See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/215778026

Psycholinguistic models of speech production in Bilingualism and

Multilingualism

Chapter · January 2007

CITATIONS READS

9 5,634

2 authors:

Andrea Marini Franco Fabbro

University of Udine University of Udine
124 PUBLICATIONS   1,807 CITATIONS    271 PUBLICATIONS   5,025 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

An analysis of language skills in children with developmental disorders View project

Language Impairments in Parkinson's Disease View project

All content following this page was uploaded by Andrea Marini on 19 May 2014.

The user has requested enhancement of the downloaded file.

In: Speech and Language Disorders in Bilinguals ISBN 978-1-60021-560-5
Editors: Alfredo Ardila and Eliana Ramos, pp: 47-67 © 2007 Nova Science Publishers, Inc.

Chapter 3

PSYCHOLINGUISTIC MODELS OF SPEECH

PRODUCTION IN BILINGUALISM AND
MULTILINGUALISM

Andrea Marini1,2,3 and Franco Fabbro1,2

1. University of Udine, Udine, Italy
2. IRCSS “E. Medea” “La Nostra Famiglia”, San Vito al Tagliamento (UD), Italy
3. IRCCS “Santa Lucia”, Roma, Italy

INTRODUCTION
Bilingualism is a complex psychological and socio-cultural phenomenon encompassing a
number of individual and social dimensions (Butler and Hakuta, 2004). It constitutes a
widespread phenomenon, as in the world approximately 7000 languages are spoken in just
160 countries. Therefore over 50% of the world population is bi- or even multilingual
(Tucker, 1998; Grosjean, 1982, 1994). Moreover, in many countries the number of spoken
languages is constantly growing due to massive immigration.
The issue of language representation and use in multilingual speakers and the related
problem of multilingual competence can be approached at different levels of description.
Indeed, they constitute a matter of interest not only for sociolinguistic models of language use
and stratification, but also for psycholinguistic theories of language development and
functioning, as well as neurolinguistic models of language representation in the brain.
In this chapter some of the most influential psycholinguistic models of language
representation and speech production in multilingual speakers are introduced. The results
from psycholinguistic and neurolinguistic investigations will be taken into account in order to
discuss the assumptions of such models.
2 Andrea Marini and Franco Fabbro

MODELS OF LANGUAGE REPRESENTATION IN MONOLINGUAL

INDIVIDUALS
Psycholinguistics investigates the functional architecture of language and the processes
involved in both oral and written language production and comprehension. Language is
considered a complex cognitive function involving interaction of a number of different levels
of processing: a word level for lexical processing; a sentence level for syntactic processing; a
pragmatic level where words or sentences are contextualized and inferences are drawn; a
text/discourse level where those sentences that constitute a written text or a spoken discourse
are integrated in order to retrieve its general meaning or gist (Marini et al., 2005a, b; Marini
and Nocentini, 2003; Caplan, 1992; Kintsch and van Dijk, 1978). Lexical and syntactic
processing form the within-sentence or microlinguistic dimension of language responsible for
intra-phrasal functions. Lexical processing organizes phonological or graphemical patterns
into morphological strings and words, and determines the morphosyntactic context each word
requires. All these aspects of lexical knowledge are assumed to be represented in a mental
lexicon where phonetic, phonological, morphophonological, morphological, morphosyntactic
and semantic information are stored for each word. Syntactic processing organizes the
morphosyntactic contexts required by words in phrases and clauses in order to generate well-
formed sentences. Pragmatic and discourse processing form the between-sentence or
macrolinguistic dimension of language. They are responsible for inter-phrasal functions and
utterance contextualization.
Each aspect of language processing is subserved by neurofunctionally distinct systems
that can be separately compromised, inhibited or preserved following a brain lesion or after
transient inhibition due to electrical stimulation, paroxystic charge or pharmacological
inhibition (e.g. Ojemann, 1991). According to the Declarative/Procedural model (Paradis,
1994; 2004; Ullman, 2001; 2004), language learning and representation are assumed to be
subserved by two anatomically and functionally distinct long-term memory systems
(declarative and procedural, respectively). Procedural memory is a type of implicit memory
subserved by frontal/basal ganglia circuits as well as portions of the parietal cortex, superior
temporal cortex and the cerebellum (Fabbro, 1999; Ullman, 2001). Procedural memory
underlies implicit linguistic competence. In the course of first language acquisition, it is
involved in the process of learning and consequently executing sensori-motor and cognitive
skills such as, for example, those involved in the articulation of the sounds of a language (i.e.
phones) and in syntax. Declarative (or explicit) memory is subserved by bilateral medial and
temporo-parietal structures, including the hippocampal region and the parahippocampal
cortex (Fabbro, 1999; Ullman, 2001). It is implicated in conscious learning of facts and
events and consists of two subtypes, semantic and episodic memory, respectively. Semantic
memory is the system storing one’s encyclopedic knowledge of the world (e.g. knowledge
about the meaning of words, as well as knowledge about historical, geographical, social
facts). Episodic (or autobiographical) memory refers to one’s past experiences that can be
consciously recalled. It is assumed that grammar (i.e. syntactic and morphosyntactic implicit
competence) is acquired incidentally through procedural memory, whereas lexical-semantic
explicit knowledge is overtly learned and stored in declarative memory (Paradis, 1994;
Ullman, 2001; Fabbro, 1999; 2001). For example, procedural memory acquires and applies
implicit syntactic rules such as principles of government and binding (Chomsky, 1988; 1995;
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 3

Pinker, 1999) or procedures of syntactic parsing (Frazier and Fodor, 1978), but it is
declarative memory that deals with lexico-semantic aspects of the words the form the
sentences. Furthermore, the procedural system “learns” and applies word formation rules to
morphemes and lexical items stored in declarative memory. Moreover, once the access to a
lexical item is granted, implicit memory procedures automatically generate the argument
structure of that particular word and assign the thematic roles to the required arguments
(morphosyntactic processing). As to phonetics, the articulatory programs necessary to
produce the target phones of a language become automatized and are transferred to
procedural memory, so that the speaker does not have to think about the articulatory
movements while speaking.
According to Levelt (1989), the process of speech production is mediated by the central
role of the mental lexicon and involves three subsystems: a pre-linguistic conceptual system
(conceptualizer); a linguistic system (formulator); an output system (articulator) where
production actually takes place. In the conceptualizer the speaker retrieves from long-term
declarative memory all available data to formulate a mental model of the intended message.
Indeed, declarative memory provides knowledge about what to say (message planning), what
has previously been said (linguistic context) and the particular situation, place and time in
which the communicative exchange takes place (Marini, 2001; Levinson, 1983; Johnson-
Laird, 1980). During the phase of message planning, a supervisory attentional system (SAS:
Green, 1998) is in charge of controlling the speakers’ communicative intentions. Moreover,
such system modulates the amount of information that the speaker intends to communicate
and its relevance with respect to what has previously been said (Grice, 1975). This conceptual
information is then sent to the formulator where the preverbal message is converted into a
speech plan. At this stage, lexical processing matches the intended meaning formulated in the
pre-linguistic phase with the corresponding lexical items stored in the mental lexicon. This
match is performed through a multi-stage process which entails a phase of lexical selection
and one of lexical access. The process of lexical selection, allows speakers to select the
lexical items that correspond to the intended meaning (Levelt, 1989; Levelt, Roelofs and
Meyer, 1999). How exactly this selection takes place is still debated, but it is likely achieved
through an activation/inhibition mechanism (Green, 1986; 1998; Paradis, 1989; 1993). Each
word is supposed to have its own specific activation thresholds as a function of its frequency
of use and time elapsed since last activation (Luria, 1973; Paradis, 1989; 1993): The lower the
threshold level, the easier the access; The higher the threshold level, the more difficult the
access. In Green’s theory, the activation of the target word is achieved through the co-
occurring inhibition of its semantically-related competitors. Such inhibition would be
obtained by raising the competitors’ activation thresholds. For example, if the speaker’s
intention is to speak out the word “banana” (e.g. in a picture naming task), the activated
concept corresponding to the idea of “banana” enters the lexicon where a selection
mechanism is needed in order to select the target word (“banana”) among all other
semantically related lexical items (“pear”, “orange”, “apple”, etc…). Consequently, the
activation threshold of the competitors is raised while the target word (“banana”) gets
selected. At the end of the lexical selection process, the target word has been activated and the
formulator gains access to its semantic, morphosyntactic and morphological features (lemma
level of word representation) and then to its phonological form (lexeme level of word
representation). In case of single word production (e.g. picture naming tasks) the lexical
information is then transmitted to the output system where articulatory configurations
4 Andrea Marini and Franco Fabbro

corresponding to the phonemes to be uttered are programmed and then implemented. In case
of sentence production, the process of lexical selection and the access to a word’s lemma
form the “functional level” of sentence processing, where the morphosyntactic structures
requested by the selected lexical item (i.e. its argumental structure) guides the process of
sentence generation by means of thematic roles assignment and phrase generation (noun
phrase [NP], verbal phrase [VP], adjective phrase [AP], prepositional phrase [PP]). At the
second level of sentence processing, the “positional level”, the information contained in the
lemmas of the selected lexical items is used to generate the grammatical relations among the
phrases and to build up well-formed syntactic representations (Chomsky, 1988; 1995; Pinker,
1999; Caplan, 1992). It is now possible to access the syllabic and phonologic representations
of the lexical items (their lexemes) and this information is sent to the output system where
articulatory configurations corresponding to the phonemes to be uttered are programmed and
then implemented.

WHAT IS MULTILINGUALISM?
There is no accepted definition of bilingualism and/or multilingualism among researchers
as our intuitive knowledge of “bilingualism” falls short of a precise definition. According to
Bloomfield (1933), bilinguals are those people who have a native-like control of two
languages. We might therefore extend such interpretation to multilingual speakers, suggesting
to consider multilinguals those people who have native-like control of more than two
languages. However, such a rigid account of bilingualism and multilingualism introduces to
the problem of the definition of quite abstract notions such as “balanced”, “ideal” or even
“perfect” bilingualism. Who is a perfect bilingual or multilingual? It is no doubt hard to find
completely balanced or “perfect” bilinguals. When we turn to consider multilinguals, the
issue becomes even more complex because we should be obliged to exclude from the count of
multilingual speakers all those people who know three or more languages but simply not as
much as they know their mother tongue (L1). Such a view of bilingualism (and consequently
multilingualism) has been fiercely challenged by a number of studies. Rather, the focus has
been shifted toward the degree of proficiency in two or more languages within the same
individual. According to Weinreich (1953), a bilingual is a person who can use alternatively
two languages to communicate, whereas Haugen (1953) introduced the notion of bilingualism
as the simple ability to produce well-formed meaningful utterances in a second language (L2).
According to Grosjean (1989; 1994; 1995; 1999), bilinguals are those people who use two or
more languages in their every-day lives. However, not all multilingual speakers use their two
languages all the time. Therefore an additional distinction must be introduced between
multilinguals who use their languages daily and dormant multilinguals who know more than
one language but use only one of them in their ordinary communicative interactions.
Furthermore, Grosjean (1989) points out that bilinguals must not be considered as the sum of
two complete or incomplete monolinguals, because the interaction of two or more linguistic
competences in the same person produces a different but complete linguistic entity. It is an
integrated whole with its own characteristics that cannot be simply reduced to the sum of the
parts. Indeed, multilingual individuals can acquire their languages at different times and in
different ways. Furthermore, they can use them with different people in different situations.
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 5

Therefore, bilinguals find themselves at various points of “a situational continuum which

induce a particular speech mode. At one end of this continuum, bilinguals are in a totally
monolingual speech mode … At the other end of the continuum, they are with bilinguals who
share their two languages … and with whom they normally mix languages (code-switch and
borrow): they are here in a bilingual speech mode … we should keep in mind that
intermediary modes exist between the two” (Grosjean, 1989, 8).
Results from neuropsychological and neuroimaging investigations suggest that also at
the neuroanatomical level the languages spoken by multilingual individuals have a complex
neurofunctional organization (Fabbro, 2001). For example, several studies on bilingual or
polyglot aphasic patients provide indirect arguments in favour of the hypothesis that in
multilingual speakers the different languages are represented in distinct brain regions. Indeed,
in some cases polyglot patients loose only one of the languages mastered before the insult
(Albert and Obler, 1978; Paradis, 1995) and the different languages can be recovered at
different degrees (Paradis, 1978). Neurosurgical studies provide further evidence for
distributed cortical organization of the languages in multilingual speakers (Gomez-Tortosa et
al., 1995). However, electrical stimulation and neuroimaging studies suggest that the cortical
representation of the different languages in multilingual speakers may be complicated by
several factors. For example, Ojemann and Whitaker (1978) and Rapport, Tan and Whitaker
(1983) report the case of bilingual and polyglot epileptic patients who underwent language
assessment while receiving cortico-electrical stimulation before neurosurgery. Interestingly,
the authors detected centers common to all languages known by the patients along with
centers showing differential inhibition effects. Furthermore, the L2 tended to have a more
diffuse representation in the left hemisphere as opposed to the mother tongue. More recently,
several studies have shown that the cortical representation of languages in multilingual
speakers varies in function of two major factors: age of L2 acquisition and proficiency in the
second language (e.g. Perani et al., 1998; Perani et al., 1996; Dehaene et al., 1997, Kim et al.,
1997) (see below). Overall, results from neuroimaging investigations support the hypothesis
that bilinguals are not simply two monolinguals in one person suggesting that even the
multilingual brain is not the sum of two or more monolingual language systems “but is rather
a unique and complex neural system which may differ in individual cases” (Abutalebi, Cappa
and Perani, 2001, 188).
In what follows, a pragmatic definition of multilingualism will be privileged. Namely,
multilingualism (or multilingual competence) will be intended as the command and use of
two or more linguistic systems, whatever the level of proficiency and the age of acquisition of
those languages. Bilingualism (or bilingual competence) will be considered a type of
multilingualism, that is the command and use of two linguistic systems whatever the level of
proficiency and the age of acquisition of both languages.

FACTORS AFFECTING THE ACQUISITION OF LANGUAGES IN

MULTILINGUAL SPEAKERS
Several classifications have been proposed to distinguish among different types of
bilingualism (Butler and Hakuta, 2004; Edwards, 2004). When the focus is on the proficiency
in the two languages, a distinction has been introduced between balanced and dominant
6 Andrea Marini and Franco Fabbro

bilinguals (Peal and Lambert, 1962): a balanced bilingual masters the two languages to the
same extent, whereas a dominant bilingual is more fluent in one language than in the other.
As to when a second language has been acquired, it has been proposed to distinguish between
early and late bilingualism: the expression early bilingualism refers to the early acquisition
(in infancy) of both languages, whereas late bilingualism designates that the L2 has been
acquired much later than the mother tongue. As to the effect that a second language exerts on
the mother tongue, Lambert (1974) proposed to distinguish between additive and subtractive
bilingualism, where additive bilinguals are those who improve their L2 without loosing L1
proficiency, and subtractive bilinguals those who lower the proficiency in their L1 while
achieving a second language.
Some of these distinctions have proved empirically useful to distinguish among groups
of bilingual participants for psycholinguistic and neurolinguistic experiments (Hamers and
Blanc, 1990; Vaid, 1986). Indeed, several investigations have demonstrated that factors such
as context and modalities of L2 acquisition (e.g. acquiring a second language in a Country
where that language is currently spoken by means of its continuous use in real-life contexts,
rather than learning it at school by means of conscious and effortful assimilation of rules and
words) or the age when it has been learnt are particularly critical to determine good levels of
proficiency in that language (Neville, Mills and Lawson, 1992; Neville, Coffey, Lawson, et
al., 1997; Weber-Fox and Neville, 1996; Kim et al., 1997; for a review see Fabbro, 2001).
One of the challenges for second language learners is the difficulty (in most cases
impossibility) to acquire a native-like pronunciation of the L2. As to this respect, second
language learners are faced with three main problems: an articulatory problem, consisting in
the difficulty to acquire the correct articulatory configurations that allow them to produce the
target phones; a word-level prosody problem, that is the production of words with correct
intonation; a sentence-level prosody problem, consisting in the difficulty to learn how to
correctly produce the intonation pattern of sentences during a conversation. Results from a
number of investigations suggest that only those children who have been adequately exposed
to the second language before the age of 8 will acquire native-like control of articulatory and
prosodic aspects of speech production in L2 (Fabbro, 2004). For example, two groups of
children who migrated to the USA and to The Netherlands when they were younger than 8 are
reported to have acquired native-like phonetic/phonological competence in English and
Dutch, respectively (Asher and Garcia, 1969; Snow, 1987; Flege and Fletcher, 1992)
Interestingly, the accuracy in L2 pronunciation gets gradually lower for those people who
moved to other Countries when they were older than 8 and lower than 21 years of age and
remains low for those who have migrated after the age of 22. Therefore, it is possible to
postulate three critical periods for the acquisition of native-like pronunciation in a second
language: 1) Multilingual speakers acquiring two or more languages between 1 and 8 years of
age will have a native-like pronunciation in all those languages; 2) Multilingual speakers
learning a second, a third (or even more) languages between 9 and 21 years of age will
develop a relatively good pronunciation in all of them; 3) Multilingual speakers learning two
or more languages after the age of 22 will probably have a marked foreign accent. The
existence of a “critical period” for language learning has been reported also for what regards
grammatical development. For example, results from a series of experiments, where bilingual
participants were tested with L2 grammaticality judgment tasks, showed a significant
reduction in accuracy for those who had learnt the second language after the age of 8
(Johnson and Newport, 1989; Long, 1990). Further evidence for age-related effects on
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 7

bilingual acquisition comes from electrophysiological studies where the cortical organization
of the mother tongue and a second language was assessed using event-related potentials
(ERPs) in a group of early bilingual participants compared to a group of late bilingual
volunteers who had acquired the L2 after 7 years of age (Neville et al., 1992; 1997; Weber-
Fox and Neville, 1996). In the early bilingual group the authors found that for both languages
closed-class words were represented in the left frontal lobe, whereas open-class words usually
involved post-rolandic cortical structures. Conversely, in late bilinguals L2 closed-class
words were represented together with open-class words in post-rolandic areas and not in left
frontal areas. Moreover, in an fMRI study, Kim et al. (1997) compared twelve proficient
bilingual participants with diverse first and second languages in covert production tasks. The
volunteers were divided in two groups: an early acquisition group, formed by six participants
who had been exposed to both L1 and L2 during early infancy, and a late acquisition group,
formed by six subjects who had learnt the second language after puberty. In the early
bilingual group the authors found similar activation of Broca’s and Wernicke’s areas for both
languages. In the late bilingual group, L1 and L2 shared similar activations of Wernicke’s
area but different activations (separated by approximately 8 mms) within Broca’s area. These
results suggested that the age of acquisition is a major factor affecting the cortical
organization of a second language. It is noteworthy, however, that the authors did not specify
in detail the levels of L2 proficiency for the participants involved in the experiment. This is a
significant omission because the differences registered in Broca’s activation in the late
acquisition group may reflect differences at the morphosyntactic (functional and/or
positional) and phonetic (articulatory) levels of sentence processing (see also Abutalebi et al.,
2001; Fabbro, 2001). In a subsequent fMRI study, Chee, Tan and Thiel (1999) investigated
the cerebral activation in two groups of Chinese (Mandarin) – English highly proficient
bilinguals during word-stem completion tasks. The first group included fifteen participants
who learned English before the age of 6 (early bilingual group). The second group included
nine bilinguals who learned English after age 12 (late bilingual group). In this experiment, the
authors found no group-related difference in the activation of frontal areas (left prefrontal
region including the inferior frontal gyrus, left supplementary motor area and bilateral
occipital and parietal regions). The divergent results obtained by Kim et al. (1997) and Chee
et al. (1999) may depend not on the age of L2 acquisition (early vs. late acquisition), but on
the levels of proficiency in the two languages and on the time spent speaking them.
Therefore, the differential activation of left frontal areas reported by Kim et al. (1997) may
reflect different levels of L2 proficiency between the two groups rather than the age of L2
acquisition. A possible solution to this problem can be found in the results of a recent fMRI
experiment where three groups of Italian-German bilingual participants were compared in a
series of grammatical and semantic judgment tasks (Wartenburger et al., 2003). The first
group (Early Acquisition High Proficiency group [EAHP]) was formed by 11 bilingual
participants who learned the two languages together since birth and showed high proficiency
in both of them. The second group (Late Acquisition High Proficiency group [LAHP])
included 12 bilingual participants who learned German (L2) before 6 years of age but with
high proficiency in both languages. The third group (Late Acquisition Low Proficiency group
[LALP]) consisted of 9 bilingual volunteers who learned German (L2) after the age of 6 and
had were low proficient in that language. The EAHP and LAHP groups had a perfect phonetic
and morphosyntactic competence in the L2, whereas the participants included in the LALP
group made a number of morphosyntactic errors coupled with a strong foreign accent. On the
8 Andrea Marini and Franco Fabbro

grammatical judgement task, the LAHP group showed greater activation in language-related
areas than the EAHP group. At the same time, with respect to the LALP group, greater
activation was found in the left temporo-parietal junction, the right lingual gyrus and right
inferior parietal lobule in the LAHP group. These results suggest that age of acquisition and,
to a lesser degree, proficiency level affect the cortical representation of grammatical
processes in L2.

MODELS OF LANGUAGE REPRESENTATION IN BILINGUAL AND

MULTILINGUAL INDIVIDUALS
The first systematic investigations of bilingualism began approximately 100 years ago as
diary studies by researchers reporting on the linguistic development of their own children.
Ronjat (1913) described the linguistic development of his son in a bilingual German-French
environment. Leopold (1939; 1949) provided a detailed report of his daughter Hildegard’s
simultaneous acquisition of English and German. These studies represent a rich source of
observations. However, the analyses of bilingual language development performed by the
authors lacked of a systematic linguistic approach and were based on subjective impressions
rather than rigorous scientific methodology.
One of the first attempts to formulate a model of lexical representation in bilinguals is
the “compound-coordinate model” of bilingualism initially proposed by Weinreich (1953)
and then revised by Ervin and Osgood (1954). In its original configuration, this model
assumes that in bilingual individuals the phonological (“signifier”) and semantic (“signified”)
aspects of the words in the two languages are represented in three possible configurations:
coordinate, compound and subordinate, respectively. In a coordinate system the phonological
representation of translation equivalents (i.e. words of different languages sharing the same
meaning) matches two different meaning representations. In such a system, the bilingual
individual would have two separate conceptual representations and two separate word forms
for the Italian word “casa” and the corresponding English word “house” (figure 1). In a
compound representation, the two phonological forms are identified with a common meaning
representation. Therefore, the two word forms “casa” and “house” are represented in the same
conceptual system (figure 2). A subordinate system is a word association model where the
meaning representation of an L2 word (e.g. “house”) is accessed through the meaning
representation of the word of the mother tongue (e.g. “casa”) (figure 3).
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 9

Figure 1. Representation of a coordinate system (adapted from Woutersen, Cox, Weltens, and De Bot,
1994).

Figure 2. Representation of a compound system (adapted from Woutersen, Cox, Weltens, and De Bot,
1994).
10 Andrea Marini and Franco Fabbro

Figure 3. Representation of a subordinate system (adapted from Woutersen, Cox, Weltens, and De Bot,
1994).

Erving and Osgood (1954) suggested that coordinate or compound representation of two
languages depends on the modalities of L2 acquisition. An individual learning two languages
in separate cultural environments will develop a coordinate representation. This is for
example the case of an English child who learns English as his mother tongue in an English-
speaking Country and then Italian as L2 in an Italian-speaking environment. In contrast, a
child learning two languages together in the same context will develop a compound structure.
An individual learning a second language after a certain age in formal contexts will develop a
subordinate representation. While speaking, subordinate bilinguals formulate what they want
to say in their L1 and then translate it into L2. Therefore, in a subordinate representation the
meanings of the L2 words are accessed only through a translation process.
The coordinate-compound model introduced important theoretical issues on the debate
about language representation in bilingual individuals such as the hypothesis of a distinction
between conceptual and lexical representation of words, and the problem of language storage
in bilingual speakers.
As to the first problem, results from psycholinguistic and neurolinguistic studies support
the hypothesis that the two languages of a bilingual individual access a common semantic
system (e.g. Illes et al., 1999; Costa, Miozzo and Caramazza, 1999; Hermans, Bongaerts, de
Bot and Schreuder, 1998; Potter, So, von Eckardt, and Feldman, 1984). For example, in an
fMRI study aimed at investigating the brain activation patterns during semantic judgement
tasks (concrete/abstract judgements) in a group of 8 late Spanish-English bilingual speakers
who had learned their L2 after the age of 10, Illes et al. (1999) identified a shared frontal lobe
system for semantic processing for both languages (left inferior frontal gyrus; in six of the
participants the activation included also the right inferior frontal gyrus).
As to the problem of whether lexical knowledge of bilinguals’ two languages is merged
into one large lexical system or is kept separate in two distinct lexicons (i.e. a lexicon for L1
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 11

and a lexicon for L2), there is much less agreement, as two opposite hypotheses have been put
forward and are currently debated (Grosjean, 1995): a two-lexicons view opposed to a one-
lexicon view. Proponents of the two-lexicons hypothesis (Potter et al., 1984; De Groot, 1993;
Kroll and Stewart, 1994) claim that the lexical representations in bilingual speakers are stored
in two different lexicons and that information acquired in one language is available in the
other only through a translation process. In contrast, according to the one-lexicon view (e.g.
De Bot and Schreuder, 1993; De Bot, 1992; Paradis, 1987), the two languages of a bilingual
speaker are stored in different subsystems of the same lexicon which can be activated to
different extents, depending on which language is currently being used.
Within the theoretical framework of the two-lexicons view, the coordinated-compound
model proposed by Weinreich has been extended and reformulated. Such models are usually
defined “hierarchical” as they postulate that a word’s conceptual representation (its
“meaning” in the broadest sense) is separated from its lexical representation and that the
words of the two languages of a bilingual individual are stored in two separate lexicons, one
for the words of the L1, the other for those of the L2. Potter et al. (1984) proposed a Word
Association Model and a Concept Mediation Model based on Weinreich’s subordinate system
and compound system, respectively. According to the Word Association Model, L2 words do
not have a direct access to the corresponding concepts but must first be translated in the
corresponding L1 lexical representations that, in turn, have direct access to the conceptual
system. This model is assumed to represent the organization of the two languages in non-
fluent low proficient bilinguals such as for example second language learners in the early
stages of L2 acquisition. In the Concept Mediation Model the lexical items of both lexicons
(L1 and L2) are supposed to have direct access to the conceptual system by means of direct
conceptual-lexical links. This system is hypothesized to reflect the competence of bilinguals
who have reached high levels of proficiency in their L2. Taken together, the Word
Association Model and the Concept Association Model entail a developmental hypothesis of
language representation in bilinguals, as the shift from lexical to conceptual mapping is
assumed to reflect growing levels of proficiency in the L2. Such hypothesis has been tested in
a number of studies (see Kroll and de Groot [1997] for an extensive review on this issue). For
example, Chen and Leung (1989) and Kroll and Curley, (1988) compared the reaction times
(RTs) of more and less fluent bilinguals in a picture naming task and a word translation task.
Under the assumption that picture naming is achieved through concept mediation, it was
hypothesized that similar RTs in both tasks would imply the recourse to a concept mediation
strategy, whereas shorter RTs at the word translation task would imply the recourse to a
lexical translation strategy. Indeed, less proficient bilingual individuals at earlier stages of L2
acquisition were faster in the word translation task than in the picture naming task (use of the
lexical translation strategy), whereas in the more proficient bilinguals the RTs were
comparable across the two tasks (recourse to the concept mediation strategy). Therefore,
experimental results seem to support the proficiency-related shift from a Word Association to
a Concept Association representation of the bilinguals’ two languages. However, neither of
the two models could account for the translation asymmetry effect (Kroll and Dussas, 2004;
Kroll and Sholl, 1992) reported in both early and advanced bilinguals in a number of
empirical observations, that is the fact that bilinguals usually translate faster and more
accurately in the L2 to L1 rather than in the L1 to L2 direction. In order to explain the
translation asymmetry effect, Kroll and Stewart (1994) proposed a Revised Hierarchical
12 Andrea Marini and Franco Fabbro

Model where both the Word Association and the Concept Association Model were
incorporated (figure 4).

Figure 4. The Revised Hierarchical Model (adapted from Kroll and Stewart, 1994).

The Revised Hierarchical Model maintains the idea of independent lexical

representations for the two languages known by the bilingual and of a shared conceptual
system for both languages. However, it also introduces some important innovations to the
previously presented models. We will briefly consider some of the main assumptions of the
model. It is assumed that the extension of the lexicon of the L1 is larger than that of the L2, as
even highly proficient bilinguals usually know more words in the native language than in
their subsequently learned languages. In order to account for the translation asymmetry effect,
the connections between the two languages (L1 and L2) and the concept-system are supposed
to be asymmetrical and to reflect the modalities of L2 acquisition. It is postulated that
translating from L2 to L1 entails a direct access to the L1 lexical form, whereas translating
from L1 to L2 is supposed to activate not only the corresponding translation equivalent in L2,
but also the conceptual representation of the L1 word thus engaging a semantic route as well.
The forward translation direction (from L1 to L2) is assumed to be particularly challenging
for less proficient bilinguals, because in these individuals the link between the second
language and the conceptual store is still weak. Indeed, experiments on translation generally
support these predictions (e.g. Kroll, Michael, Tokovicz, and Dufour, 2002).
Even if appealing, the two-lexicons view has been challenged by a number of
neurophysiologic and neuroimaging studies that, at least at the macroscopical level,
evidenced a similar cerebral representation for L1 and L2 lexicons in both early and late
bilinguals (Klein et al., 1999; Chee et al., 1999; Illes et al., 1999; Hernandez et al., 2000). For
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 13

example, in a Positron Emission Tomography (PET) study Klein et al. (1999) found that
Chinese (L1) – English (L2) bilinguals showed similar cerebral activation patterns for both
languages when involved in verb generation tasks (left inferior frontal, dorso-lateral frontal
cortex, temporo-parietal cortices, and right cerebellum). Furthermore, results from the afore
mentioned study by Chee, Than and Thiel (1999) provide additional evidence in favour of the
hypothesis that the cortical representation of words even in typologically different languages
such as English and Mandarin may involve the same cortical areas. Overall, as it will be
shown in the next paragraph, results from both experimental and clinical studies suggest that
age and modalities of acquisition, as well as language use probably affect the
neuroanatomical and functional organization of the lexicon.

THE FUNCTIONAL ARCHITECTURE OF THE BILINGUAL SPEECH

PRODUCTION SYSTEM
Several hypotheses have been put forward to account for the functional architecture of
the bilingual speech production system. In most cases these are the reformulation of models
originally proposed for monolingual speakers. Indeed, psycholinguistic theorizing has usually
held monolingualism as the canonical form of language use (Vaid, 2002), implicitly assuming
that monolingual acquisition and processing are the norm. Such monolingual prejudice has
proved to be misguiding in cognitive research on multilingualism and has conditioned the
modelling of bilingual and multilingual competence. In fact, quite no one can be considered
purely monolingual as even monolingual children usually grow up in a social environment
where more than one variety of the same language is spoken (e.g. the so-called “dialects”).
Therefore, they learn to use a certain variety of the language they are exposed to in specific
linguistic and/or extra-linguistic contexts, and other varieties in different contexts. In other
words, even those children who are generally regarded as monolingual will develop a
multilingual competence, based on lexical, grammatical and pragmatic/discursive knowledge
relative to each of the varieties of the language to which they are exposed. Let’s consider the
case of a Canadian born in Quebec. He may acquire Quebecois as his mother tongue (L1) and
use it in familiar contexts, standard French as a second language (L2) used as language of
education and in formal contexts, and eventually English as a third language (L3), used not in
daily/life interactions but, for example, to write scientific articles or to give lectures at
international congresses.
The issue of language production in polyglots raises several problems. A major question
concerns the functional architecture of the multilingual lexicon, that is how words are
represented, selected and accessed in multilingual speakers. One of the most influential
models of language production is the “bilingual production model” (de Bot, 1992; de Bot and
Schreuder, 1993), based on Levelt’s “speaking model” (1989). In what follows we will
attempt to integrate the model proposed by de Bot (1992) with data from recent
psycholinguistic, neurophysiologic and neuroimaging investigations.
De Bot’s bilingual production model hypothesizes the existence of three subsystems for
language production (a conceptualizer, a formulator and an articulator, respectively), a
subsystem for comprehension (speech-comprehension system), and a lexicon involved in both
production and comprehension.
14 Andrea Marini and Franco Fabbro

In multilingual speakers, the conceptualizer is assumed to be language independent. It

elaborates the conversational conventions and the contextual information in order to signal
“which language to choose”. It is at the level of the formulator that the language to be used in
the interaction will be chosen. The formulator is considered a procedural system working on a
system of declarative knowledge: the lexicon. Adopting Paradis’ (1987) subset hypothesis,
De Bot (1992) proposed that L1 and L2 words form different subsets within one and the same
lexicon. Experimental and clinical studies suggest that age and modalities of acquisition, as
well as language use probably affect the organization of both formulator and lexicon. In
multilingual individuals who have acquired two or more languages from birth, the formulator
is probably represented in common cortico-subcortical cerebral structures. In this case the
different languages would be maintained separate merely by neurofunctional mechanisms.
When a second language is learned after the age of 8, however, the neurofunctional systems
accounting for grammar and phonology are most likely to be separate also at the
neuroanatomical level (Fabbro, 1996; Kim, Relkin, Lee and Hirsch, 1997). Furthermore, in
fluent bilinguals the lexicon is probably represented in common neural structures (parieto-
temporal areas) and the neurofunctional separation between languages might depend on word
use relationships.
In the formulator the preverbal message is converted into a speech plan by means of two
processes: lexical selection and lexical access, respectively. When considering multilingual
competence, it becomes important to explain how a multilingual speaker selects the correct
lexical node in the target language preventing cross-language interference. As to this respect,
two major hypotheses have been proposed. The first states that the intended meaning in the
preverbal message selectively activates the lexical items only of the selected language and not
those of the other language(s) (Soares and Grosjean, 1984; Macnamara and Kushnir, 1971).
Such view has been generally abandoned in favour of an alternative hypothesis claiming that
the intended meaning activates the subsets of words belonging to the languages mastered by
the speaker and that the target word is selected by creating an imbalance in their activation
levels (De Bot, 1992; Green, 1986, 1998). It is postulated that each subset has its own specific
activation thresholds1 (Luria, 1973; Green, 1986, 1998; Paradis, 1989; 1993) and can be
selected by means of activation and inhibition processes2 (Inhibitory Control Model, Green,
1998). Inhibition is generally automatic and avoids interferences among the languages that
form the multilingual competence. The selection of the target lexical node is obtained by
creating a differential level of activation in the two or more lexicons of the multilingual
speaker by means of both activation of the target words and reactive inhibition of its
competitors. In other terms, the selection of a word in one language automatically triggers an
inhibition process of the competitors in the other language(s). Consider the case of a trilingual
individual who knows Italian (L1), English (L2) and Dutch (L3). In this case, the activation of
the word ‘bike’ (L2) will inhibit the corresponding lexical items in L1 (i.e. Italian
“bicicletta”) and L3 (i.e. Dutch “fiets” ) as well as semantically and phonologically similar

1
A cerebral lesion can lower the activation threshold of a language, which is thus not lost, but simply inaccessible
through the usual activation threshold. Generally, such activation threshold is lower for comprehension than
for expression. For this reason, there are cases of bilingual aphasic patients with preserved comprehension of
a language associated to impaired production in it (Paradis, 1996).
2
It is likely that inhibitory relations among languages are organized differently in simultaneous interpreters in
which both languages are concurrently activated for many hours a day. In this particular profession, the
activation thresholds of the two languages are thus quantitatively different as opposed to those of the other
polyglots.
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 15

words in the three languages. The multilingual speaker’s languages can receive three states of
activation (Green, 1986). They can be selected, active or dormant depending on the
communicative situation. A language is selected when it has been chosen for the interaction.
The selected language controls the speech output. A language is active when it has not been
chosen as the main code for the communicative exchange but is kept active during the whole
interaction. It works in parallel to the selected language but has no direct access to the
outgoing speech processing. A dormant language does not receive any activation and
consequently does not play a role in ongoing processing. In the multilingual competence, one
language is always selected while the other(s) language(s) can be either active or dormant. In
this framework it is possible to explain the code-switching phenomenon. Language switching
occurs quite commonly for instance in bilingual communities where both languages have the
same social status (see Grosjean, 1998). The habit of bilingual individuals to alternate
between languages within one coherent discourse may determine low thresholds of activation
of switching phenomena and a reduction in both languages’ mutual inhibition. In such
bilingual communities, it is quite normal for individuals to switch from one language to the
other while speaking informally. In other communities, however, such behavior is
intentionally avoided for sociolinguistic reasons. For example, in Brussels, where both French
and Flemish are official languages, a public official or a bank employee would never switch
between the two languages and would stick to his/her interlocutor’s language, even though
most natives understand and speak both languages. In Belgium each linguistic community has
a strong identity which is fiercely defended also through language use. This switching
mechanism is not peculiar only to polyglot speakers, but also to monolingual individuals who
use it to select among different linguistic registers according to the communicative context.
For example, an individual may ask his/her interlocutor to close the door by using a courtesy
form (“Will you please close the door?”) or a direct form (“Close the door”). In Paradis’
(1993) opinion, the system accounting for the selection of one of the two registers is similar
to that accounting for the selection of one language rather than another. Many neurologists
and neurolinguists have discussed the possible neurological organization of the switch
mechanisms. Some of them, generally defined as localizationists, claim that the activation of
a specific language (for example L1) and the concurrent inhibition of the other languages (e.g.
L2, L3, and the like) is subserved by specific brain areas. For example, Leischner (1943)
claimed that this center is localized in the left supramarginal gyrus, whereas, according to
Stengel and Zelmanovicz (1933) and Zatorre (1989), it is localized in anterior structures of
the left hemisphere. In Lebrun’s opinion (1971), the switching mechanism is localized in the
right hemisphere. Other authors have criticized the idea of an anatomical center governing
language switching. Goldstein (1948), for example, suggested that any cerebral lesion impairs
switching between mental processes, and that the faculty of switching is only an example of
the more general faculty of abstraction. Paradis (1993) claims that switching mechanisms are
only one aspect of a more general system involved in decision-taking processes. In Paradis’
opinion, the switch function is part of a general system responsible for the selection of
behaviors such as standing up or sitting down, speaking Italian or English, etc. Recent
evidence suggests that the language switching mechanism involves the dorsolateral prefrontal
cortex (Fabbro, Skrap, and Aglioti, 2000; Hernandez, Dapretto, Mazziotta, and Bookheimer,
2001; Holtzeimer, Fawatz, Wilson, and Avery, 2005). Indeed, Fabbro et al. (2000) report the
case of a bilingual patient with a lesion to the left anterior cingulated and to the frontal lobe
(marginally involving also the right anterior cingulate area). After that lesion, the patient
16 Andrea Marini and Franco Fabbro

presented pathological switching between languages in the absence of any other linguistic
impairment.
Once the target lexical node has been selected, the formulator gains access to its
semantic, morphosyntactic and morphological features (lemma level) and then to its
phonological form (lexeme level). In case of single word production the lexical information is
directly transmitted to the articulator. In case of sentence production, however, the
information stored in the lemma level of word representation is used to generate the
argumental structure and hence the syntactic organization of the sentence. At this point, the
articulator converts the speech plan into actual speech. During speech production in bilinguals
both selected and active languages are simultaneously activated at all levels with the
exception of the articulatory subsystem, which remains inactive for the non-selected language
(de Bot, 1992). The articulator transforms the strings of syllables of the selected language in
articulatory patterns. It is assumed that syllable articulatory programs are automatized and
that the level of automatism correlates with the level of proficiency which, in turn, is a
function of age of L2 acquisition. In other words, a multilingual speaker can pronounce as a
native speaker words and sentences in two or more languages only if he/she has correctly
automatized the syllable articulatory programs before the age of 8. Advanced bilinguals,
usually early bilinguals, may have only one large articulatory system containing all syllable
articulatory programs for all the languages they master, whereas less proficient bilinguals,
usually late bilinguals, may have independent stores for each language (e.g. Flege and
Fletcher, 1992).

REFERENCES
Abutalebi J, Cappa SF and Perani D (2001). The bilingual brain as revealed by functional
neuroimaging. Bilingualism Lang. Cogn., 4: 179-190
Albert, M.L., and Obler, L.K. (1978). The bilingual brain. New York: Academic Press.
Asher J, and Garcia R (1969). The optimal age to learn a foreign language. Modern Language
Journal, 28: 334-341
Bloomfield L (1933). Language. New York: Holt, Rinehart and Winston
Butler YG and Hakuta K (2004). Bilingualism and Second Language Acquisition. In T K
Bhatia and W C Ritchie (eds.) The Handbook of Bilingualism. (pp. 114-144). Blakwell
Publishing. Malden , MA
Caplan D (1992). Language. Structure, processing and disorders. Cambridge, Massachusetts:
MIT Press
Chee, M. W. L., Tan, E. W. L., and Thiel, T. (1999). Mandarian and English single word
processing studied with functional magnetic resonance imaging. The Journal of
Neuroscience, 19, 3050–3056.
Chen H-C. and Leung Y.S. (1989). Patterns of lexical processing in a nonnative language. J.
Exp. Psychol. Learn. Mem. Cogn., 15, 316-325.
Chomsky N (1988). Language and problems of knowledge: the Managua Lectures, MIT
Press, Cambridge, Massachusetts
Chomsky, N. (1995). The minimalist program. Cambridge, MA: MIT Press.
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 17

Costa, A., Miozzo, M., and Caramazza, A. (1999). Lexical selection in bilinguals: Do words
in the bilingual’s two lexicons compete for selection? Journal of Memory and Language,
41, 365–397.
Costa A, and Caramazza A (1999). Is lexical selection in bilingual speech production
language-speciWc? Further evidence from Spanish–English and English–Spanish
bilinguals. Bilingualism: Language and Cognition, 2, 231–244.
Costa A, and Santesteban M (2004a). Bilingual word perception and production: Two sides of
the same coin? Trends in Cognitive Sciences, 8(6), 253.
Costa A, and Santesteban M (2004b). Lexical access in bilingual speech production: Evidence
from language switching in highly proWcient bilinguals and L2 learners. Journal of
Memory and Language, 50, 491–511.
de Bot K (1992). A bilingual production model: Levelt’s speaking model adapted. Applied
Linguistics, 13, 1–24.
de Bot, K., and Schreuder, R. (1993). Word production and the bilingual lexicon. In R.
Schreuder and B. Weltens (eds.), The bilingual lexicon (pp. 191-214). Amsterdam: John
Benjamins.
de Groot A.M.B. (1993). Word-type effects in bilingual processing tasks: Support for a
mixed-representational system. In R. Schreuder and B. Weltens (eds.) The bilingual
lexicon (pp. 27-51). Amsterdam: John Benjamins
Dehaene, S., Dupoux, E., Mehler, J., Cohen, L., Paulesu, E., Perani, D., van de Moortele,
P.F., Lehéricy, S., and Le Bihan, D. (1997). Anatomical variability in the cortical
representation of first and second language. Neuroreport, 8(17):3809-15
Edwards, J.V. (2004). Foundations of bilingualism. In T.K.Bhatia and W.C. Ritchie (eds.),
The handbook of bilingualism (pp. 7-31). Blakwell Publishing. Malden , MA
Erwin, S., and Osgood, C. (1954). Second language learning and bilingualism. Journal of
abnormal and social psychology. Supplement. 49, 139-146
Fabbro, F. (1996) . Neurofunctional aspects of bilingual aphasia. Journal of the Israeli Speech
Hearing and Language Association, 19: 13-15.
Fabbro F (1999). The neurolinguistics of bilingualism: An introduction. Hove, UK:
Psychology Press.
Fabbro F (2001). The bilingual brain, cerebral representation of languages. Brain and
Language, 79, 211–222.
Fabbro F (2004). Neuropedagogia delle lingue. Astrolabio, Roma, Italy
Fabbro F, Skrap M, and Aglioti S (2000). Pathological switching between languages after
frontal lesions in a bilingual patient. Journal of Neurology, Neurosurgery, and
Psychiatry, 68, 650–652.
Flege, J.E. and Fletcher, K.L. (1992). Talker and listener effects on degree of
perceived foreing accent. Journal of the Acoustical Society of America, 91:
370-389.
Frazier L, and Fodor JD (1978). The sausage machine: a new two-stage parsing model.
Cognition, 6, 291–325.
Goldstein, K. (1948). Disturbances of language in polyglot individuals with aphasia. In
Language and language disturbances (pp. 138-146). New York: Grune and Stratton.
Green DW (1986). Control, activation and resource. Brain and Language, 27, 210–223.
18 Andrea Marini and Franco Fabbro

Green DW (1998). Mental control of the bilingual lexico-semantic system. Bilingualism:

Language and Cognition, 1, 67–81.
Grice HP (1975). Logic and conversation. In Cole P, Morgan JL (eds.) Syntax and semantics:
Vol. 3: Speech acts, Academic Press, New York , 41-58
Grosjean F (1982). Life with Two Languages. An Introduction to Bilingualism. Cambridge:
Harvard University Press
Grosjean F (1989). Neurolinguists beware! The bilingual is not two monolinguals in one
person. Brain and Language, 36, 3-15
Grosjean F (1994). Individual bilingualism. In: Asher R E (ed.) The Encyclopaedia of
Language and Linguistics. Pergamon Press, Oxford: 1656-60
Grosjean F (1995). A psycholinguistic approach to code-switching: The recognition of guest
words by bilinguals. In L. Milry and P. Muysken (eds.), One speaker, two languages
(259-275). Cambridge: Cambridge University Press.
Grosjean, F. (1998). Studying bilinguals: Methodological and conceptual issues.
Bilingualism: Language and Cognition, 1: 131-149.
Grosjean F (1999) Individual bilingualism. In B. Spolsky (ed.) Concise Encyclopaedia of
Educational Linguistics, pp. 284-90. London: Elsevier
Hakuta K (1986) Mirror of language: The debate on bilingualism. New York: Basic Books
Hamers JF, and Blanc MHA (1990). Bilinguality and Bilingualism. Cambridge: Cambridge
University Press.
Haugen, E. (1953). Bilingualism in the Americas. Montgomery, AL: University of Alabama
Press.
Hermans, D., Bongaerts, T., de Bot, K., and Schreuder, R. (1998). Producing words in a
foreign language: Can speakers prevent interference from their first language?
Bilingualism: Language and Cognition, 1: 213-229.
Hernandez A.E., Martinez, A., and Kohnert, K. (2000). In search of the language switch: An
fMRI study of picture naming in Spanish-English bilinguals. Brain and Language, 73:
421-431.
Hernandez, A.E, Dapretto, M., Mazziotta, J., and Bookheimer, S. (2001). Language switching
and language representation in Spanish-English bilinguals: An fMRI study. Neuroimage,
14: 510-520.
Holtzeimer, P., Fawatz, W., Wilson, C., and Avery, D. (2005). Repetetive transcranial
magnetic stimulation may induce language switching in bilingual patients. Brain and
Language, 94, 274-277.
Illes, J., Francis, W., Desmond, J., Gabrieli, J., Glover, G., et al. (1999). Convergent cortical
representation of semantic processing in bilinguals. Brain and Language, 70: 347-363.
Johnson, J.S.and Newport, E.L. (1989). Critical period effects in second language
learning: The influence of maturational state on the acquisition of English as a
second language. Cognitive Psychology, 21: 60-99.
Johnson-Laird PN (1980). Mental models in cognitive science. Cognitive Science, 4, 71-115.
Kim, K.H.S., Relkin, N.R., Lee, K.-M. and Hirsch J. (1997). Distinct cortical
areas associated with native and second languages: Nature, 388: 171-174.
Kintsch W, and Van Dijk T (1978). Toward a model of text comprehension and production.
Psychological Review. 85, 363-394
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 19

Klein, D., Milner, B., Zatorre, R., Zhao, V., and Nikelski, J. (1999). Cerebral organization in
bilinguals: A PET study of Chinese-English verb generation. NeuroReport, 10: 2841-
2846.
Kroll J.F. and Curley J. (1988). Lexical memory in novice bilinguals: the role of concepts in
retrieving second language words. In P.M. Gruneberg and R. Sykes (eds.) Practical
Aspects of Memory (Vol. 2): John Wiley and Sons. London, UK
Kroll J.F. and Dussas P.E. (2004). The comprehension of words and sentences in two
languages. In T.K. Bhatia and W.C. Ritchie (eds.) The handbook of bilingualism (pp.
169-200). Blackwell Publishing. Malden, MA
Kroll, J.F., and de Groot, A.M.B. (1997). Lexical and conceptual memory in the bilingual:
Mapping form to meaning in two languages. In A. de Groot and J. Kroll (eds.), Tutorials
in bilingualism: Psycholinguistic perspectives (pp. 169-199). Mahwah, NJ: Lawrence
Erlbaum Associates.
Kroll J.F., Michael E., Tokovicz N. and Dufour R. (2002). The development of lexical
fluency in a second language. Second Language Research, 18, 137-171
Kroll J.F. and Sholl A. (1992). Lexical and conceptual memory in fluent and nonfluent
bilinguals. In R.J. Harris (ed.) Cognitive processes in bilinguals (pp. 191-206).
Amsterdam: Elsevier Science.
Kroll JF, and Stewart E (1994). Category interference in translation and picture naming:
Evidence for asymmetric connections between bilingual memory representations. Journal
of Memory and Language, 33, 149–174.
Lambert WE (1974). Culture and language as factors in learning and education. In F F Aboud
and R D Meade (eds.) Cultural Factors inLearning and Education. Bellingham, W A:
Western Washinghton State University.
Lebrun, Y. (1971) The neurology of bilingualism. Word, 27 (1-3), 179-186
Leischner, A. (1943). Aphasia of deaf-mutes. In Paradis, M. (Ed.), Readings on Aphasia in
Bilinguals and Polyglots. Montreal: Didier, 1983, pp.423-444.
Leopold W. (1939-1949). Speech development of a Bilingual Child (4 vols.) Evanston, IL:
North-western University Press
Levelt WJM (1989). Speaking: From intention to articulation. Cambridge, MA: MIT Press.
Levelt WJM, Roelofs A, and Meyer AS (1999). A theory of lexical access in speech
production. Behavioral and Brain Sciences, 22, 1–37.
Levinson S (1983). Pragmatics. Cambridge University Press, Cambridge
Long MH (1990). Maturational constraints on language development. Studies in second
language acquisition, 12: 251.285
Luria, A.R. (1973). The working brain: An introduction to neuropsychology. New York:
Penguin Education.
Macnamara, J., and Kushnir, S.L. (1971). Linguistic independence of bilinguals: The input
switch. Journal of Verbal Learning and Verbal Behavior, 10: 480-487.
Marini, A. (2001). Elementi di psicolinguistica generale, Springer Verlag, Milano, Italy
Marini A., Carlomagno S., Caltagirone C., and Nocentini U. (2005a). The role played by the
right hemisphere in the organization of complex textual structures. Brain and Language.
93/1, 46-54
Marini A., Boewe A. and Carlomagno S. (2005b). Assessment of age-related differences in
the processing of textual descriptions. Journal of Psycholinguistic Research. 34/5, 439-
463
20 Andrea Marini and Franco Fabbro

Marini, A., and Nocentini, U. (2003). Comunicazione verbale e emisfero destro. Springer
Verlag, Milano, Italy
Martin R C, Newsome M R and Vu H (2002). Language and lexical processing. In V S
Ramachandran (Ed.) Encyclopaedia of the Human Brain. Vol. 2 (pp. 631-643). Academic
Press. San Diego.
Meisel J M (2004). The Bilingual Child,. In T K Bhatia and W C Ritchie (eds.) Handbook of
Bilingualism. pp. 91-113. Blakwell Publishing. Malden , MA
Neville, H.J., Mills, D.L. and Lawson, D.S. (1992). Fractionating language: different neural
subsystems with different sensitive periods. Cerebral Cortex, 2: 244-258.
Neville, H.J., Coffey, S.A., Lawson, D.S., Fischer, A., Emmorey, K. and Bellugi,
U. (1997). Neural systems mediating American Sign Language: Effects of
sensory experiences and age of acquisition. Brain and Language, 57: 285-308.
Ojemann, G.A. (1991). Cortical organization of language. The Journal of Neuroscience, 11:
2281-2287
Ojemann, G.A., and Whitaker, H.A. (1978). The bilingual brain. Archives of Neurology, 35:
409-412.
Paradis, M. (1978). Bilingual linguistic memory: Neurolinguistic considerations. Paper
presented at the Annual Meeting of the Linguistic Society of America, Boston, 28
December.
Paradis, M. (1987). Neurolinguistic perspectives on bilingualism. In M. Paradis and G.
Libben, The assessment of bilingual aphasia (pp.1-17). Hillsdale, NJ: Lawrence Erlbaum.
Paradis, M. (1989). Bilingual and polyglot aphasia. In F. Boller and J. Grafman (eds.),
Handbook of Neuropsychology. Vol. 2, 117-140. Amsterdam: Elsevier.
Paradis, M. (1993). Linguistic, psycholinguistic, and neurolinguistic aspects of ‘interference’
in bilingual speakers: The activation threshold hypothesis. International Journal of
Psycholinguistics, 9: 133-145.
Paradis, M. (1994). Neurolinguistic aspects of implicit and explicit memory: implications for
bilingualism and second language acquisition. In N. Ellis (Ed.), Implicit and Explicit
Language Learning. London: Academic Press, 393-419
Paradis, M. (1995). The need for distinctions. In M. Paradis (ed.), Aspects of bilingual
aphasia (pp. 1-9). Oxford: Pergamon Press.
Paradis, M. (1996). The cognitive neuropsychology of bilingualism. In: A. De Groot and J.
Kroll (Eds.), Tutorials in Bilingualism: Psycholinguistic Perspectives.. Hillsdale, NJ:
Erlbaum.
Paradis M (2004). A neurolinguistic theory of bilingualism. Amsterdam: John Benjamins.
Peal E and Lambert WE (1962). The relation of bilingualism to intelligence. Psychological
Monographs, 76, 1-23
Perani, D., Paulesu, E., Sebastian-Gallés, N., Dupoux, E., Dehaene, D., Bettinardi, V., Cappa,
S., Fazio, F., and Mehler, J. (1998). The bilingual brain: Proficiency and age of
acquisition of the second language. Brain, 121: 1841-1852.
Perani, D., Dehaene, S., Grassi, F., Cohen, L., Cappa, S., Paulesu, E., Dupoux, E., Fazio, F.,
and Mehler, J. (1996). Brain processing of native and foreign languages. NeuroReport, 7:
2439-2444.
Pinker S (1999). Words and rules: the ingredients of language. New York: Basic Books.
 Psycholinguistic Models of Speech Production in Bilingualism and Multilingualism 21

Potter M.C., So K., Eckardt V. and Feldman L. (1984). Lexical and conceptual representation
in beginning and proficient bilinguals. Journal of Verbal Learning and Verbal Behavior,
23, 23-38
Rapport, R.L., Tan, C.T., and Whitaker, H.A. (1983). Language function and dysfunction
among Chinese- and English-speaking polyglots: Cortical stimulation, Wada testing, and
clinical studies. Brain and Language, 18: 342-366.
Ronjat J. (1913). Le développement du langage observé chez un enfant bilingue. Paris :
Champion.
Snow C (1987). Relevance of the notion of a critical period to language acquisition. In
Bornstein MH (ed.) Sensitive Periods in Development: Interdisciplinary Perspectives,
Erlbaum, Hillsdale, NJ
Soares, C. and Grosjean, F. (1984). Bilinguals in a monolingual and a bilingual speech mode:
The effect of lexical access. Memory and Cognition, 12: 380-386.
Stengel, E. and Zelmanowicz, J. (1933). Über polyglotte motorische Aphasie. Zeitschrift für
die gesamte Neurologie und Psychiatrie, 149: 292-311.
Tucker GR (1998). A global perspective on multilingualism and multilingual education. In J
Cenoz and F Genesee (eds.) Beyond bilingualism: Multilingualism and multilingual
education. Clevedon, England: Multilingual Matters. 3–15
Ullman MT (2001). The declarative/procedural model of lexicon and grammar. Journal of
Psycholinguistic Research, 30(1), 37–69.
Ullman MT (2004). Contributions of memory circuits to language: the declarative/procedural
model. Cognition, 92, 231-270
Vaid J (Ed.) (1986). Language Processing in Bilinguals: Psycholinguistic and
Neuropsychological Perspectives. Hillsdale, NJ.: Erlbaum.
Vaid J (2002). Bilingualism. In V S Ramachandran (ed.) Encyclopaedia of the Human Brain.
Academic Press, Elsevier. Vol. 1, 417-434
Wartenburger I, Heekeren HR, Abutalebi J, Cappa SF, Villringer A, and Perani D (2003)
Eatly settings of grammatical processing in the bilingual brain. Neuron, 37:159-170
Weber-Fox, C.M. and Neville, H.J. (1996). Maturational constraints on functional
specializations for language processing: ERP and behavioral evidence in bilingual
speakers. Journal of Cognitive Neuroscience, 8: 231-256.
Weinreich U (1953). Languages in contact. New York: The Linguistic Circle of New York
Woutersen M., Cox A., Weltens B. and De Bot K. (1994). Lexical aspects of standard dialect
bilingualism. Applied Psycholingusitics, 15, 447-473.
Zatorre, R.J. (1989). On the representation of multliple languages in the brain: Old problems
and new directions. Brain and Language, 36: 127-147.

View publication stats