Research, 32, 143-150.
196
DeDe, G., Caplan, D., Kemtes, K., & Waters, G. (2004). The relationship between age,
verbal working memory, and language comprehension. Psychology andAging,
19(4),601-616.
Emery, O. (1986). Linguistic decrement in normal aging. Language and Communication,
6,47-64.
Estill, R, & Kemper, S. (1982). Interpreting idioms. Journal ofPsycholinguistic
Research, 11, 559-581
Feier, C., & Gerstman, L. (1980). Sentence comprehension abilities throughout the adult
life span. Journal ofGerontology, 35, 722-728.
Fiebach, C., & Friederici, A. (2003). Processing concrete words: fMRI evidence against a
specific right hemispheric involvement. Neuropsychologia, 42, 62-70.
Fleming, v., & Harria, J. (2008). Complex discourse production in mild cognitive
impairment: Detecting subtle changes. Aphasiology, 22, 729-740.
Fliessbach, K., Weis, S., Klaver, P., Elger, c., & Weber, B. (2006). The effect ofword
concreteness on recognition memory. Neuroimage. 32(3),1413-1421.
Folstein, M., Folstein, S., & McHugh, P. (1975). Mini-Mental State: A practical method
for grading the cognitive state ofpatients for the clinician. Journal ofPsychiatric
Research, 12, 189-198.
Gibbs, R (1987). Linguistic factors in children's understanding ofidioms. Journal of
Child Language, 14, 569-586.
Gibbs, R (1986). Skating on thin ice: Literal meaning and understanding idioms in
conversation. Discourse Processes, 9(1), 17-30.
Gibbs, R. (1980). Spilling the beans on understanding and memory for idioms in
conversation. Memory & Cognition, 8, 149-156.
Gibbs, R., Bogdanovich, J., Sykes, 1., & Barr, D. (1997). Metaphor in idiom
comprehension. Journal ofMemory and Language, 37, 141-154.
Gibbs, R., & O'Brien, J. (1990). Idioms and mental imagery: The metaphorical
motivation for idiomatic meaning. Cognition, 36, 35-68.
Gibbs, R., Strom, L., Spivey-Knowlton, M. (1997). Conceptual metaphors in mental
imagery for proverbs. Journal ofMental Imagery, 21, 83-110.
197
Giora, R. (1999). On the priority ofsalient meanings: Studies of literal and figurative
language. Journal ojPragmatics, 31, 919-929.
Giora, R. (1997). Understanding figurative and literal language: The graded salience
hypothesis. Cognitive Linguistics, 7(1), 183-206.
Giora, R., & Fein, O. (1999). On understanding familiar and less-familiar figurative
language. Journal ojPragmatics, 31,1601-1618.
Glass, A. (1983). The comprehension ofidioms. Journal ojPsycholinguistic Research,
12, 429-442.
Glosser, G., & Deser, T. (1992). A comparison of changes in macrolinguistic and
microlinguistic aspects ofdiscourse production in normal aging. Journal oj
Gerontology, 47(4), 266-272.
Gregory, M., & Waggoner, J. (1996). Factors that influence metaphor comprehension
skills in adulthood. Experimental Aging Research, 22, 83-98.
Gunter, T., Jackson, J., & Mulder, G. (1995). Language, memory, and aging: An
electrophysiological exploration ofthe N400 during reading ofmemory-demanding
sentences. Psychophysiology, 32, 215-229.
Harris, 1., Rogers, W., & Qualls, c. (1998). Written language comprehension in younger
and older adults. Journal ojSpeech, Language and Hearing Research, 41,603-617.
Hasher, L., & Zacks, R. (1988). Working memory, comprehension, and aging: A review
and new view. In G. H. Bower (Ed.), The psychology ojlearning and motivation
(Vol. 22, pp. 193-225). New York: Academic Press.
Hillert, D. (2004). Spared access ofidiomatic and literal meanings: A single-case
approach. Brain and Language, 89, 207-215.
Huber-Okrainec, J., Blaser, S., & Dennis, M. (2005). Idiom comprehension deficits in
relation to corpus callosum agenesis and hypoplasia in children with spina bifida
meningomyelocele.Brain and Language, 93(3),349-368.
Jessen, E, Heun, R., Erb, M., Granath, D., Klose, U., Papassotiropou10s, A., & Grodd, W.
(2000). The concreteness effect: Evidence for dual coding and context availability.
Brain and Language, 74, 103-112.
Juncos-Rabadan, O. (1996). Narrative Speech in the elderly: Effects ofage and education
on telling stories. International Journal ojBehavioral Development, 19(3),
669-685.
198
Juncos-Rabadan, 0., & Iglesias, F. (1994). Decline in the elderly's language: Evidence
from cross-linguistic data. Journal ofNeurolinguistics, 8(3),183-190.
Kellogg, R, Olive, T., & Piolat, A. (2007). Verbal, visual, and spatial working memory in
written language production. Acta Psychologica, 124, 382-397.
Kemper, S. (1986). Imitation of complex syntactic constructions by elderly adults.
Applied Psycholinguistics, 7, 277-288.
Kemper, S., Greiner. L., Marquis, 1., Prenovost, K., & Mitzner, T. (2001). Language
declines across the life span: Findings from the Nun study. Psychology andAging
16(2),227-239.
Kemper, S., Herman, R, & Lian, C. (2003). Age differences in sentence production.
Journals ofGerontology: Psychological Science, 58, 260-269.
Kemper, S., Herman, R, & Liu, C. (2004). Sentence production by younger and older
adults in controlled contexts. Journals ofGerontology: Psychological Sciences,
58B, 220-224.
Kemper, S., Kynette, D., Rash, S., O'Brien, K., & Sprott, R (1989). Life-span changes to
adults' language: Effects ofmemory and genre. Applied Psycholinguistics, 10,
49-66.
Kemper, S., Thompson, M., & Marquis, 1. (2001). Longitudinal change in language
production: Effects ofaging and dementia on grammatical complexity and
propositional content. Psychology andAging, 16, 600-614.
Kempler, D., & Van Lancker, D. (1993). Acquisition and loss offamiliar language: Idiom
and proverb comprehension. In F. R Eckman (Ed.), Language acquisition and
language disorders (pp. 249-257). Amsterdam: JohnBe~aminsPublishing.
Kempler, D., Van Lancker, D., Marchman, v., & Bates, E. (1999). Idiom comprehension
in children and adults with unilateral brain damage. Developmental Psychology,
15(3),327-349.
Kempler, D., Van Lancker, D., & Read, S. (1988). Proverb and idiom comprehension in
Alzheimer Disease. Alzheimer Disease andAssociated Disorders, 2(1), 38-49.
Kempler, D., & Zelinski, E. (1994). Language function in dementia and normal aging. In
F. A. Huppert, C. Brayne, & D. O'Connor (Eds.), Dementia and normal aging (pp.
331-365). Cambridge, England: Cambridge University Press.
199
Kiefer, M., Weisbrod, M., Kern, 1., Maier, S., & Spitzer, M. (1998). Right hemisphere
activation during indirect semantic priming: evidence from event-related potentials.
Brain and language, 64(3), 377-408.
Kiehl, K., Liddle, P., Smith, A., Mendrek, A., Forster, B., & Hare, R. (1999). Neural
pathways involved in processing concrete and abstract words. Human Brain
Mapping, 7, 225-233.
Kounios, J., & Holcomb, P. (1994). Concreteness effects in semantic processing: ERP
evidence supporting dual-coding theory. Journal ofExperimental Psychology,
20(4),804-823.
KwongSee, S., & Ryan, E. (1995). Cognitive mediation ofadult age differences in
language performance. Psychology andAging, 10(3), 458-468.
Kynette, D. & Kemper, S. (1986). Aging and the loss ofgrammatical forms. Language
and Communication, 6, 65-72.
Laurent, J., Denhieres, G., Passerieux, c., Iakimova, G., & Hardy-Bayle, M. (2006). On
understanding idiomatic language: The salience hypothesis assessed by ERPs.
Brain Research, 1068(1), 151-160.
Libben, M. & Titone, D. (2008). The Multidetermined Nature of Idiom Processing.
Memory and Cognition, 36(6), 1103-1121.
Liu, Y, Xiang, J., Wang, Y, Vannest, J., Byars, A., & Rose, D. (2008). Spatial and
frequency differences of neuromagnetic activities in processing concrete and
abstract words. Brain Topogr, 20(3), 123-129.
Lyons, K., Kemper, S., Ferraro, R., LaBarge, E., Balota, D., & Storandt, M. (1993).
Language andAlzheimer's disease: A reduction in syntactic complexity. Presented
at the annual meeting ofthe American Psychological Society, Chicago, IL.
Mackenzie, C. (2000). Adult spoken discourse: the influences ofage and education. INT.
J Language Communication Disorders, 35(2),269-285.
Mashal, N., Faust, M., Hendler, T., & Jung-Beeman, M. (2008). Hemispheric differences
in processing the literal interpretation ofidioms: converging evidence from
behavioral and fMRI studies. Cortex, 44(7), 848-860.
McDonald, C., Bauer, R., Filoteo, J., Grande, L., Roper, S., Buchanan, R., & Gilmore, R.
(2005). Semantic priming in patients with right frontal lobe lesions. Journal ofthe
International Neuropsychological Society, 11, 132-143.
200
Mellet, E., Tzourio, N., Denis, M., & Mazoyer, B. (1998). Cortical anatomy ofmental
imagery ofconcrete nouns based on their dictionary definition. NeuroReport, 9,
803-808.
Moran, C., Nippold, M., & Gillon, G. (2006). Working memory and proverb
comprehension in adolescents with traumatic brain injury: A preliminary
investigation. Brain Injury, 20(4), 417-423.
Morawski, M. (2009). Idiom processing in aphasic patients. Poznan Studies in
Contemporary Linguistics, 45(2), 245-260.
Myers, P., & Linebaugh, C. (1981). Comprehension ofidiomatic expressions by
right-hemisphere-damaged adults. In R. H. Brookshire (Ed.), Clinical Aphasiology:
Proceedings ofthe Conference (pp. 254-261). Minneapolis, :LVlJ'J: BRK Publishers.
Nenonen, M., Niemi, J., & Laine, M. (2002). Representation and processing ofidioms:
evidence from aphasia. Journal ofNeurolinguistics, 15, 43-58.
Needham, W. (1992). Limits on literal processing during idiom interpretation. Journal of
Psycholinguistic Research, 21, 1-16.
Nippold, M. (2007). Later language development: School-age children, adolescents, and
young adults (3rd ed). Austin, TX: Pro-Ed.
Nippold, M., & Duthie, J. (2003). Mental imagery and idiom comprehension: A
comparison ofschool-age children and adults. Journal ofSpeech, Language and
Hearing Research, 46, 788-799.
Nippold, M., & Rudzinski, M. (1993). Familiarity and transparency in idiom explanation:
a developmental study ofchildren and adolescents. Journal ofspeech and hearing
research, 36(4), 728-737.
Nippold, M., & Taylor, C. (2002). Judgment ofidiom familiarity and transparency: A
comparison of children and adolescents. Journal ofSpeech, Language and Hearing
Research, 45, 384-391.
Nippold, M., Uhden, L., & Schwarz, 1. (1997). Proverb explanation through the lifespan:
A developmental study ofadolescents and adults. Journal ofSpeech, Language and
Hearing Research, 40, 245-253.
Nittono, H., Suehiro, M., & Hori, T. (2002). Word imageability and N400 in an incidental
memory paradigm. International Journal ofPsychophysiology, 44, 219-229.
201
Noppeney, U., & Price, C. (2004). Retrieval of abstract semantics. Neuroimage, 22(1),
164-170.
Nunberg, G., Sag, 1., & Wasow, 1'. (1994). Idioms, Language, 70 (3), 491-538
Obler, L., & Albert, M. (1989). Language decline in aging. ITL. Review ofApplied
Linguistics, 83-84, 63-73.
Obler, L., Fein, D., Nicholas, M., & Albert, M. (1991). Auditory comprehension and
aging: Decline in syntactic processing. Journal ofApplied Psycholinguistics, 12,
433-452.
Oliveri, M., Romero,L., & Papagno, C., (2004). Left but not right temporal involvement
in opaque idiom comprehension: A repetitive transcranial magnetic stimulation
study, Journal ofCognitive Neuroscience, 16,848-855.
Papagno, C. (2001). Comprehension ofmetaphors and idioms in patients with
Alzheimer's disease: A longitudinal study. Brain, 124, 1450-1460.
Papagno, C., Capasso, R, Zerboni, H., & Miceli, G. (2007). A reverse concreteness effect
in a subject with semantic dementia. Brain and Language, 103, 90-91.
Papagno, C., & Caporali, A. (2007). Testing idiom comprehension in aphasic patients:
The effects oftask and idiom type. Brain and language, 100(2),208-220.
Papagno, C., Curti, R, Rizzo, S., Crippa, E, & Colombo, M. (2006). Is the right
hemisphere involved in idiom comprehension? A neuropsychological study.
Neuropsychology, 20(5), 598-606.
Papagno, C., & Genoni, A. (2004). The role of syntactic competence in idiom
comprehension: a study on aphasic patients. Journal ofNeurolinguistics, 17,
371-382.
Papagno, C., Lucchelli, E, Muggia, S., & Rizzo, S. (2003). Idiom comprehension in
Alzheimer's disease: The role ofthe central executive. Brain, 126,2419-2430.
Papagno, c., Tabossi, P., Colombo, M., & Zarnpetti, P. (2004). Idiom comprehension in
aphasic patients. Brain andLanguage, 89, 226-234.
Paivio, A. (1971). Imagery and Verbal Processes. New York: Holt, Rinehart & Winston.
Paivio, A. (1986). Mental representations: a dual coding approach. Oxford, England:
Oxford University Press.
202
Peters, J., & Daum, 1. (2008). Differential effects ofnormal aging on recollection of
concrete and abstract words. Neuropsychology, 22(2),255-261.
Petros, T., Norgaard, L., Olson, K., & Tabor, L. (1989). Effects oftext geme and verbal
ability on adult age differences in sensitivity to text structure. Psychology and
Aging, 4(2), 247-250.
Pexman, P., Hargreaves, 1., Edwards, 1., Hemy, L., & Goodyear, B. (2007). Neural
correlates ofconcreteness in semantic categorization. Journal ofCognitive
Neuroscience, 19, 1407-1419.
Popiel, S., & McRae, K. (1988). The figurative and literal senses ofidioms, or all idioms
are not used equally. Journal ofPsycholinguistic Research, 17,475-487.
Proverbio, A, Crotti, N., Zani, A, & Adomi, R. (2009). The role of left and right
hemispheres in the comprehension ofidiomatic language: An electrical
neuroimaging study. BMC Neuroscience, 10: 116.
Qualls, C., & Harris, J. (2003). Age, working memory, figurative language type, and
reading ability: Influence factors in African American adults' comprehension of
figurative language. American Journal ofSpeech-language-Pathology, 12, 92-102.
Radvansky, G., (1999). Aging, memory and comprehension. Current Directions in
Psychological Science, 8(2), 49-53.
Rassiga, C., Lucchelli, E, Crippa, E, & Papagno, C. (2009). Ambiguous idiom
comprehension in Alzheimer's Disease. Journal ofClinical and Experimental
Neuropsychology, 31(4), 402-411.
Richards, L., & Chiarello, C. (1995). Depth ofassociated activation in the cerebral
hemispheres: Mediated versus direct priming. Neuropsychologia, 33, 171-179.
Richardson, J. (2003). Dual coding versus relational processing in memory for concrete
and abstract words. The European Journal ofCognitive Psychology, 15(4),
481-509.
Richardson, L. (1990) Narrative and sociology. Journal ofContemporary Ethnography,
19(1), 116-135.
Roberts, R, & Kreuz, R (1994). Why do people use figurative language? Psychological
Science, 5(3), 159-163.
Sabsevitz, D., Meddler, D., Seidenberg, M., & Binder, J. (2005). Modulation ofthe
semantic system by word imageability. Neuroimage, 27, 188-200.
203
Sadoski, M. (1995). Concreteness Effects in Text Recall: Dual Coding or Context
Availability? Reading Research Quarterly, 30(2), 278-288.
Sadoski, M., Goetz, E., & Fritz, J. (1993). A causal model ofsentence recall: Effects of
familiarity, concreteness, comprehensibility, and interestingness. Journal of
Reading Behavior, 25(1),5-16.
Sadoski, M., Goetz, E., & Rodriguez, M. (2000). Engaging texts: Effects ofconcreteness
on comprehensibility, interest, and recall in four text types. Journal ofEducational
Psychology, 92(1), 85-95.
Schweigert, W. (1991). The muddy waters ofidiom comprehension. Journal of
Psycholinguistic Research, 20(4), 305-314.
Schweigert, W. (1986). The comprehension offamiliar and less familiar idioms. Journal
ofPsycholinguistic Research, 15(1), 33-45.
Schweigert, W., & Moates, D. (1988). Familiar idiom comprehension. Journal of
Psycholinguistic Research, 17, 281-296.
Shibahara, N., & Lucero-Wagoner, B. (2002). Hemispheric asymmetry in accessing word
meanings: Concrete and abstract nouns. Perceptual and Motor Skills, 94,
1292-1300.
Swaab, T., Baynes, K., & Knight, R. (2002). Separable effects ofpriming and
imageability on word processing: An ERP study. Cognitive Brain Research, 15,
99-103.
Swinney, D., & Cutler, A. (1979). The access and processing of idiomatic expressions.
Journal ofVerbal Learning and Verbal Behavior, 18, 523-534.
Szuchman, L., & Erber, J. (1990). Young and older adults' metaphor interpretation: The
judgements ofprofessionals and nonprofessionals. Experimental Aging Research,
16(2),67-72.
Tabossi, P., Fanari, R., & Kinou, W. (2009). Why are idioms recognized fast? Memory &
Cognition, 37(4), 529-540.
Titone, D., & Connine, C. (1994). Descriptive norms for 171 idiomatic expressions:
Familiarity, compositionality, predictability and literality. Metaphor and Symbolic
Activity, 9, 247-270.
204
Titone, D., Holzman, P., & Levy, D. (2002). Idiom processing in schizophrenia: Literal
implausibility saves the day for idiom priming. Journal ofabnormal psychology,
111(2), 313-320.
Tombaugh, T., & McIntyre, N. (1992). The mini-mental state examination: A
comprehensive review. Journal ofthe American Geriatrics Society, 40(9), 922-935.
Tompkins, c., Boada, R., & McGarry, K. (1992). The access and processing of familiar
idioms by brain-damaged and normally aging adults. Journal ofSpeech and
Hearing Research, 35(3), 626-637.
Towne, R., & Entwisle, L. (1993). Metaphoric comprehension in adolescents with
traumatic brain injury and in adolescents with language learning disability.
Language, Speech, and Hearing Services in Schools, 24, 100-107.
Uekeermann, J., Thoma, P., & Daum, I. (2008). Proverb interpretation changes in aging.
Brain and Cognition, 67, 51-57.
Ulatowska, H., Cannito, M., Hayashi, M., & Flemming, S. (1985). Language abilities in
the elderly. In H. Ulatowska (Ed.), The aging brain: Communication in the elderly
(pp. 125-139). San Diego, CA: College-Hill Press.
Ulatowska, H., Chapman, S., & Johnson, J. (1995). Processing ofproverbs in aphasics
and old-elderly. Clinical Aphasiology, 23, 179-193.
Van Lancker, D., & Kempler, D. (1987). Comprehension of familiar phrases by left- but
not by right-hemisphere damaged patients. Brain and Language, 32, 265-277.
Ventry, I., & Weinstein, B. (1982). The hearing handicap inventory for the elderly: A new
tool. Ear and Hearing, 3, 128-134.
Wapner, W., Hamby, S., & Gardner, H. (1981). The role ofthe right hemisphere in the
apprehension ofcomplex linguistic materials. Brain andLanguage, 14, 15-33.
Waters, G., & Caplan, D. (2005). The relationship between age, processing speed,
working, memory capacity, and language comprehension. Memory, 13,403-413.
West, C., & Holcomb, P. (2000). Imaginal, semantic, and surface-level processing of
concrete and abstract words: An electrophysiological investigation. Journal of
Cognitive Neuroscience, 12(6), 1024-1037.
Wilmer, E., & Gardner, H. (1977). The Comprehension ofmetaphor in brain-damaged
patients. Brain, 100, 717-729.
Zempleni, M., Haverkort, M., Renken, R., & Stowe, L. (2007). Evidence for bilateral
involvement in idiom comprehension: An event-related functional MRI study.
Neurolmage, 34, 1290-1291.
205