Bilingual language processing

Part 1: Basics
 Psycholinguistic research on Language
processing
 Psycholinguistics concerns itself with language processing:

 How is the computing done in word, sentence and discourse

meaning

 This is a mostly unconscious process

 How does the mental representation of these processes look like?

 Basically, what happens when we are busy ‘doing language’

 Experimental technique
Online method: Online methods are important as they take us to the processing of
language as it happens. This process is not accessible once the task is over.

 This means trying to understand the underlying mechanism as the participant is

doing some ‘tasks’.

 These tasks are designed in a way so as to help us understand the effect of one
variable at a time or the interrelationship between more than one variable.

 For example, prime and the effect of masking condition can be combined to
investigate bilingual language comprehension in L1 or L2.
Lexical processing
Processing at word level
 Methods and research questions

Questions:
 Are the two languages of a bilingual active simultaneously?
 How are the two systems (lexical, conceptual, syntactic) connected?

Methods:
 Behavioural: Memory measures, Reaction Time studies, etc.
 Non-behavioural: brain mapping, eye tracking etc.
 Memory measures

 Rationale: People’s memory for a certain expression could indicate how the
expression was mentally represented at the time of memorizing it. Particularly,
the errors indicate the representations constructed.

 Two stages: Learning and Test. Typically there would be a gap of certain time
between the stages, when the participant engages in an unrelated task, to clear
the working memory from the task just performed.

 Then they are presented with task of free recall or old/new recognition task.
Free recall

Stage 1 Stage 2
 Bird  Recall as many words as you
 can from the list
House
 Elephant
 Ball
 Ethiopia
 Tree
 Monkey
 River
Recognition

Stage 1 Stage 2
 Bird  Was the word presented before?
 House
 Elephant  1. baboon
 Ball  2. Mountain
 Ethiopia  3. ball
 Tree  4. house
 Monkey
 River
 Reaction time methods

 Reaction time, or RT, is used as a dependent variable in many experimental

paradigms.

 This is the output data of the experimental work.

 This refers to the time taken to react after the onset of the stimuli.

 The RT is the result that is calculated as a function of the type of stimulus and
its relative processing difficulty.

 This means, that the more complex the stimulus, more will be the RT.

 This variable is widely used.

 Reaction time method
 RT is a productive method, used for a range of processing studies.

 Some of them are:

 Lexical decision
 Priming paradigm
 Translation recognition
 Self paced reading
 Sentence processing
 Picture/digit naming etc.
Experimental tasks can be divided into two
broad categories:

 Comprehension tasks

 Production tasks
 Comprehension tasks

 The task where the participants need to comprehend/understand the given

stimulus.

 That understanding is checked with a yes/no question. The participants need to

answer by pressing the assigned keys on the keyboard of the computer.

 So, after showing a word on the screen, the participant can be asked a question,
like is this a word? Is it meaningful? Is it a possible structure? And so on.

 For example:
 Display: unbuttonable Vs buttonable
 Probe question: is the word correct?
 There are many types of comprehension tasks based on this basic
structure.

 The design of the experiment depends on the research question.

 Thus, there are number independent variables that are studied using the
dependent variable of RT.
 Lexical Decision Task
 How is it done: A string of letters flash on the screen. The
participants need to decide if it is a real word or not. The method to
react is key press (customizable).

 What is measured (dependent variable):

 The RT, i.e. the time taken to press the key.
 Response accuracy: whether the answer is correct or not.

 Relation to mental lexicon:

 Mental lexicon needs to be accessed for this task.
 Real words will find the mental entry, non-words will not.
 GIRL/GIRK/BIRK/BARK and so on.

 Effects:
 Frequency
 Early learned Vs late learned words
 Priming paradigm

 Process: the target word is preceded by another word, which is

called the prime. Can be used in Lexical decision task

 Dependent variable: RT

 Priming effect:
 Related primes lead to faster recognition
 This relation can be semantic, phonological, orthographical
 Masked Vs non-masked prime (do the reading from reference list)
Example:
 ‘Bread’ followed by ‘butter’ Vs ‘bread’ followed by ‘cycle’

 If the two words are related in any way, the recognition is faster.

 Because, reading ‘bread’ activates the related concepts, and when one
candidate from those related concepts is the target word, recognition
happens faster.
 Masking condition
 Masking conditions are used with priming.
 There are two types of mask: forward and backward, depending on whether they
appear before or after the prime.
 A backward mask helps to minimise the effect of prime on the target

Source: Francis et al., 2022

 TE recognition

 Two words follow one another.

 The probe question: is the second word a translation of the first?

 E.g.
 Is ‘rasta’ a translation of ‘road’?
 In this simple task, many variables can be incorporated, like word
frequency, translation direction etc.
 Semantic
categorization task

 Can be a bilingual
processing task

 Can involve both, L1 and

L2
Production (Reaction time)
 Types of task
 Word level
 Picture naming: Line drawing appear on the screen and subjects need to name them
aloud.
 Digit naming

 Syntactic level
 Syntactic priming
 Recall task

 Reaction time in production studies measures the speech onset latency.

 That is, how quickly subjects start the utterance (word /sentence).
Source: Lucas, C. W et al. 2021. The Cologne Picture Naming Test for Language Mapping and Monitoring (CoNaT): An Open Set of 100 Black and
White Object Drawings. Frontiers in Neurology.
Digit naming

Source: Bakos, Sarolta et al, 2019

Picture-Word Interference Paradigm

 This is a manipulation of picture naming task.

 The subjects need to name a picture (given on the screen)

while:
 Hearing a word, through headphone,
OR
 Seeing a word written (superimposed) on the target picture
First, a bit of history

 Early research in bilingualism (1950s -1970s)

 Classifying bilingualism
 Degree of bilingualism and language dominance
 Lexical organization
 Bilingual processing
 Classifying bilingualism
 All bilinguals are not same.

 Weinreich (1954) was the first to identify and classify bilinguals into three groups
 Compound, Coordinate and Subordinate

 Osgood (1954) also made a distinction but a two-way system

 Compound and Coordinate

 Among these two, Weinreich made a distinction between lexical and conceptual
representation.

 This was also what De Groot (1993) incorporated in her model (RHM) many years
later.
 Assessing degree of bilingualism and language
dominance
 The beginning of psychological perspective on bilingualism is traced to Johnson
(1953).

 He assessed speakers' degree of bilingualism by comparing the number of words

they could produce

 Lambert (1955) was among the first to use reaction Time (RT) for this purpose.

 His study showed that bilinguals reacted faster to directions given in their dominant
language and this difference between the languages decreased as L2 proficiency
increased.

 Lambert, Havelka and Gardner (1959) developed a set of tests, including RT

(developed by Lambert 1955) and other six types of tests.
 These tests were
 Word completion test
 Word detection test
 Word naming test
 Verbal response set test
 Translation test

 They used these tests with RT measures, and found that

 ‘Bilingualism is reflected in many aspects of linguistic behaviour’
 These were followed by Ervin (1961) and Magiste (1979)

 Ervin used picture naming latencies in two languages to determine language

dominance

 Magiste used two encoding and two decoding tasks:

 Encoding: Picture naming and digit naming

 Decoding: Following directions and reading words aloud
 Bilingual lexical organization
 Initial psychological studies in bilingualism in the 1950s also had a close
association with research in human memory system.

 As a result, studies using language were often part of the larger investigation into
human memory

 Within this, the separate and shared memory hypothesis were studied

 Kolers (1963;1965) word association and free recall study seemed to favour
separate hypothesis.

 These two types of studies dominated bilingual memory research for some time
 Free recall task

 List of words were presented in two conditions: single language and mixed
language.

 Then the participants were asked to recall as many words as they could.

 The performance between the two conditions were checked to assess the
organization of memories of the two languages.
 Word association task
 Words in one language and their translation in another language were presented as
stimulus.
 Participants were then asked to provide the first word that comes to their mind, in
any one of the languages.

 For example, if the word pair is house-ghar (English Hindi) and participants answers
are road-rasta Or road-khidki.

 The first is example of interlingual translation response and higher number of such
responses were taken as evidence for shared hypothesis.

 Kolers (1963) found a low percentage of such responses and hence endorsed
separate memory hypothesis.
 Since both of these methods depended on using individual words as
stimulus, research on memory organizations somewhere tilted towards
lexical organization.

 Gradually studies with a separation between lexical and conceptual

presentation started (Kolers 1966) .

 This picked up momentum in the 1970s.

See also: Liepman & Saegert (1974); Winagrad, Cohen & Barresi (1976); MacLeod (1976); Paradis
(1978).
 Exploring bilingual processing

 In lexical processing domain, studies on language switching and interference was

also an important development.

 This line of work was motivated by two main ideas:

 The concept of language switch (Penfield & Roberts, 1959). A property that
allows bilinguals to switch on/off one language so that it can be kept from
interfering with the other.

 Second was the ongoing work on separate Vs shared memory hypothesis, using
cross-language interference tests.
 Language switch

 If bilinguals took more time to read 100 words from a mixed list as
compared to a single language list, it was due to the time it takes to
switch from one language to another.

 This could also mean that the two languages are represented
separately, leading to the switch in the first place.
First study to report switch cost was Kolers (1966).

 Three tasks:
 Read passages for comprehension
 Read passages aloud
 Read and then orally summarize the passages

 The first two tasks had three different conditions:

 Single language
 Alternating language between sentences
 Mixed language, both within and across sentences

 In the third task participants had to summarize in:

 One language
 Alternating languages
 Mixed language
 Result:
 First task did not show any impact of mixed language condition
 Second and third task showed the effect of mixed language condition

 Thus, this study provided evidence for language switch

 Soon afterward, a number of studies investigated the same and confirmed the
findings.

See also: Macnamara (1967); Macnamara, Krauthemmer & Bolgar (1968); Macnamara & Kushnir (1971);
Caramazza, Yeni-Komshian & Zurif (1974); Neufeld (1976)
 Cross language interference
 In such studies, the focus was to find out if performance in language A would
affect the performance in language B.

 Penfield and Roberts (1959) suggested that this will not happen since the other
language is ‘switched off’.

 Thus, if interference is proved, it will question the idea of language switch.

 And it will provide support for shared memory hypothesis.

 Studies in this domain usually use within language and between language
tasks. (this has been discussed before. Hence not detailing here)
 1980s
 From the beginning of this decade, hierarchical models of distinct lexical and
conceptual representations were popular.

 An important symposium in 1981, focused on how the different types of surface form,
like picture Vs words; L1 words Vs L2 words are connected to the underlying shared
conceptual system

 Five studies were presented in this symposium and then they appeared in the 1984
issue of Journal of Verbal learning and verbal behaviour.

 Two of these went on to become torch-bearers of sort for future studies.

 These are:
 Potter et al., (1984)
 Scarborough, Gerard & Cortese (1984)
 Another change during this time was in terms of methodology:

 Prior to this, response type and accuracy were main measures.

 RT was limited to language dominance and degree of bilingualism.

 And sometimes in case of language processing (Dalrymple-Alford & Budayr, 1966; Hamers &
Lambert, 1972; Kolers, 1966; Macnamara et al., 1968; Preston & Lambert, 1969).

 Word association and free recall studies were more common.

 However, during 80s, computers were available in large numbers, and this made RT
studies more feasible.

 As a result, investigation into bilingual lexical representation in hierarchical model with

RT data is a significant development of this decade.
 1990s
 This decade saw a rapid growth of research on lexical representation and processing
in bilinguals.

 Several state-of-the-art reviews came out, highlighting the progress made thus far
and providing an in-depth analysis of various issues at hand. (de Groot & Kroll 1997;
Harris 1992; Schreuder &Weltens 1993).

 Two main journals of the field stated during this decade. These are:
 Bilingualism: Language and cognition (1998)
 International Journal of Bilingualism (1997).

 These are journals dedicated to psycholinguistic studies on bilingualism

 On the research front

 Rapid growth of research output

 Diverse paradigms adopted

 Theoretical sophistication

 New phenomena uncovered

 New questions emerged

 One of the most significant developments of this period was proposals of
several new models on representation and processing.

 Models of representation
 Distributed conceptual feature model (DCF)
 Revised hierarchical model (RHM)

 Models of language interaction during processing

 Bilingual interactive activation model (BIA)
 Inhibitory control model (IC)
 Cross–language priming studies

 Cross language priming studies had already begun in the 1980s.

 In this decade, the studies included newer varieties in terms of:

 Word types:
 Concrete Vs abstract
 Cognate Vs non-cognates
 Prime-target relationships:
 Translation pairs
 Associative and semantic pairs
 Phonologically or orthographically related pairs
 Similarly, comparison across tasks also gained popularity:

 Lexical decision and semantic association tasks

 Participants differing on proficiency levels

 Different prime duration and so on.

 A significant new development

 Masked priming:

 De Groot & Nas (1991); Williams (1994) showed that masked condition had effect on
cross language priming in some cases, like non-cognate associative priming.

 This paved the way for using masked Vs unmasked conditions in processing
based studies.
 Major findings from cross language
priming studies during this time
Asymmetry in priming:

 Priming effect from dominant language to the less dominant language, not vice
versa.

 Both unmasked and masked conditions, cognate/non-cognate yielded same result.

(Altarriba 1992; Keatley, Sprinks & de Gelder 1994; Gollan, Forster & Frost 1997)

 Manipulation of factors, such as cognate status, prime-target relation and tasks

resulted in different outcome.
(De Groot & Nas 1991; Sanchéz-Casas, Davis & Garcia-Albea 1992; Williams 1994)
 The 2000s

 Masked priming studies have come a long way now.

 It is more complex than initially thought to be.

 Differences in findings with respect to lexical and semantic priming were

reported.

 Asymmetry in priming has been an important theoretical point.

 Similarly, switch cost asymmetry (or even absence in some cases).

 Are the lexical entities in both languages of a bilingual simultaneously
activated?

 Cross linguistic overlap (orthography, semantics and phonology)

 Either overt or covert
 Its role in bilingual lexical processing
has been utilized to find the answers

 In a series of experiments Dijkstra et al (1998) investigated whether such

overlap facilitates processing

 Also interesting was to see if degree of overlap has any role

 Facilitation would signal non-selective access, whereas no facilitation would

mean selective access.
 English LDT with homographs
 OP, O and P false friends Dutch and English

 OP [15]: /spot/ [mockery in Dutch]

 O [15]: /glad/ only same orthographically, but not phonologically [D: xlaf; slippery]

 P [15] : cow/kou [kou= cold in Dutch]

 Control words: 15 matched words in English each in three conditions. These words
have no counterpart in the other language.

 Non-words: 150 [50 English sounding, 50 Dutch sounding and 50 neutral].

 Before the experiment, the participants read written instructions in English.

 The participants were instructed to press a button with the index finger of their
preferred hand if they thought the letter string was an English word

 And to react with their other hand if it was a nonword.

 Result: O items showed faster processing. OP items did not show any RT
difference
 Only P items showed inhibition.

 This points toward selective access.

 Cognates
 Three categories of cognates were used in this study

 With SOP, SO and SP pairs

 SOP e.g., film , tent, hotel etc.

 SP:. Wheel/weil; homophonic cognate

 SO items, ‘Fruit’ is pronounced / frut / in English, but /fryt/ in Dutch

 Findings suggest a strong facilitation in all cases except in only phonological

match conditions.

 For homophonic cognates, there was no effect.

 Neighborhood effect
 The 'neighborhood size’ of a word is the number of words that can be created by
replacing one letter of that word.
 For instance, the word ‘chair’ has a neighborhood size of two (only ‘chain’ and
‘choir’ can be formed by replacing one letter of the word ‘chair’.
 Van Heuven et al. (2014) found that word recognition depends on the
neighborhood size of the word in both languages.
 So, the speed with which Dutch–English bilinguals recognized an English word
like /farm/ did not only depend on the number of English neighbors (e.g., firm,
fart), but also on the number of neighbor this English word has in Dutch (e.g.,
darm [colon], faam [fame]).
 This strongly suggests that lexical, orthographic representations from one’s native
language are active during word recognition in one’s second language.
 Phonological effect

 Phonological priming has been studied through cognates and homographs.

However, only a few studies have tried to investigate phonological cross-lingual
interactions more directly.

 Bilinguals are faster to recognize words from their second language if these words
are primed by non-words that sound like that word (pseudo-homophones).
[Brysbaert, Van Dyck and Van De Poel (1999); Duyck et al. 2004].
 For instance, Dutch–French bilinguals will be faster to recognize the French word sourd
[deaf ] if they saw the non-word soer just before, which sounds like the French sourd if
it is pronounced in Dutch.
 Using the same stimuli, but with French–Dutch bilinguals this time, Van Wijnendaele
and Brysbaert (2002) showed that the influence across languages also goes in the
opposite direction.
 They found that L2 pseudo homophone primes (soer which sounds like sourd if
pronounced in their second language) also facilitate the processing of L1 targets
(sourd).
 This strongly suggests that access to phonological representations is not selective
with respect to language.
 Non-word: /Soer/ [pronounced in Dutch way /sourd/]. This is similar to French word
/sourd/.
Word naming study with word body neighbors

 Word body: Words that share their medial vowel and final consonant with the
target word.
 E.g., save and wave are word body neighbors

 Jared and Kroll (2001) showed that people were significantly slower to pronounce
words in their second language when a part of that word could be pronounced in a
different way in their first language, even when the experiment was entirely run in
their second language.
 For instance, French–English bilinguals took longer to pronounce the English word
‘bait’ than the English word ‘bump’,

 because ‘bait’ contains ‘ait’, which is pronounced differently in French, whereas the
word part ‘ump’ is not used in French.

 The opposite effect was not always found.

 Auditory word processing
 Cross language facilitation effect is seen also in case of auditory processing.
 For example:
Weber and Cutler (2004)
 Subjects: Dutch–English bilinguals

 Eye tracking test: Look at a display while listening to an auditory stimuli. Lexical
competition in non-native spoken word recognition.

 Hearing words in English (e.g., desk) made longer eye fixations on pictures with
names in Dutch that were phonologically related to the English word (e.g., a picture
of a lid, for which the name in Dutch, deksel, starts with the same sounds).
 However, the opposite effect (influence from English names for pictures
on Dutch words) was not found:

 Dutch–English listeners did not look longer at a picture of a desk when they
heard the Dutch word deksel.

 These results suggest that interference from phonological

representations from the other language is larger when processing
one’s second language than when processing one’s native language.
 Similar findings were reported by Marian, Spivey and Hirsch (2003)

 Russian– English bilinguals instructed in English to ‘pick up the

marker’ often looked at a stamp, because its Russian translation
(marka : stamp]) has phonological overlap with the English spoken
word.
 Semantic and associative similarity

 Associative relatedness is a normative description of the probability that one word

will call to mind a second word.

 Associative relations are assumed to reflect word use rather than word meaning
(e.g., “needle-thread,” “spider-web”).

 On the other hand, semantic relatedness reflects the similarity in meaning or the
overlap in featural descriptions of two words (e.g., “whale-dolphin”, “duck-chicken”).

 It is possible for words to be either highly associated yet semantically dissimilar (e.g.,
“coat-rack”) or weakly associated yet semantically similar (e.g., “radish-beet”).
 Semantic priming

 Lexical decisions are faster when a word is immediately preceded by a semantic

associate than by an unrelated word or presented in isolation.

 Studies examining semantic priming across several languages found that this
advantage for a related prime held even when it appeared in a language different
from the target language.
 Semantic priming without association: whale-dolphin, missile-bomb,…[robust
priming effect]

 Associative priming without semantic overlap (phrasal associates): robust finding

help-wanted .

 But there have been issues of methodological issues with this design.
 By and large the finding that semantically related prime can facilitate the target has
been taken as evidence that L1 and L2 share the conceptual level.

 There are not separate sets of conceptual units for the two languages.

 This view of course does not preclude the existence of some conceptual units that
could be specific to one language,

 For example, if that language has a word expressing a concept that does not have a
particular lexical correlate in the other language
 However, a similar asymmetry is visible here too

 Semantic priming effects are generally larger when the primes are in L1
and the targets in L2 than vice versa in both same-script and different-
script bilinguals, especially when the SOA is relatively short.

 In contrast, the effect has been shown to be equally large in both

directions with longer SOA

(Frenck & Pynte, 1987; Kirsner et al., 1984).

 A study on Bodo-Assamese bilinguals used the
following design

Prime Type Word Nonword Word Nonword

Associated word फोथा−কপাল फोथा−মপাল उथुमाय−পেত उथुमाय−ঘেত

pair “bindi−forehead” “navel−stomach”

जोगोनार−কপাল जोगोनार−মপাল राव−পেত राव−ঘেত

Control pair
“pumpkin−forehead” “language−stomach”
 Result

 In unmasked condition, effect was seen only among early bilinguals. Not late
bilinguals. Also L1-L2 effect was seen but not L2-L1.

 In masked condition, there was no effect in either direction, pointing to the

fact that there is no automatic priming for word pairs that are only
associatively related but not semantically related.

 Masked priming studies are employed to test automatic and not strategic
processing.
 Metaphor processing

There are two general models proposed in the monolingual literature on

metaphor processing:

1. Direct access model: Assumes that in the course of language

processing, non-literal interpretation of a metaphoric expression may be
accessed directly ‘without first requiring an initial interpretation
computed and rejected’. Though it is possible that the literal expression
is accessed initially, but it is not obligatory.

2. Indirect Access Model: Assumes obligatory access to metaphor’s literal

meaning, and only if the literal interpretation does not fit, a search for
the non-literal interpretation is initiated.
 Now, how do bilinguals process metaphors?

 Studies found evidence that proficiency is directly correlated with metaphor

understanding in a bilingual’s L2.

 In a study [Johnson and Rosano, 1993] L2 learner participants were given metaphors
and asked to provide as many interpretation as possible.

 The results showed that the metaphor understanding capacity correlated with L2
fluency, with high proficient L2 speakers performing on par with monolinguals.
Heredia and Muñoz (2015)
 In another study, highly fluent bilinguals from heterogeneous backgrounds listened
to a short passage and named visually presented targets [in L2] that were
figuratively [boxer] or literally [pastry] related to the metaphoric reference
mentioned previously [creampuff, describing a weak and soft fighter].

 The crucial manipulation was the onset of the target word, either at 0 ms or 1000
ms after the passage.
Design:
 Auditory stimuli:
Stu went to Saturday night fights. There was one boxer he hated. This guy always
lost. Just as the match was about to start Stu went to get some snacks. He stood in
line for 10 minutes. When he returned the bout had been canceled. “What
happened?” he asked a friend. The friend replied, “the creampuff /fighter/referee did
not even show up” [metaphoric/literal/baseline control]
 Pastry/boxer
 Participants were instructed to listen to sentences presented over
headphones, understand them, and to name the string of letters
appearing in the middle of the computer screen as quickly and as
accurately as possible.
 At 0 ms post-offset, bilinguals showed evidence for the availability of the nonliteral
interpretation of the metaphoric reference and no evidence for the activation of
the vehicle’s word-level literal related meaning

 However, by 1000 ms post-offset, the results showed evidence for the availability
of the literal interpretation, and no activation for the nonliteral interpretation.

 Hence time seems an important predictor for processing words at metaphoric Vs

literal level.

 Other studies have also looked at this feature.

Production studies: Switching cost
 The Inhibitory Control model
 Proposed by David Green (1986; 1998)

 This is concerned with language control in bilingual language

production.

 Deals with the mechanism involved in deciding which language to use

while doing a specific task.

 Four components of the control system:

1. Conceptualizer
2. Supervisory Attentional System (SAS)
3. Language Task Schema
4. Lexical Semantic System
Green 1998
 Conceptualizer specifies the intended message to be communicated based on
current goal. For a bilingual speaker, production begins with a goal or an
objective.
 Based on the goal conceptualizer specifies the conceptual content to be
activated
 SAS selects the right language task schema to be performed.
 Language and word selection take place at the lemma level.
 Activated lemmas are checked with the activated concepts to make sure
correct mapping
 This triggers the inhibition of all lemmas with incorrect language tags.
 The model suggests that stronger language is more difficult to inhibit.
 Green’s model of activation, inhibition and control draws our attention to
the language control processes of a bilingual.

 First, the bilingual mind is capable of preventing interference from the

non-target language, even though two languages are available.

 Secondly, these people also have language control mechanism in place

that is employed in changes of target language.

 Hence, language control takes care of both inhibition and control of

execution of language production
 Here the notion of switching cost is a very important one.

 This refers to the processing cost involved in switching between two

languages.

 It is reflected in terms of greater RT when there is a switch between

languages as opposed to when there is no language switch while
performing a task.

 This phenomenon is observed both in studies involving switch between

languages as well as tasks.
 But first let us look at the studies that sought to establish that language
mixing incurs cost in terms of RT.

 Typically these studies compared language processing in single language

block and mixed language block.

 The results were almost unequivocal.

Bilingual picture naming task:

The behavioral part of the fMRI study by Hernandez et al [2000].

 In this study the participants were Spanish English bilinguals who reported English to
be their dominant language.

 180 pictures had to be named in single and mixed-language blocks.

 The main results revealed the mixing cost with single language blocks having lower
RT than mixed blocks [English 1010 ms Vs. 1276 ms; Spanish 1148ms Vs 1298ms]
 In Festman [2012] balanced Russian English bilinguals took part in a picture naming
task .

 An alternate design of 2 trials in Russian followed by 2 in German followed by 2 in

Russian and so on.

 A total of 240 pictures were used.

 Results showed two groups perform similarly on both non-switch and switch trials.
 However, in yet another study by Gollan and Ferreira [2009] had unbalanced bilinguals for
a picture naming task.

 A crucial aspect of this study was that the switch was voluntary as opposed to inbuilt in the
design.

 Which means that the participants were free to choose which language to use to name the
pictures in.

 The finding showed that the subjects chose to use L2 [weak language] to name pictures
whose names were frequent words and used L1 [stronger language] to name those which
were less frequent.

 This was taken as evidence for low or absent switch cost when switching was voluntary.
 Digit naming

 Pictures and digits are similar in that they provide no cue on how to pronounce the
concept (i.e. there is no grapheme part to it), they do differ on a variety of levels.

 Compared to picture naming, digit naming is less demanding because digits have less
visual complexity, are more frequently used and are usually monosyllabic.

 First, digits represent an abstract concept, namely quantity, whereas pictures

represent concrete objects.
 A consequence of this difference is that the actual representation of digits is arbitrarily
mapped to their meaning, which is not the case for pictures.
 A second difference is that digits represent a specific semantic group(e.g.,
Cappelletti, Warrington, & Butterworth, 1995; Herrera & Macizo, 2010, 2011), which
is not the case for all picture sets.

 Related to this is the question of whether digit and picture naming involves the
activation of semantic information.
 While it is generally agreed that picture naming involves activation of semantic
information (Glaser & Düngelhoff, 1984; Smith & Magee, 1980), there is no
conclusive evidence as to whether digit naming requires activation of semantic
information.
 Experiments

 Wang et al [2009] tested 15 Chinese-English unbalanced bilinguals response latencies

[RT] increased for both languages in the mixed paradigm as compared to the single
language task.

 Prior and Gollan [2011] conducted study on two high proficient bilingual groups:
Spanish-English and Mandarin-English.

 Though the mixed language task incurred more RT compared to pure language [single
language] blocks,
 There was no difference in the switching between dominant and non-dominant languages.
 Hence
 Switch cost is part of production tasks.

 Balanced Vs unbalanced difference is noticed

 Similarly, there is difference between voluntary Vs non-voluntary switch

 Now, the next question:

Is switch cost same or different for L1-L2 Vs L2-L1 ?
 Understanding switching cost asymmetry

 When two tasks are of equal difficulty, the switching cost is symmetrical in either
direction. That is, it takes the same time to switch from Task A to Task B and vice
versa.

 However, if one task is more difficult than the other, there is asymmetry because
it takes longer to switch to the easier task.

 Inhibition plays a major role in explaining this. Easy tasks need to be inhibited
more strongly than the difficult ones.

 Thus, more time is needed to switch back to the easy task as the strong inhibition
applied to it has to be overcome
 For bilinguals the same asymmetry in switching cost is observed.

 Larger switch cost is usually observed when switching into a more dominant
language than the weaker one.
 The first important study in this regard was done by Meuter and Allport in 1999.

 They had 16 non-balanced bilingual participants who had English either as their L1 or
L2.

 The task was to name aloud the numerals presented on the screen.

 The crucial manipulation was the task switch cue through the changing color of the
rectangle within which the numbers appeared.

 Larger switch costs when switching into a more dominant language [L1] as opposed to
L2 as dominant language is difficult to suppress and hence more difficult to retrieve.
 Some other studies : Unbalanced bilinguals

 Participants: 40 participants: 20 from Padova University in Italy with Italian as L1 and

English as L2. Another 20 from Trent University Canada, with English as their L1 and French
as L2.

 Stimuli: Numerals and number names.

 Switch cue: Colored box.

 Result: switch trials were 37 ms slower than non-switch trials. Size of switch cost was larger
for numerals than for number names. Switch cost to L1 [50 ms] was larger than L2 [25ms].

(Reynolds, Schloffel & Peressotti 2016)

 Similar findings have been reported from another study as well

 Participants: German English bilinguals.

 Cue: colored square in the middle of the screen
 Result: smaller switch cost in digit naming than in picture naming.
 However, another study by Costa (2004) proved that when the proficiency in L1
and L2 are equal, the switching cost is symmetrical.

 In many cases, the asymmetric switch cost can be explained by the IC model,
discussed before.

 The Inhibitory Control (IC) model suggests that, in both production and
recognition, the amount of inhibition on the non-target language is greater for the
stronger compared to the weaker language.
 Gullifer, Kroll and Dussias [2013]
 Investigated inter-sentential switching cost in Spanish English bilinguals.
 The participants had to read sentences word by word and then name the target word aloud.
 They did not find any switch cost difference

 Macizo, Bajo and Paolini [2012]

 Conducted a word reading task with 20 Spanish English bilinguals.
 Participants had to read aloud 144 experimental words in a combination of switch and non-
switch trials.
 The result showed cost on switch trials and higher cost for L2-L1.
 Adaptive Control Hypothesis (Green and
Abutalebi, 2013)

 More recently, the ACH model has proposed a mechanism that connects
language control mechanism with interactional context.
 Interactional context: “Recurrent pattern of conversational exchanges within a
community of speakers”
 The main idea is that control processes are not automatic, rather they adapt to
the demands placed on them by these contexts.
 Hence speakers from different interactional contexts will display differences in
their control mechanism.
Three interactional contexts

 The interactional contexts are based on the level of code-

switching in the speaker’s linguistic environment.

 These are:
 Single language context
 Dual language context
 Dense code-switching context
 The latest developments in bilingual language processing points to a
more dense relationship between language and other mental
mechanism, involving control.

 More on this in another segment.

Sentence processing
 What is sentence processing?
 A process through which the meaning of a sentence is understood.

 This in turn, has three main components: word recognition, parsing and semantic
integration.

 At the first stage: lexical information of individual words, like meaning, syntactic
properties, morphological structures, lexical associations etc become available

 Parsing: syntactic structure is built on the basis of the retrieved lexical information,
morphosyntactic cues (word order, case marking) and an individual’s syntactic
knowledge.

 Finally, semantic integration: mental representation of the meaning of the sentence

based on derived input and the individual’s linguistic and world-knowledge.
Different focal domains within
sentence processing studies
 Word recognition within sentence context

 Mental representation of meaning in sentence processing

 Parsing, the process of building a syntactic structure

 Syntactic parser

 Sentences are parsed [breaking it down to its parts] to understand the

meaning.

 Some grammatically complex sentences are easy to parse [Ram saw the
ghost who harassed Sita the other day].

 Some easy sentences are difficult to process [the horse raced past barn
fell].

 Difficulty is often connected to ambiguity.

 Parsing and ambiguity
 Ambiguous sentences are useful to study sentence processing.

 Because these offer an opportunity to examine how the parser handles incoming
information

 Which is to be used for building the syntactic structure in the mind

 Questions that can be studied:

 Does the parser focus on only one interpretation first, followed by the others?
 Factors responsible for one or the other interpretation
 Recovery from the incorrect interpretation etc
 Or does the system allow for multiple interpretations at the same time?
 Some examples of ambiguous sentences

 “I went to the bank.”

(The bank could be a place where money is kept, or it could be the edge of a
river.”

 “Visiting relatives can be exhausting.”

(What is exhausting: when relatives visit you, or when you visit relatives?)

 “The cat chased the mouse until it stumbled and fell.”

(Which fell, the cat or the mouse?)
 Ambiguous sentences help us understand the parsing strategies better.

 These are explored through garden path effect (GPE)

 A temporarily ambiguous sentence resulting in initial incorrect interpretation is called

garden path sentence

 GPE was studies by Bever, who made the sentence famous :‘The horse ran past the barn
fell’

 His work connected delay in response latency in reading comprehension with garden path
sentences.

(Bever 1970)
 Another form of GPE is known as filler-gap effect

 Examples:
A. My brother wanted to know who Ruth will bring us home to at Christmas
B. My brother wanted to know if Ruth will bring us home to Mom at Christmas

 Here the longer reading time of ‘us’ was explained as initial consideration of the Wh-
word as direct object of the verb ‘bring’, which must be corrected when ‘us’ is
encountered.

 In contrast this initial incorrect analysis will not happen for example B.

(Stowe 1986)
 Types of ambiguity
 Temporary ambiguity: Sentences that allow two types of interpretations in the middle of
the sentences, but it also contains a disambiguating element later in the sentences that
helps remove the ambiguity.
 He remembered the story was told by his uncle
 NP: Direct object of preceding verb OR subject of the succeeding verb

 Global ambiguity: These type of sentences do not have any disambiguating element and
hence the sentence remains ambiguous till the end.
 The police saw the robber with binoculars
 Who had the binoculars is not clear till the end
 Models of syntactic parsing
 The main question in sentence processing is which information is processed first
while processing sentences.

 Based on the answers, the models are called modular and interactive models.

 Modular models are based on the modularity of mind theory (Fodor)

 Interactive models are based on connectionist view of the mind.

 Garden path model (Frazier & Rayner 1982)

 Constraint based model (MacDonald 1994)

(Fodor 1983; Frazier 1987; MacDonald, Pearlmutter & Seidenberg 1994; Trueswell, Tanenhaus & Garnsey 1994)
 Garden path Model
 This is a domain specific model, handling a specific cognitive function.

 This function is, in terms of information, encapsulated only within that module.

 In this model, parsing takes place in two stages:

1. Stage one: generates syntactic structure based on syntactic information only
1. Minimal attachment principle
2. Late closure principle
2. Stage two: the initial structure is then evaluated and if needed, revised.

 This happens on the basis of both syntactic and non-syntactic information.

 Constraint-based model
 This model is based on connectionist view of mind

 According to this model, all information, syntactic and non-syntactic (lexical, semantic,
contextual) are immediately accessed during parsing.

 In view of this model, parser generates multiple structures simultaneously

 These candidate structures compete for selection

 And the final interpretation is based on multiple sources of information.

 Within this model, the garden path effect is an outcome of multiple competing
interpretations
 And not due to reanalysis
 Comparing the two models

GPM CBM
 Based on modular mind theory  Based on interactive view of mind
 Only syntactic information is initially  All information considered
considered in parsing
 Sentence processing is parallel
 Sentence processing is serial, with only
 Competition leads to garden path
one interpretation at a time
effect
 Reanalysis leads to garden path effect
Behavioural Methods: Bilingual
sentence processing
 Self-paced reading: Sentence processing

 Measurement of time taken to read words/phrases/sentences.

 There are different ways of measurement depending on how fine

grained the measurement is sought to be.

 Currently, the most common way is to present sentences word-by-

word.

 This way, one can measure the time taken for each part of the
sentence.
Range of issues that can be investigated
with this method
 Syntactic: ambiguous sentences, scrambling, unambiguous but
structurally complex sentences etc.

 Interpretation of pronouns and reflexives

 Effect of priming
 Criticism

 Lack of ecological validity

 Real life reading is not reflected in this design
 Unnatural way of moving forward by pressing a button, as opposed to
automatic eye movement

 Spill-over effect
 Often the effect of a word is not visible till a word or two later
 Item recognition measures

 Consists of recognizing whether or not an item was present in the text presented
preceding the task.

 The latency of the decision reflects the relative accessibility of the term. ‘Negation
decreases the accessibility of the negative noun’. (X baked some bread but no
cookies).
 Other methods

 Process of sentence processing is incremental. Method of analysing

sentence processing can include:

 Eye tracking
 ERPs
 Monitoring task
 Bilingual sentence processing
 Sentence processing in monolinguals and bilinguals are similar in some
sense, like both focus on parsing. However, in case of bilinguals, there is an
added question of influence of one language over the other while
processing:

 Thus, the research goals in bilingual sentence processing are:

 Parsing (and how language specific properties affect that)

 Sensitivity to morpho-syntactic violations

 Parsing
 How bilinguals parse syntactically ambiguous sentences for which solutions are
different between the two languages

 Often this meant investigating the role of L1 on L2 parsing:

 Verb subcategorization
 Relative clause attachment
 syntactic and lexico-semantic information in L2 parsing

 Also studied are areas like

 Role of AoA, proficiency, working memory
 Sensitivity to morpho-syntactic
violations
 A comparatively new trend that started in the 2000s.

 Grammaticality judgement, self-paced reading and eye tracking are main methods
used.

 Focus is more on acquisition than parsing:

 Whether L2 learners are able to develop native like competence in grammatical
morphemes and syntactic agreement involving these morphemes.

 How sensitive bilinguals are to the syntactic and/or semantic anomalies? Are they
equally sensitive as monolinguals?
Some studies
 Sentence processing and eye tracking
 Studies using eye movement have proven to be quite useful for investigating
human sentence processing (visual).

 Research on eye-tracking has investigated a wide range of questions which

include:
 Recovery mechanisms following garden-pathing,
 Timing of processes driving comprehension,
 Role of anticipation and expectation in parsing,
 Role of semantic, pragmatic, and prosodic information, etc.
Spivey et al (2002)

 Looked at eye movements in response to auditory garden path

sentences within the context of visual arrays.

 Participants were asked to “put the apple on the towel in the box”.

 The eye movements of the participants were recorded to assess how the
sentence was interpreted.
 As per the garden path model “on the towel” should initially be
understood as the place where the apple should be put because it is the
simplest syntactical structure.

 With the help of visual context, the following was found:

 When the visual context consisted of two apples, one on a towel and the other
on a napkin, participants rapidly interpreted “on the towel” as a way of
identifying which apple was to be moved.

 Results show that visual contexts reduce ambiguity and prevent garden path
effects.
 Sentence processing and working memory

Zhou et al. (2017)

 Aim: To investigate how working memory capacity (WMC) was related to
processing of wh-extractions in Chinese-English bilinguals.

 Participants: Chinese-English bilinguals, aged 19-26.

 Tasks: Grammatical judgment task for sentence processing and Operation Span
(OSPAN) to measure their WMC.

 Results: Those with high WMC had faster grammatical judgment of the critical
region in subject- and object-extractions.
 Code switching and sentence processing

 Comprehension of a language switch has been described as a “passive” process and

may be considered less costly than the more active process of producing a language
switch.

Gross et al., (2019)

 Aim: examined processing of code-switching in bilingual children.
 Participants: Spanish-English bilingual children, aged approximately 6–12.
 Tasks:
 Auditory moving window paradigm to measure children’s processing of code-switched
sentences.
 Cognitive control was measured by the Dimensional Change Card Sort.
 Results:
 Children exhibited significant costs in processing speed when
listening to code-switched sentences.

 No costs incurred in a measure of offline comprehension.

 Children with higher language skills exhibited a greater moderating

effect of cognitive control.
 Syntactic/semantic anomaly and ERP
 Example sentences:
1. The house has ten rooms in total.
2. The house has ten cities in total.
3. The turtles moves slowly.
4. The turtles move slowly.

 Sentences had syntactic/semantic anomalies.

 ERP was measured on critical words (rooms, cities, moves, move)

 For monolinguals different N400 effect is predicted between sentences 1 & 2

 Studies found no critical difference in this between monolingual , early and late
bilinguals.

 This means, all the subjects showed

 N400 effect on the critical words and
 the amplitude of difference between correct and incorrect critical words

 However, delayed N400 effect was found in L2. this means semantic integration is late
in bilinguals.

(Ardal et al 1990; Weber-Fox & Neville 1996)

 For syntactic anomaly, the critical ERP indicators are ELAN and P600.

 As such the pattern in these two are expected to be different between monolinguals
and bilinguals.

 In this case, however, a clear difference is visible between monolingual and bilinguals.

 Bilinguals were not found to have the ELAN and P600 effect due to syntactic violations.

 This indicated the first-pass parsing and second-pass reanalysis may be different
between these groups.

(Ojima et al 2005; Weber-Fox & Neville 1996)

 Current findings
 There is no fundamental difference between first and second language sentence
processing among sequential bilinguals.

 Proficiency, speed of lexical access, cognitive and computational resources might

indicate if the processing style will be similar.

 Not only there is L1- L2 influence, but L2 to L1 influence is also a possibility

 Age of acquisition does not seem to be a deciding factor

 Both behavioural and electrophysiological studies on time course of processing show

that L2 speakers may be able to learn even subtle aspects of L2
 This contradicts the view of hard constraints in native Vs non-native processing

 These findings point out that bilingual sentence processing a dynamic and
open system.
 Writing
 Psycholinguistic studies of bilingualism is not complete without also looking into
writing.

 Bilingual text production, reader-oriented writing in domains like academics,

educational and professional.

 There are two main strands of work :

1. Intra-subject studies: bilinguals’ writing in their different languages
2. Inter-subject or between – subject studies: bilinguals’L2 writing Vs monolinguals’ L1
writing;
 Two methods have been applied to study this domain:

1. Survey method
 By asking participants on their own writing process questionnaires and interviews

2. Laboratory setting controlled studies

 Here the participants are asked to engage in actual writing
 Writing as a process
 It’s a recursive process, not a linear one. This involves a number of sub-
processes, that interact with one another in a cyclical fashion.
 Planning: This is a conceptual activity. Knowledge of content, metacognitive
awareness of similar texts, lead to the planning of what the final product will
be.
 Advance planning and online planning.

 Formulation: transformation of the intended message into writing.

 Revision: solving mismatches between intention and linguistic form.

 Demands of bilingual writing

 Demands are linguistic and cognitive

 Linguistic level: the writer’s linguistic and genre knowledge. This adds to the
processing demands on the bilingual writer.

 Cognitive level: deals with the allocation of attention derived from the
cognitively demanding and problem-solving nature of writing.
 Research data

 Fluency: Like in verbal language proficiency, written language also has been probed
in terms of fluency.

 It is measured as total number of words written, or the number and length of pauses.

 Comparison can be between the L1 and L2 of the subjects.

 Typical findings suggest shorter texts in L2, more pauses, and lower number of
words per pause.

(Leki, Cumming & Silva 2008)

Attention allocation:
 Text generation is a process in both monolingual and bilingual writing, it is the
process of transforming ideas into language.

 Findings suggest bilinguals spend more of composing time in generating text: 50%
by monolinguals Vs upto 80% by bilinguals.

 Similar results were obtained in intra-subject studies too. Survey studies on Chinese
scholars in English also point to the same fact, noting that they found it difficult to
‘make claims about their research with appropriate amount of force’.

(Wang & Wen 2002; Manchon, Murphy & Roca De Larios 2005; Flowerdew & Li 2009)
 One interesting point in this domain is the debate on interaction Vs separation of
language abilities and writing skills in bilinguals’ L2 writing.

 Most research evidence suggest an independence of the proficiency and writing ability

 Writing ability already acquired in one language might get transferred to another
language if proficiency allows.

 Hence proficiency is an addition, but not primary prerequisite.

(Cumming 2009 Manchon, Roca De Larios & Murphy 2009)

 References
 Altarriba, J. (1992). The representation of translation equivalents in bilingual memory. In R. J. Harris (Ed.), Cognitive processing in bilinguals (pp. 157–174).
North-Holland. [Link]

 Ardal, S., Donald, M. W., Meuter, R., Muldrew, S., & Luce, M. (1990). Brain responses to semantic incongruity in bilinguals. Brain and Language, 39(2), 187–205.
doi:10.1016/0093-934X(90)90011-5

 Bakos, Sarolta & Mehlhase, Heike & Landerl, Karin & Bartling, Jürgen & Schulte-Körne, Gerd & Moll, Kristina. (2019). Naming processes in reading and spelling
disorders: An electrophysiological investigation. Clinical Neurophysiology. 131. 10.1016/[Link].2019.11.017.

 Bever, Thomas G. 1970. The cognitive basis for linguistic structures. Cognition and language development, ed. by John R. Hayes, 279–362. New York: Wiley &
Sons.

 Bilingualism and lexical representation

 Bock, Kathryne. 1996. Language production, Methods and Methodologies.

 Brysbaert, M., Van Dyck, G., & Van de Poel, M. (1999). Visual word recognition in bilinguals: Evidence from masked phonological priming. Journal of Experimental
Psychology: Human Perception and Performance, 25(1), 137–148. [Link]

 Caramazza, A., Yeni-Komshian, G., & Zurif, E. B. (1974). Bilingual switching: The phonological level. Canadian Journal of Psychology / Revue canadienne de
psychologie, 28(3), 310–318. [Link]

 Costa, A., & Santesteban, M. (2004). Lexical access in bilingual speech production: Evidence from language switching in highly proficient bilinguals and L2
learners. Journal of Memory and Language, 50, 491-511.

 Dalrymple-Alford, E. C., & Budayr, B. (1966). Examination of some aspects of the Stroop Color-Word Test. Perceptual and Motor Skills, 23(3, PT. 2), 1211–1214.
[Link]

 de Groot, A. M. B. (1993). Word-type effects in bilingual processing tasks: Support for a mixed-representational system. In R. Schreuder & B. Weltens (Eds.), The
bilingual lexicon (pp. 27–51). John Benjamins Publishing Company. [Link]

 de Groot, A. M. B., & Kroll, J. F. (Eds.). (1997). Tutorials in bilingualism: Psycholinguistic perspectives. Lawrence Erlbaum Associates Publishers.

 de Groot, A. M., & Nas, G. L. (1991). Lexical representation of cognates and noncognates in compound bilinguals. Journal of Memory and Language, 30(1), 90–
123. [Link]

 Dijkstra, T. et al. [Link] homograph recognition: effects of task demands and language intermixing. Bilingualism language and cognition. 1. 51-66.

 Dijkstra, T., Van Heuven, W. J. B., & Grainger, J. (1998). Simulating cross-language competition with the bilingual interactive activation model. Psychologica
Belgica, 38(3-4), 177–196.

 Duyck, P.-F., David, P. and Quilici, S. (2004), A review of relationships between interspecific competition and invasions in fruit flies (Diptera: Tephritidae).
Ecological Entomology, 29: 511-520. [Link]
 Ervin, S. and C. Osgood, 1954. Second language learning and bilingualism, In: C. Osgood and T. Sebeok (eds.), Psycholinguistics.
Suppl. J. Abn. Sot. Psychol. 49, 139-146.
 Ervin, S. M. (1961). Semantic Shift in Bilingualism. The American Journal of Psychology, 74(2), 233–241.
[Link]
 Festman, J., & Münte, T. F. (2012). Cognitive control in Russian–German bilinguals. Frontiers in Psychology, 3, Article 115.
[Link]
 Flowerdew, J., & Li, Y. (03 2009). English or Chinese? The trade-off between local and international publication among Chinese
academics in the humanities and social sciences. Journal of Second Language Writing, 18, 1–16. doi:10.1016/[Link].2008.09.005
 Foster, W. (1998). Editor’s Foreword. Educational Administration Quarterly, 34(3), 294–297.
[Link]
 Francis, Zoë & Jebanesan, Aravinth & Inzlicht, Michael. (2022). Leading Us Unto Temptation? No Evidence for an Asymmetry in
Automatic Associations Between Goals and Temptations. Collabra: Psychology. 8. 10.1525/collabra.31030.
 Frazier, L., & Rayner, K. (1982). Making and correcting errors during sentence comprehension: Eye movements in the analysis of
structurally ambiguous sentences. Cognitive Psychology, 14(2), 178–210. [Link]
 Frenck, C., & Pynte, J. (1987). Semantic representation and surface forms: A look at across-language priming in bilinguals. Journal of
Psycholinguistic Research, 16(4), 383–396. [Link]
 García, O., & Lin, A. (03 2016). Translanguaging in Bilingual Education. doi:10.1007/978-3-319-02324-3_9-1
 Glaser, W. R., & Düngelhoff, F.-J. (1984). The time course of picture-word interference. Journal of Experimental Psychology: Human
Perception and Performance, 10(5), 640–654. [Link]
 Glaser. 1992.
 Goldinger, S.D. 1996. Auditory lexical decision. Language and Cognitive Processes. 11.559-67
 Gollan, T. H., & Ferreira, V. S. (2009). Should I stay or should I switch? A cost–benefit analysis of voluntary language switching in
young and aging bilinguals. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(3), 640–665.
[Link]
 Gollan, T. H., Forster, K. I., & Frost, R. (1997). Translation priming with different scripts: Masked priming with cognates and
noncognates in Hebrew–English bilinguals. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23(5), 1122–
1139. [Link]
 Granville B. Johnson Jr. (1953) Bilingualism as Measured by a Reaction-Time Technique and the Relationship between a Language
and a Non-Language Intelligence Quotient, The Pedagogical Seminary and Journal of Genetic Psychology, 82:1, 3-9, DOI:
10.1080/08856559.1953.10533652
 Green DW, Abutalebi J. Language control in bilinguals: The adaptive control hypothesis. J Cogn Psychol (Hove). 2013
Aug;25(5):515-530. doi: 10.1080/20445911.2013.796377. PMID: 25077013; PMCID: PMC4095950
 Green, D. W. (1998). Mental control of the bilingual lexico-semantic system. Bilingualism: Language and Cognition, 1, 67–81.
 Green, D. W., & Abutalebi, J. (2013). Language control in bilinguals: The adaptive control hypothesis. Journal of cognitive
psychology (Hove, England), 25(5), 515–530. [Link]
 Gross, J. J., Uusberg, H., & Uusberg, A. (2019). Mental illness and well-being: an affect regulation perspective. World
psychiatry : official journal of the World Psychiatric Association (WPA), 18(2), 130–139. [Link]
 Gross, M. C., Lopez, E., Buac, M., & Kaushanskaya, M. (2019). Processing of code-switched sentences by bilingual children:
Cognitive and linguistic predictors. Cognitive development, 52, 100821. [Link]
 Gullifer JW, Kroll JF and Dussias PE (2013) When language switching has no apparent cost: lexical access in sentence
context. Front. Psychol. 4:278. doi: 10.3389/fpsyg.2013.00278
 Hamers, J. F., & Lambert, W. E. (1972). Bilingual interdependencies in auditory perception. Journal of Verbal Learning &
Verbal Behavior, 11(3), 303–310. [Link]
 Harris, R. J. (Ed.). (1992). Cognitive processing in bilinguals. North-Holland.
 Heredia, R. R., & Muñoz, M. E. (2015). Metaphoric reference: A real-time analysis. In R. R. Heredia & A. B. Cieślicka
(Eds.), Bilingual figurative language processing (pp. 89–116). Cambridge University Press.
[Link]
 Hernandez, A. E., Martinez, A., & Kohnert, K. (2000). In search of the language switch: An fMRI study of picture naming in
Spanish-English bilinguals. Brain and language, 73(3), 421–431. [Link]
 Jared, D., & Kroll, J. F. (2001). Do bilinguals activate phonological representations in one or both of their languages when
naming words? Journal of Memory and Language, 44(1), 2–31. [Link]
 Johnson, J., & Rosano, T. (1993). Relation of cognitive style to metaphor interpretation and second language
proficiency. Applied Psycholinguistics, 14(2), 159–175. [Link]
 Keatley, C. W., Spinks, J. A., & de Gelder, B. (1994). Asymmetrical cross-language priming effects. Memory & Cognition,
22(1), 70–84. [Link]
 Kirsner K., Brown H., Abrol S., Chadha N., Sharma N.
 Kolers, P. A. (1963). Interlingual word associations. Journal of Verbal Learning & Verbal Behavior, 2(4), 291–300.
[Link]
 Kolers, P. A. (1965). Bilingualism and Bicodalism. Language and Speech, 8(2), 122–126. [Link]
 Kolers, P. A. (1966). Reading and talking bilingually. The American Journal of Psychology, 79(3), 357–376. [Link]
 Leki, I., Cumming, A., & Silva, T. (01 2008). A Synthesis of Research on Second Language Writing in English. A Synthesis of Research on
Second Language Writing in English, 1–259. doi:10.4324/9780203930250
 Liepmann, D., & Saegert, J. (1974). Language tagging in bilingual free recall. Journal of Experimental Psychology, 103(6), 1137–1141.
[Link]
 Love, T. & D. Swinney. 1996. Co-reference processing and levels of analysis in object relative constructions: demonstration of antecedent
reactivation with the cross modal priming paradigm. Journal of Psycholinguistic Research. 25.1. 5-24.
 MacDonald, M. (05 1994). Probabilistic Constraints and Syntactic Ambiguity Resolution. Language and Cognitive Processes, 9, 157–201.
doi:10.1080/01690969408402115
 Macizo, P., Bajo, T., & Paolieri, D. (2012). Language switching and language competition. Second Language Research, 28(2), 131–149.
[Link]
 MacLeod, C. M. (1976). Bilingual episodic memory: Acquisition and forgetting. Journal of Verbal Learning & Verbal Behavior, 15(4), 347–364.
[Link]
 Macnamara, J. (1967), The Bilingual's Linguistic Performance—A Psychological Overview. Journal of Social Issues, 23: 58-77.
[Link]
 Macnamara, J. (1967), The Effects of Instruction in a Weaker Language. Journal of Social Issues, 23: 121-135.
[Link]
 MacNamara, J. (1967). The linguistic independence of bilinguals. Journal of Verbal Learning & Verbal Behavior, 6(5), 729–736.
[Link]
 Macnamara, J., & Kushnir, S.L. (1971). Linguistic independence of bilinguals: The input switch. Journal of Verbal Learning and Verbal
Behavior, 10, 480-487.
 Macnamara, J., Krauthammer, M., & Bolgar, M. (1968). Language switching in bilinguals as a function of stimulus and response
uncertainty. Journal of Experimental Psychology, 78(2, Pt.1), 208–215. [Link]
 Mägiste, E. (1979). The competing language systems of the multilingual: A developmental study of decoding and encoding
processes. Journal of Verbal Learning & Verbal Behavior, 18(1), 79–89. [Link]
 Manchon, M., Murphy, L., & Larios, J. (12 2007). Lexical Retrieval Processes and Strategies in Second Language Writing: A Synthesis of
Empirical Research. International Journal of English Studies, 7. doi:10.6018/ijes.7.2.49041
 Marian, V., Spivey, M., & Hirsch, J. (2003). Shared and separate systems in bilingual language processing: converging evidence from
eyetracking and brain imaging. Brain and language, 86(1), 70–82. [Link]
 Meuter, R. F. I., & Allport, A. (1999). Bilingual language switching in naming: Asymmetrical costs of language
selection. Journal of Memory and Language, 40(1), 25–40. [Link]
 Neufeld, R. W. (1976). Evidence of stress as a function of experimentally altered appraisal of stimulus aversiveness
and coping adequacy. Journal of Personality and Social Psychology, 33(5), 632–646.
[Link]
 Onifer, W. & D Swiney. 1981. Accessing lexical ambiguities during sentence comprehension: effect of frequency of
meaning and contextual bias.
 Pedro Macizo, Amparo Herrera, Patricia Román, & María Cruz Martín (2010). Second language acquisition influences
the processing of number words. Procedia - Social and Behavioral Sciences, 9, 1128-1134.
 Penfield, W., & Roberts, L. (1959). Speech and brain mechanisms. Princeton University Press
 Potter, M. C., So, K.-E, Von Eckhardt, B., & Feldman, L. B. (1984). Lexical and conceptual representation in beginning
and more proficient bilinguals, Journal ofVerbalLearning & VerbalBehavior, 23, 23-38
 Preston, M. S., & Lambert, W. E. (1969). Interlingual interference in a bilingual version of the Stroop color-word
task. Journal of Verbal Learning & Verbal Behavior, 8(2), 295–301. [Link]
 Prior, A., & Gollan, T. H. (2011). Good language-switchers are good task-switchers: evidence from Spanish-English and
Mandarin-English bilinguals. Journal of the International Neuropsychological Society : JINS, 17(4), 682–691.
[Link]
 Quarterly Journal of Experimental Psychology, 32 (1980), pp. 585-594
 Sánchez-Casas, R. M., Davis, C. W., & García-Albea, J. E. (1992). Bilingual lexical processing: Exploring the
cognate/non-cognate distinction. European Journal of Cognitive Psychology, 4(4), 293–310.
[Link]
 Scarborough, D. L., Gerard, L., & Cortese, C. (1984). Independence of lexical access in bilingual word
recognition. Journal of Verbal Learning & Verbal Behavior, 23(1), 84–99.
[Link]
 Schreuder, R., & Weltens, B. (Eds.). (1993). The bilingual lexicon. John Benjamins Publishing Company.
[Link]
 Shiro Ojima, Hiroki Nakata, Ryusuke Kakigi; An ERP Study of Second Language Learning after
Childhood: Effects of Proficiency. J Cogn Neurosci 2005; 17 (8): 1212–1228. doi:
[Link]
 Smith, M. C., & Magee, L. E. (1980). Tracing the time course of picture–word processing. Journal of
Experimental Psychology: General, 109(4), 373–392. [Link]
 Spivey, M. J., Tanenhaus, M. K., Eberhard, K. M., & Sedivy, J. C. (2002). Eye movements and
spoken language comprehension: effects of visual context on syntactic ambiguity
resolution. Cognitive psychology, 45(4), 447–481. [Link]
 Uriel Weinreich (1954) Is a Structural Dialectology Possible?, WORD, 10:2-3, 388-400, DOI:
10.1080/00437956.1954.11659535
 Van Hell, J. G & Ton Dijkstra. [Link] language knowledge can influence native language
performance in exclusively native contexts. Psychonomic bulletin and review. 9.780-789.
 van Heuven, W. J., Mandera, P., Keuleers, E., & Brysbaert, M. (2014). SUBTLEX-UK: a new and
improved word frequency database for British English. Quarterly journal of experimental
psychology (2006), 67(6), 1176–1190. [Link]
 Van Wijnendaele, I., & Brysbaert, M. (2002). Visual word recognition in bilinguals: Phonological
priming from the second to the first language. Journal of Experimental Psychology: Human
Perception and Performance, 28(3), 616–627. [Link]
 Vasishth, S., von der Malsburg, T., & Engelmann, F. (2013). What eye movements can tell us about
sentence comprehension. Wiley interdisciplinary reviews. Cognitive science, 4(2), 125–134.
[Link]
 Wang, M., Perfetti, C. A., & Liu, Y. (2005). Chinese–English biliteracy acquisition: cross-language
and writing system transfer. Cognition, 97(1), 67–88. doi:10.1016/[Link].2004.10.001
 Wang, M., Yang, C. & Chen, C. (2009). The contribution of phonology, orthography, and
morphology inChinese–English biliteracy acquisition. Applied Psycholinguistics, 30, 291–
314. doi:10.1017/S0142716409090122.
 Weber, A., & Cutler, A. (2004). Lexical competition in non-native spoken-word
recognition. Journal of Memory and Language, 50(1), 1–25.
[Link]
 Weber-Fox, C. M., & Neville, H. J. (1996). Maturational Constraints on Functional
Specializations for Language Processing: ERP and Behavioral Evidence in Bilingual
Speakers. Journal of cognitive neuroscience, 8(3), 231–256.
[Link]
 Williams, C. (1994) Arfarniad o Ddulliau Dysgu ac Addysgu yng Nghyd-destun
AddysgUwchradd Ddwyieithog, [An evaluation of teaching and learning methods in the
context of bilingual secondary education]. Unpublished Doctoral Thesis (University of
Wales, Bangor)
 Winograd, E., Cohen, C., & Barresi, J. (1976). Memory for concrete and abstract words in
bilingual speakers. Memory & Cognition, 4(3), 323–29. [Link]
 Zhou H, Rossi S and Chen B (2017) Effects of Working Memory Capacity and Tasks in
Processing L2 Complex Sentence: Evidence from Chinese-English Bilinguals. Front. Psychol.
8:595. doi: 10.3389/fpsyg.2017.00595