Words as Social Tools An Embodied View on Abstract

Concepts

Visit the link below to download the full version of this book:

https://medipdf.com/product/words-as-social-tools-an-embodied-view-on-abstract-c
oncepts/

Click Download Now

From Anna
To Lola, Luca and Francesco, with all my
love, phantasy and energy. To my parents,
Vanna and Gigi, with gratitude
From Ferdinand
To Bettina, Anja and Erik for their
unlimited love and support. To Walter
Huber, Walter Sturm and Klaus Willmes,
who shaped the landscape of German
Clinical Neuropsychology and with whom I
had the honour and the pleasure to work and
to share the common department
Acknowledgments

First of all, we would like to thank Sandro Rubichi for inviting us to write this
book, to the editor for accepting our proposal, and to Martin Fischer for his
accurate, extremely useful, and insightful work as reviewer. A special thank you to
Irene Mittelberg for the many insightful comments and suggestions. Thanks also
to Massimiliano Miatton, Felice Cimatti, Claudia Scorolli, and Carmen Granito,
who read the manuscript and provided us with useful comments, and to Luca
Tummolini for continuous discussions on the topic. Thanks to Olga Capirci,
Gabriele Gianfreda, and Virginia Volterra for discussions and feedback on the sign
language part. Thanks to Laura Barca, Cristina Burani, Cristiano Castelfranchi,
Fabian Chersi, Federico Da Rold, Davide Marocco, Domenico Parisi, Giovanni
Pezzulo, Lucia Riggio, and Corrado Roversi for stimulating discussions on
abstract concepts and words. Thanks as well to all the members, present and past,
of the EMbodied COgnition lab (http://www.emcolab.unibo.it). Thanks to all the
people who, in different ways, supported us.

vii
Contents

1 The Problem of Definition . . . . . . . . . . . . . . . . . . . . . . . ....... 1

1.1 Abstract Concepts and Word Meanings:
How to Define Them? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Abstraction and Abstractness . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Abstractness and the Glue of Language . . . . . . . . . . . . . 4
1.3 Some Hints from Psycholinguistics . . . . . . . . . . . . . . . . . . . . . 7
1.3.1 Concreteness Effect and Its Accounts . . . . . . . . . . . . . . 7
1.3.2 Problems in Selecting Abstract Words:
Perceptual Strength . . . . . . . . . . . . . . . . . . . . ....... 8
1.3.3 The Abstractness of Emotions . . . . . . . . . . . . ....... 9
1.3.4 The Problems Highlighted by Psycholinguistics ....... 11
1.4 Abstract and Concrete Words: Dichotomy,
Continuum or Other? . . . . . . . . . . . . . . . . . . . . . . . . ....... 12
1.5 Conclusion: Definition is a Hard Job . . . . . . . . . . . . . ....... 14
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 15

2 The WAT Proposal and the Role of Language . . . . . . . . . . . . . . . 19

2.1 The WAT Proposal, in a Nutshell . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Some Reasons Why Language is so Important for ACWs . . . . . 22
2.3 What is Crucial in Language? Sounds, Labels, Explanations? . . . 24
2.3.1 Phonology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.3.2 Auditory Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.3 Labels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.4 Explanations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.4 Which Mechanisms? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.5 Conclusion: WAT and the Scaffolding Role of Language. . . . . . 34
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3 Embodied and Hybrid Theories of Abstract

Concepts and Words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2 Grounding in Action of both Concrete
and Abstract Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

ix
x Contents

3.3 Differences in Content Between Concrete

and Abstract Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3.1 Situations and Introspective Properties. . . . . . . . . . . . . . 44
3.3.2 Emotions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4 Metaphors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.5 Multiple Representation View . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.5.1 Representational Pluralism: Dove . . . . . . . . . . . . . . . . . 52
3.5.2 Grounding and Sign Tracking: Jesse Prinz . . . . . . . . . . . 55
3.5.3 Hybrid Models: Distributional and Embodied
Approaches . . . . . . . . . . . . . . . . . . . . . . . . ........ 57
3.6 Conclusions: Many Theories, One Unifying Theory? . ........ 63
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ 64

4 Word Learning and Word Acquisition . . . . . . . . . . . . . . . . . . . . . 71

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2 Social Aspects in Word Learning . . . . . . . . . . . . . . . . . . . . . . 71
4.2.1 Cultural Psychology and Vygotsky . . . . . . . . . . . . . . . . 72
4.2.2 Studies on Testimony . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.2.3 Comparative Studies on Apes and Children . . . . . . . . . . 75
4.3 Embodiment and Statistics in Word Learning . . . . . . . . . . . . . . 77
4.4 Hybrid Approaches of Word Learning . . . . . . . . . . . . . . . . . . . 79
4.5 Age of Acquisition and Modality of Acquisition . . . . . . . . . . . . 82
4.6 Acquisition of Novel Words in Adults:
An Embodied Approach . . . . . . . . . . . . . . . . . . . .......... 83
4.7 Conclusion: A Possible Acquisition Trajectory . . . .......... 90
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... 91

5 What Can Neuroscience Tell Us About Abstract Concepts ...... 95

5.1 Concreteness Effect . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 95
5.2 Reversed Concreteness Effect in Patients with Deep
Dyslexia and Herpes Encephalitis . . . . . . . . . . . . . . . . ...... 96
5.3 Neuroimaging of Abstract and Concrete Concepts
and Mental Imagery. . . . . . . . . . . . . . . . . . . . . . . . . . ...... 97
5.4 Neuroimaging of Abstract and Concrete Concepts
and Emotional Valence . . . . . . . . . . . . . . . . . . . . . . . ...... 102
5.5 Conclusion: Hints from Neuroscience . . . . . . . . . . . . . ...... 104
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 105

6 Language, Languages, and Abstract Concepts. . . . . . . . . . . . . . . . 111

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2 Abstract Concepts and Rich Linguistic Context:
Computational Linguistics Evidence . . . . . . . . . . . . . . . . . . . . 111
Contents xi

6.3 Abstract Concepts and Sign Language: Some Examples

from Italian Sign Language . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.4 Abstract Concepts and Differences Between Languages . . . . . . . 116
6.5 Conclusion: Influence of Language on Abstract Concepts . . . . . 121
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Afterword: A Short Story on Abstract Concepts. . . . . . . . . . . . . . . . . 125

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Author Biography

Anna M. Borghi is an associate professor in Psychology at the University of

Bologna; her research is conducted both at the Department of Psychology of the
University of Bologna and at the Institute of Cognitive Sciences and Technologies,
National Research Council in Rome, Italy. She has been a visiting scholar at
various American universities (University of Chicago, University of Wisconsin-
Madison, Emory University-Atlanta). She is Associate editor of the journal
Frontiers in Cognition and of Psychological Research. She has just finished to
coordinate the FP7 funded ROSSI (Emergence of Communication in Robots
through Sensorimotor and Social Interaction) project. Her research themes include
affordances, categorization, the relationships between concepts and perception and
action, and between language and the motor system, abstract concepts.

Ferdinand Binkofski is professor and chairman for Cognitive Neurology at the

RWTH Aachen University, he is also affiliated with the Institute for Neuroscience
and Medicine at the Research Center Jülich. In Aachen he also heads study courses
in speech and language pathology. Before his move to Aachen he was one of the
managing directors of the Neuro Image Nord neuroimaging center of the north
German universities of Hamburg, Kiel, and Lübeck. He has managed or is still
managing several national and international research networks in behavioral and
cognitive neuroscience. His research is dedicated to higher motor control
and cognitive neurology. He is exploring the link between action and perception
and between the language and the motor system.

xiii
Chapter 1
The Problem of Definition

There is no abstract art. You must always start with something.

Afterward you can remove all traces of reality.
Picasso
I hardly need to abstract things, for each object is unreal
enough already, so unreal that I can only make it real by means
of painting.
Max Beckmann

1.1 Abstract Concepts and Word Meanings: How to Define

Them?

This book is about abstract concepts and abstract word meanings. Humans have the
amazing ability to distinguish objects and entities, forming categories—categories of
prays and predators, of artefacts and living beings, of animals and plants. Concepts
can be seen as the cognitive side of categories (Barsalou et al. 2003), a sort of
‘‘mental glue’’ that links our past with our current experiences in the world (Murphy
2002). Concepts are grounded in our sensory and motor system, that is they reac-
tivate previous experiences with their referents, helping us to act in the environment
in which we are immersed (Barsalou 1999; Borghi 2005; Gallese and Lakoff 2005).
For example, possessing a concept of ‘‘ball’’ helps us predicting what to do when we
see a novel ball. Once formed, many concepts are then typically associated with
names. In this book we will equate concepts with word meanings. This might seem
problematic. Word meanings are typically less broad and more constrained than the
corresponding concepts, since language contributes in rendering the boundaries
between categories more tight (Cangelosi and Parisi 1998; Cangelosi and Harnad
2000); in addition, there exist concepts without a clear linguistic counterpart. In this
sense, even other animals besides humans can possess concepts. At the same time,
however, it is possible that humans ability to categorize is influenced from the start
by the important fact that we are a linguistic species. Furthermore, the influence of
language is so pervasive and literature on human concepts is so influenced by studies
on language that in most cases it is not possible to distinguish between concepts and
word meanings. For these reasons, we will use the term ‘‘concepts’’ and the term
‘‘words’’ to refer to both concepts and word meanings, unless otherwise specified.
This book deals with the marvelous capacity humans have to form not only
concrete but also abstract concepts, as well as to use abstract word meanings. As the
quotes at the beginning of the chapter suggest, defining abstract concepts—as
abstract art—is not an easy matter, and it is a controversial one. With abstract

A. M. Borghi and F. Binkofski, Words as Social Tools: An Embodied View 1

on Abstract Concepts, SpringerBriefs in Cognition,
DOI: 10.1007/978-1-4614-9539-0_1,  The Author(s) 2014
2 1 The Problem of Definition

concepts and word meanings we refer to concepts mediated by words as ‘‘freedom’’

and ‘‘truth’’. To clarify: we do not intend to claim that words are abstract. All words
are material and perceivable—for example, they can be heard, pronounced or read.
When we use the term ‘‘abstract words’’, in keeping with the literature, we refer to
the fact that their meaning is abstract. There exist indeed words the referents of
which are not material, perceivable, concrete objects and entities, as ‘‘balls’’ and
‘‘cats’’. Their referents are instead mental states, events, conditions, as ‘‘thought’’,
‘‘justice’’, ‘‘totalitarism’’. This book will deal with them. Notice that not all abstract
concepts are abstract in the same way but that there might be different degrees of
abstractness. As argued by Larry Barsalou (2003), concepts become increasingly
abstract (e.g., from ‘‘pen’’ to ‘‘truth’’) as they become more detached from physical
entities, and more associated with mental states.
Below we list the main characteristics that, in our view, characterize abstract
concepts.
Different grounding Abstract concepts are not grounded in physical entities and
in concrete, single objects, as concrete concepts. This does not imply at all that
abstract concepts are not grounded or that their referents are not material, since
they might be grounded in situations, events, mental states, and in complex
relations between objects etc. Notice however that there is always a continuum
between abstract and concrete concepts. Very concrete concepts have abstract
aspects and vice versa. Consider for example the notion of ‘‘penny’’: it refers to
a concrete, manipulable element, but it has some abstract properties, as the
possibility to be exchanged, the value it is attached to it, etc. Furthermore,
words that refer to concrete concepts can also have a metaphorical meaning,
beyond the literal one. For example, we can speak of a mental ‘‘journey’’.
Complexity Abstract concepts are more complex than concrete ones. As argued
by Larry Barsalou: ‘‘abstract concepts often capture complex configurations of
physical and mental events’’ (Barsalou 2003, p. 1185), i.e. they evoke spatial,
temporal and causal relations. This definition relies on data showing that
abstract concepts evoke properties and relations more than objects and events
(e.g., Wiemer-Hastings and Xu 2005).
Meaning variability Abstract concepts meaning is highly changeable compared
to the meaning of concrete words. It is much easier to gain consensus on what
‘‘bottle’’ means and evoke, than on what ‘‘truth’’ means and evoke. In addition,
an abstract concept such as ‘‘truth’’ is more exposed to the different experiences
compared to a concept such as ‘‘bottle’’. This does not mean that ‘‘bottle’’ is a
stable concept while truth is not, since both concepts are continuously updated
and filled by new experiences with the category members. However, the
meaning of abstract concepts is more variable and less stable, both across
subjects and within the same subject.

So far we have introduced those that, in our opinion, are the main character-
istics of abstract concepts and words. In the next paragraphs we will further deal
with the problem of definition, trying to elucidate which concepts can be classified
1.1 Abstract Concepts and Word Meanings: How to Define Them? 3

as abstract in our view. First, we will distinguish between superordinate concepts

and abstract concepts, clarifying that this book does not focus on abstraction but
rather on abstractness. A major part of the chapter is dedicated to the analysis of
how psycholinguistic studies have addressed the problem of defining abstract
concepts and words: we will describe the criteria that have been proposed to
identify abstract words, and then we will discuss whether emotional terms can be
considered as abstract or not. Finally we will discuss whether the distinction
between concrete and abstract concepts can be considered as a dichotomy, as a
continuum or whether more fine-grained analyses of the different kinds of abstract
concepts are necessary.

1.2 Abstraction and Abstractness

A first question that might rise is the following: Would words as ‘‘animal’’ or as
‘‘artifact’’ be considered as abstract ones? Animals and artifacts might come in a
great variety—foxes, robins and penguins do not have much in common, and
neither do chairs and screwdrivers. In addition, the superordinate term ‘‘animal’’ is
certainly more abstract than the basic term ‘‘dog’’, since it refers to a collection of
rather diverse exemplars. But even the subordinate term ‘‘cocker’’ can be con-
sidered as partially abstract, since it abstracts and extracts some common char-
acteristics from the experience of different cockers.
The definition we proposed might appear only as a negative one. It seems that
we clarify what abstract concepts are NOT, not what they are. For these reasons, in
order to answer to the question above and to provide a positive definition it is
important to distinguish between abstraction and abstractness.
Abstraction is the process by means of which ‘‘knowledge of a specific category
has been abstracted out of the buzzing and blooming confusion of experience’’ by
forming a summative representation of that experience (Barsalou 2003: 389). This
form of abstraction is at the core of every form of categorization, since it regards
both concrete concepts, i.e. concepts endowed with perceivable referents, and
abstract concepts. While forming each category, indeed, we somehow ‘‘abstract’’
from single instances and specific experiences. Even a subordinate-level category,
such as ‘‘cocker’’, abstracts from single instances of dogs, and obviously super-
ordinate level categories such as ‘‘animal’’ abstract more than concrete ones.
Abstractness is sometimes conflated with abstraction. However, we intend to
keep abstraction and abstractness separate as much as possible, and this book will
focus on the last one. An example will clarify the reasons of this choice. Concepts
as ‘‘animal’’ and ‘‘furniture’’ (on top of the abstraction hierarchy) are more abstract
than ‘‘dog’’ and ‘‘chair’’, but their category members are all concrete instances.
Concepts as ‘‘freedom’’ and ‘‘phantasy’’, instead, are not abstract because they are
on top on a conceptual hierarchy, but because their referent/s are not concrete
objects or entities: they are not visible, manipulable or perceivable through any of
our senses. This does not imply that they are not grounded in our sensorimotor
4 1 The Problem of Definition

system. Take for example the concept of ‘‘phantasy’’: it is grounded since it refers
to a sparse collection of elements, as it evokes situations, events, and it likely elicits
internal introspective states. A further difference that might exist between concepts
on top of an abstraction hierarchy, such as ‘‘animal’’ or ‘‘vehicle’’, and abstract ones
such as ‘‘freedom’’, concerns quantification: we can easily count animals and
vehicles, while the amount of freedom or of truth might vary. This distinction based
on the easiness of quantification is difficult to operationalize, and further research is
needed in order to verify to what extent it applies to all subsets of abstract concepts
(numbers, for example, are a kind of abstract concepts that can obviously be taken
as quantifiable, even if they belong to a different symbolic system. However, we
will see that numbers represent a special kind of abstract concepts).
One of the main reasons why we intend to focus on abstractness rather than on
abstraction is a theoretical one. Explaining abstract concepts and words (abstract-
ness) constitutes a major challenge for embodied and grounded theories of cogni-
tion (EGC) (Barsalou 1999, 2003, 2008; Borghi and Pecher 2011, 2012; Borghi and
Caruana (in press); Gallese and Lakoff 2005; Myachykov et al. 2013). EGC theories
assume that our bodily characteristics constrain our cognitive processes, from
perception and action to processes traditionally considered as ‘high level’ processes
such as language and thought. Many scientists nowadays recognize that these
theories are rather powerful in explaining conceptual representation, as the burst of
recent evidence on activation of perception and action while processing concrete
concepts and words has shown. However, the number of skeptics is rather con-
sistent when abstract concepts, such as ‘‘truth’’, ‘‘phantasy’’ and ‘‘justice’’, come
into play. One of the aims of this book is to propose a theory of abstract concepts
and words in keeping with an embodied and grounded perspective of cognition.

1.2.1 Abstractness and the Glue of Language

In this book we will propose and defend the view that, the more concepts increase in
abstractness, the more they need some sort of glue that keeps the different category
members together (Murphy 2002). This glue in our view is relevant for all concepts,
but it becomes highly important for concepts on top of the abstraction hierarchy, as
superordinate terms (e.g., ‘‘plants’’), and it is particularly crucial for abstract con-
cepts (e.g., ‘‘thought’’). This also raises the question of whether abstract concepts
have a prototypical and a hierarchical structure like concrete concepts (animal-
dog-cocker). Probably they do not, because they are more a concept than a category
with prototypical examples and superordinate and subordinate members.
Let us consider first concepts that do differ in abstraction but not in degree of
abstractness, as subordinate and superordinate terms (e.g., ‘‘siamese cat’’, ‘‘ani-
mal’’). The glue that keeps together different members of superordinate categories
can be given by the presence of a common context where different category
members can be found. Murphy and Wisniewski (1989) have demonstrated that
1.2 Abstraction and Abstractness 5

Fig. 1.1 Study by Borghi et al. (2005). The interaction between concepts and locations. In a
location verification task with words superordinate concepts (e.g., fruit) yelded faster response
times with broader locations (scenes) (e.g., countryside, for fruit/orange), which could contain
more exemplars, while basic concepts (e.g., orange) yelded faster responses with object locations
(e.g., basket, for fruit/orange). The result suggests that the context can be a sort of glue linking
together different exemplars of a category

the role of context is more relevant for the recognition of superordinate concepts
(e.g., ‘‘musical instruments’’) than of basic level ones (e.g., ‘‘guitar’’). Participants
were presented with a basic or superordinate category name, followed by a picture
of a scene where the depicted objects could either fit in or not. Afterwards, a dot
was presented in the same position as that in which there was an object in the
scene; the task consisted in judging whether the cued object was a member of the
named category or not. Inappropriate scenes affected participants’ decision more
often when the name referred to superordinate than to basic level concepts. Borghi
et al.(2005) (see Fig. 1.1) have shown with both a location production task and
asking participants to evaluate the adequacy of different contexts to the category
members that superordinate concepts elicit broader contexts, where many category
members can coexist. In a similar vein, Kalénine et al. (2009) found with a
categorization task submitted to both children and adults that the advantage of
basic level over superordinate level concepts was reduced when a target-image
was preceded by a scene context rather than by an action priming context (see
Fig. 1.2).
Overall, this evidence suggests that superordinate concepts are represented as a
collection of exemplars, and a common context helps understanding their com-
munalities. In contrast, the context is not so necessary to be able to recognize
single instances of basic level categories, such as dogs or chairs.
One further powerful glue can be given by language. Having one common label
helps keep together sparse situations, experiences, mental states. This is true for
words of all hierarchical levels, from subordinate (e.g., ‘‘cocker’’) to superordinate
level (e.g., ‘‘animal’’), even if it is more crucial for more general terms, whose
6 1 The Problem of Definition

Fig. 1.2 Study by Kalénine et al. (2009). Children aged 7 and 9 and adults performed a
categorization task with basic and superordinate terms following the form ‘‘a kind of…?’’ The
target image was preceded by an action prime, consisting in a photograph of a hand in a grasping
posture, or by a context prime, consisting in a photograph representing a scene. As it can be seen
in the graph, irrespective of age results showed that the advantage of the basic-level task over the
superordinate one was greater in the hand priming condition than in the context priming
condition. This supports the idea that context works as a kind of glue linking together different
exemplars of a category, and that this is true in particular for superordinate level concepts given
the diversity of their category members

members are not characterized by high perceptual similarities. Similarly, such a

glue is important for both concrete (e.g., ‘‘pen’’) and abstract words (e.g., ‘‘free-
dom’’), even if having a common label is particularly important for abstract words.
To summarize: All concepts, even specific subordinate level ones, abstract to
some extent from the single instances of the category: for example, a cocker is not
our Fufi. However, having a common label is more crucial for general, superor-
dinate level terms, which hold together a variety of different members: for
example, ‘‘animal’’ includes cockers and other dogs, cats, birds, fishes, etc. And
possessing a unifying label becomes critical for abstract terms, such as ‘‘truth’’,
‘‘knowledge’’ and ‘‘tolerance’’, since the exemplars of abstract concepts are even
more sparse and diverse. The role language plays for abstractness will be further
developed in the next chapters.
1.3 Some Hints from Psycholinguistics 7

1.3 Some Hints from Psycholinguistics

A wide psycholinguistics literature has addressed the problem of concrete and

abstract words representation. In this paragraph we will focus on this literature, in
particular on how psycholinguists distinguish between concrete and abstract
words. Referring to this kind of research helps us understand the difficulties in
defining concrete and abstract concepts and in operationalizing the two constructs.
At the same time, we are convinced that this literature can give us some hints and
cues on how to interpret current evidence.

1.3.1 Concreteness Effect and Its Accounts

A difficult problem psycholinguistic studies had to address concerns the selection

of abstract and concrete terms. Psycholinguistics researchers usually select and
distinguish concrete and abstract words on the basis of ratings. For example,
people are submitted to huge samples of words and are required to rate them on 7
point scales on a variety of dimensions. Aside from the more obvious dimension of
concreteness, the typically used dimensions are imageability and context
availability.
The choice of testing the dimensions of imageability and context availability in
order to select abstract words is due to theoretical reasons. Two of the most influential
explanations of the so-called ‘‘concreteness effect’’ rely indeed on imageability and
on contextual availability. The ‘‘concreteness effect’’ consists in the processing
advantage and best recall of concrete over abstract words (Schwanenflugel et al.
1988; Schwanenflugel and Shoben 1983; Schwanenflugel and Stowe 1989); this
effect is rather consistent: it was found in both children and adults, and in a variety of
tasks, such as lexical decision, free recall, recognition, and language comprehension
(for a brief review see Altarriba and Bauer 2004; Altarriba et al. 1999). However,
there is some controversy since some studies have not found evidence of concrete-
ness effects (e.g., Barca et al. 2002) and an advantage in processing of abstract over
concrete concepts has been reported as well, both by Kousta et al. (2011) and Barber
et al. (2013), who controlled for items valence, and by Pexman et al. (2007) in an
edible/non edible categorization task.
According to the classical Contextual Availability Theory (CAT; Schwanenflugel
et al. 1988, 1992), while concrete concepts and words are strongly associated to a
reduced numbers of contexts, abstract concepts and words are only weakly associ-
ated to a higher number of contexts. Hence, when words are presented in isolation,
contextual information is more easily accessible for concrete than for abstract words.
This determines the worse recall of abstract compared to concrete words.
This theory has not been confirmed by some recent evidence. For example, it
has been shown that imageability and contextual relatedness influence memory
tasks in an independent fashion, not in an integrated way, as the CAT would
predict.