29.2596, Review: Applied Linguistics; Language Acquisition: LaMendola, Scott (2017)

[The LINGUIST List]

LINGUIST List: Vol-29-2596. Tue Jun 19 2018. ISSN: 1069 - 4875.

Subject: 29.2596, Review: Applied Linguistics; Language Acquisition: LaMendola, Scott (2017)

Moderators: linguist at linguistlist.org (Damir Cavar, Malgorzata E. Cavar)
Reviews: reviews at linguistlist.org (Helen Aristar-Dry, Robert Coté,
                                   Michael Czerniakowski)
Homepage: http://linguistlist.org

Please support the LL editors and operation with a donation at:
           http://funddrive.linguistlist.org/donate/

Editor for this issue: Jeremy Coburn <jecoburn at linguistlist.org>
================================================================


Date: Tue, 19 Jun 2018 11:27:52
From: Phaedra Royle [phaedra.royle at umontreal.ca]
Subject: Proceedings of the 41st annual Boston University Conference on Language Development

 
Discuss this message:
http://linguistlist.org/pubs/reviews/get-review.cfm?subid=36328957


Book announced at http://linguistlist.org/issues/28/28-2640.html

EDITOR: Maria  LaMendola
EDITOR: Jennifer  Scott
TITLE: Proceedings of the 41st annual Boston University Conference on Language Development
SERIES TITLE: BUCLD
PUBLISHER: Cascadilla Press
YEAR: 2017

REVIEWER: Phaedra Royle, Université de Montréal

SUMMARY

These two volumes contain 66 papers presented at the 41st Annual Boston
University Conference on Language Development (BUCLD) held November 4-6, 2016
a conference organized by professors and students in the Linguistics Program
at Boston University, and represents a wide range of theoretical and
methodological approaches to the study of language learning and acquisition. A
first this year: posters in addition to talks were accepted as publications in
the proceedings. Because of space limitations, only a third of the papers are
directly reviewed and little theoretical comments are made. Interested readers
are encouraged to read the original texts for a more in-depth understanding.

OVERVIEW

Linguists and psycholinguists, speech-language pathologists and others
interested in the development of phonology, morphology, syntax, semantics and
pragmatics, in monolingual, bilingual and language-disordered populations,
will find a wide variety of research articles in the BUCLD 41 proceedings. The
complete table of contents as well as pdf versions for all papers are
available on the Cascadilla Press web site at
http://www.cascadilla.com/bucld41toc.html. Pdf versions are a new and
interesting feature of this publication since 2016, before which it was
sometimes difficult to obtain individual copies of papers without buying the
book.

These can be viewed as well as “working papers” with highly innovative
approaches to new or old questions, and indicators of potential new research
that will eventually be published elsewhere. BUCLD has a diverse tradition in
topics, and the number of languages covered seems even larger than usual
(among which, American Sign Language, Arabic, Basque, Cantonese, Catalan,
Central Taurus Sign Language, Croatian, Cypriot Greek, Danish, Dutch, English,
Estonian, Finnish, French, German, Greek, Hebrew, Italian, Lithuanian,
Japanese, Korean, Maltese, Mandarin, Norwegian, Polish, Portuguese, Romanian,
Serbian, Spanish, and Turkish). Research using methodological approaches such
as eye tracking, ERPs and preferential-looking paradigms, as well as more
traditional corpus analyses and elicitation studies, with various populations
(bilinguals, deaf signers or oral language learners, children with ASD and
more) can be found.

Abed Ibrahim and Hamann (pp. 1-17) present work on the usefulness of sentence
and non-word repetition tasks for identifying Arabic-German and Turkish-German
bilingual children with and without specific language impairment (SLI). They
show that both types of tasks are useful for identifying monolingual
(German-speaking) as well as bilingual children with language impairment aged
5 to 9. The authors took care to vary both tasks on linguistic complexity
(while controlling for working-memory load), in order to identify
linguistic-specific domains of weakness in children with SLI.

Bentea and Durrleman (pp. 60-73) study the comprehension of relative clauses
in French using subject-verb number (mis-)match (e.g., _Montre moi le chat que
le/s chiens mord/ent_ [mɔʁ/d] ‘show me the cat that the dog/s bite/s’).
Results show that children have difficulty interpreting object relative
clauses if both potential antecedents (i.e., cat and dog) match in number.
However, audible agreement on the verb (which is not consistent in French) did
not improve comprehension abilities, especially in young (aged 5-8) children.
This is consistent with research on simpler structures showing that
French-speaking children show protracted mastery of agreement in production
and even comprehension in French (Pourquié & Royle, in preparation) and slow
emergence of attraction effects on verbs (after 7, Franck et al, 2004). 

In “What's a Foo? Toddlers Are Not Tolerant of Other Children's
Mispronunciations” (pp. 88-100), Bernier and White evaluate children’s
perception of child-like feature-mismatches or novel labels to known word
targets (e.g. _shoe_ [ʃu] -> _foo_ [fu], _voo_ [vu], _goo_ [gu] or _tibble_
[tɪbəl]). They find that children are quite intolerant of even 1-feature
errors (e.g. foo), although they show graded sensitivity to all types of
errors. However, children with experience interacting with other children who
spoke other languages than English also had a tendency to map 3-feature errors
to novel objects more often than children who did not have this
sociolinguistic experience. 

Boyce, Aravind and Hackl (pp. 101-113) undertake a corpus-based study of
lexical and syntactic effects on auxiliary selection in French. French has two
auxiliaries (_avoir_ ‘have’ and _être_ ‘be’), which are used with different
verbs based on argument/event structure, with the exception of reflexive
verbs, which must take _être_. The authors establish that transitive verbs are
appropriately produced with _avoir_ and that intransitives, while being
strongly mastered, show some errors not linked to cross-linguistic stability,
as predicted by Sorace (2000), but rather to age. The authors interpret the
data, after a closer look at their patterns, not to item-based learning but
rather to early error prone learning of exception types, and later perfect
mastery of new items falling into these difficult-to-learn categories.
Furthermore, auxiliary selection ceilinged at age 3, while reflexive clitic
use linked to auxiliary selection took much longer (age 6 and above), and
these two behaviours were not correlated, leaving open the question of how
mastery of reflexive clitics comes about in French. 

Brooks, Maouene, Sailor and Seiger-Gardner (pp. 114-127). Use latent semantic
analysis (LSA) and the continuous bag of words (CBOW) to study the semantic
networks of children with and without SLI. Children with SLI did not show
strong word finding difficulties in an association production task (usu. less
than 5%) but did show different patterns than controls, exhibiting more clang
(phonologically similar words), perseverant, and idiosyncratic responses than
controls. Their CBOW semantic distance scores were higher than those of
controls and their LSA similarity ones were lower. However, the authors
present and analyze qualitative data as percent response types rather than
frequencies, which would be more appropriate. Network community structures
were also more differentiated into word groups in children without, than those
with, SLI. 

Choi and Ionin (pp. 154-167) present a study on Korean and Mandarin adult
learners of the mass / count distinction in English and establish that
Atomicity is a salient cue for this feature when comparing processing of
ungrammatical non-atomic (e.g. _*sunshines_) vs. ungrammatical atomic ‘fake’
mass nouns (e.g., _*furnitures_). No effect of first-language was found on
their grammaticality-judgement and eye-tracking measures, suggesting that
atomicity is a semantic universal. 

Jensen, Slabakova and Westergaard (pp. 333-346) investigate Norwegian second
language (L2) learners’ ability to learn subject-verb agreement (morphology)
and unlearn verb second (syntax) in English. They show that the first process
is harder than the second, confirming the bottleneck hypothesis (Slabakova,
2008) that functional morphology is the most difficult domain of L2
acquisition. However, their paradigm confounds learning vs. unlearning and
domain (syntax and morphology). Anecdotally, I learned the same processes in
the other direction (English -> Danish) and found unlearning subject-verb
agreement much easier than verb second. Obviously the absence of verb
morphology played a role in my unlearning. This factor might also have
impacted results on this study (i.e. making learning harder in the other
direction).  It would be interesting to investigate more closely which factors
(e.g., unlearning vs. learning, obligatory in L1 vs. L2, morphology vs.
syntax) are driving results. 

Kapatsinski (pp. 357-372) uses singular-plural novel-word learning to
investigate how rule-like or statistical-like learning might obtain depending
on the type of input received. Results show that  participants might use both
rule and schema type learning and that cue-opacity or ambiguity is not a
factor driving pattern learning. Interestingly, this study does not attempt to
reproduce the particularity in natural languages to have more singular than
plural forms in the input, nor the fact that that different nouns can vary in
this respect (e.g., _eyes_ being more often in the plural than _hammer_). 

LaTourrette and Waxman (pp. 411-423) propose a conceptual account of
children's difficulties extending adjectives across basic-level kinds. They
compare children’s extensions of novel adjectives to ‘blobs of stuff’ and
‘pictures of things’ (i.e., kinds) using vague images that could represent
objects such as combs. Children in the ‘blob’ condition tended to extend the
novel adjective to a similar item more often than children in the ‘picture’
condition, although both groups got better over time, as well as in comparison
to a control group who were told the book was full of ‘pages’.  The authors
conclude that adjective extension is easier for children when they apply it to
non-objects, and that this procedure can produce high levels of accuracy in
comprehension. 

Ma, Gao, and Zhou (pp. 436-442) explore sensitivity to tones in young Mandarin
learners during novel word learning and recognition and test whether young
children aged three can distinguish tones 2, 3 and 4 on novel items. They do,
but have more difficulties with T3, especially in a sub-group of children,
showing that they can use tone to learn new lexical items but that T3 is more
difficult. The authors speculate this this might be related to the smaller
saliency/differences between T3 and T2 (which were paired in the task) or with
the presence of tone sandhi in Mandarin, where T3 becomes rising T2 in
combination with another T3. 

Martohardjono, Phillips, Madsen and Schwartz (pp. 452-465) attempt to study
heritage and second language (L2) Spanish-English speakers processing of
Spanish errors using ERPs.  Structures used are types 1), where a head noun in
the complex DP is ungrammatical in both English and Spanish, and 2), where a
syntactic complementizer is omitted, a structure grammatical in English though
not in Spanish. 

1) _¿Qué vecino contó Juan *el chisme que robó el carro anoche?_ 
      ‘What neighbour did Juan say *the rumour that robbed the car last
night?’  
2) _*¿Qué hermana confesó Inés ∅ (que) había comdio la tarta?_
      ‘What sister did Inés confess (that) had eaten the cake?’

Unfortunately, the experiment includes serious biases, especially in the
second condition, which render data interpretation difficult. As has been
argued by Steinhauer and Drury (2012), comparing two conditions with different
target words types is not ideal. In 2 this is _había_  in one condition, a
verb in sentences where _que_ is omitted, vs. _que_ the relativizer, in
sentences where it is not. This paradigm can cause differences in ERP waves —
the second being typically more negative (Marcinek et al, 2013) —, which can
bias the analysis. Furthermore, auditory cues to ungrammaticality could play a
role, as ungrammatical sentences were produced without controlling for
intonation or other low-level auditory cues. This is patently obvious in
Figure 1 (lower panel, p. 458) where we observe that the so-called N400 effect
of ungrammaticality of _Inés *había…_ starts BEFORE the presentation of
_había…_ that is on the baseline _Inés_, while the sentence is still
grammatical. The same problem can apply to condition 1) where the baseline
appears to show crossing lines (a ‘butterfly effect’) before the target
window, possibly enhancing the P600. The author’s interpretation of offline
judgements (which are generally inadequate in condition 2) as being incoherent
with the ERP data is thus flawed, i.e., it may be that no automatic processing
of the error was appropriately detected due to this experimental design. 

Marull (pp. 466-480) evaluates effects of language experience on L2
morphosyntactic integration and anticipation. Using visual-world and
grammaticality-judgment tasks, she attempts to provide counter-evidence for
the RAGE (Reduced Ability to Generate Expectations) theory of L2 acquisition
(Grüter & Rohde, 2013). However, she does not include participant group as a
between factor and therefore cannot justify her separate analyses for L1 and
L2 groups (Nieuwenhuis, et al, 2011). Furthermore, her design is unclear: are
all targets plural? What is the experimental/theoretical justification for the
two types of determiners being used? (The author states “Crosslinguistic
similarity” as being the reason and leaves it to the reader to figure it out).
The interpretation of results is most probably overstated, as statistical
analyses are not convincing (e.g., using ANOVAs for small numbers of
participants and extremely large standard deviations in reaction-time data,
which signal intra-subject variability that should be accounted for, Baayen et
al, 2008). 

Meir & Armon-Lotem (pp. 495-508) present a study comparing bilingual children
with SLI to bilingual controls of the same age (6 years old), but matched with
children with SLI on their weaker language, thus allowing for both age and
language matching. They use the common pairing of non-word repetition (NWR)
and sentence repetition (SR) tasks to evaluate three groups. SLI bilinguals
speaking Russian and Hebrew and two controls groups of either Hebrew- or
Russian-dominant bilinguals, compared on their weaker language to children
with SLI. They find that error patterns are quite different in children with
and without LI. Consonant cluster and syllable reduction are more common in
children with LI than those without LI, and many different syntactic patterns
are revealed. Unfortunately, it is hard to clearly grasp error patterns, as
they are presented as percentages, which is inappropriate for frequency data.
The statistics also seem inadequate. 

Petroj (pp. 532-545) investigates article distribution in American Sign
Language (ASL) concurrent with whispered English. Studying participants who
are CODAs (who simultaneously learned ASL and English), she observes that the
grammar of ASL influences article use in whispered speech when blended with
ASL, which does not use articles. She shows that prosodic preferences override
syntactic constraints in English for determiner production in whispered speech
in these children. 

Smeets (pp. 588-601) investigates ultimate attainment of L2 object movement in
Dutch, focusing on the syntax-discourse interface. She observes that advanced
English- and German- L2 learners of Dutch can attain L1 levels of sensitivity
to object vs. wide focus constraints on object movement in non-canonical
structures such as _Appels heeft Bas gekocht_ ‘Apples Bas bought’.
Furthermore, in scrambled sentences such as _dat Hans de secretaresse
binnenkort zal ontslaan_ ‘that Hans will fire the secretary soon’ with the
object given (Is there any news about the secretary?) versus wide focus (what
did Wouter say?), native and German-Dutch L2 speakers show preference for the
scrambled word order, but only a subset of English-Dutch L2 speakers do (4 of
15) . Finally she tested interpretations of scrambling with indefinite
structures, which lead to specific readings, e.g. _De entertainer heft een
paar liedjes regelmatig gezongen_ ‘The entertainer has regularly sung a few
(specific) songs’. In this case, 10 of 15 English-Dutch L2 learners reach
native-like readings (80% correct or higher). These results show on the one
hand that L2 speakers can attain native-like syntactic behaviour but, on the
other, that some aspects of the syntax-discourse interface may be more
difficult. These cannot be attributed to a general bilingual disadvantage, as
German-Dutch L2 speakers performed as natives. 

Terzi, Zafeiri, Marinis and Francis study the use of object clitics in
narratives of high-functioning Greek-speaking children with Autism. Their
results show practically no differences between 20 children with Autism (mean
age 6:11) and 20 neurotypical children matched on their receptive vocabulary
and verbal intelligence (mean age 6;7). One pattern that distinguishes the
groups is a preference for subject clitics in children with Autism. The
statistical methods seem inappropriate however, as t-tests are used for
frequency data. 

Tieu and Križ (pp. 651-664) investigate exhaustivity in French clefts — _C’est
le ballon qui est rouge_ ‘It’s the balloon that is red’ — and homogeneity in
plural definite descriptions — _Les ballons sont rouges_ ‘The balloons are
red’. In infelicitous contexts (where for example, another object was also red
in the cleft condition, or one of the balloons was another colour in the
plural definite condition) 3-to-4-year-old children accepted them as true.
5-to-6-year-old children accepted clefts more than definite descriptions, but
even adults accepted infelicitous clefts (∼ 50%), while rejecting infelicitous
definite descriptions. The authors conclude that French children begin by
interpreting clefts non-exhaustively and definite plurals existentially, and
that a homogeneous interpretation of definite plurals emerges before
exhaustivity (at ages where English-and German-speaking children have acquired
exhaustivity, at least partially). However, the fact that adults also allow
non-exhaustive readings of the cleft structure argues for a second
non-exhaustive reading for clefts in French (as controls sentences with ‘only’
were correctly interpreted as exhaustive). 

Veenstra, Antoniou, Katsos and Kissine (pp. 706-717) study attraction effects
and executive control in monolingual Dutch and sequential bilingual
French-Dutch children aged 11. They were asked to describe pictures where the
head noun matched, or not, in number with the local noun (e.g., _De cirkel/s
naast de driehoek/en is/zijn blau_ ‘The circle/s next to the triangle/s is/are
blue’). No bilingual advantage on executive control tasks was found. In fact,
the monolingual group showed better scores on the Corsi block forward subtest.
Both monolinguals and bilinguals showed attraction effects, i.e.
verb-agreement errors triggered by the intervening noun when mismatching in
number with the head noun. Finally, backward digit span and backward Corsi
block scores negatively predicted agreement errors (over both groups), that is
better scores on these measures were correlated with fewer attraction errors.

White, Goad, Su, Smeets, Mortazavinia, Garcia and Brambati Guzzo (pp. 744-752)
present preliminary data on the effects of prosody in pronoun interpretation
in Italian in sentences of the type _Lorenzo ha scritto a Roberto (#) quando
Ø/lui si è trasferito a Torino_ ‘Lorenzo wrote to Roberto (#) when (he) moved
to Turin.’ They show that 1) L2- and L1 speakers of Italian have different
interpretations of overt and null subjects as referring to a preceding subject
(Lorenzo) or object (Roberto) in the sentence, or even some other discourse
referent, 2) that the presence of a pause before the null pronoun can promote
an object referent interpretation, and 3) that the presence of contrastive
stress on overt pronouns appears to interact with (no-)pause effects on their
interpretation. 

Yatsushiro, Sauerland and Alexiadou (pp. 753-765) present an incredible amount
of cross-linguistic data (18 languages form Finnic, Semitic, Greek, Baltic,
Slavic, Germanic and Romance families) in support of the unmarkedness of
plural. Interestingly they include French which does not have overt plural
marking on the noun (i.e., the –s in only written but not usually pronounced)
so it is unclear whether the authors wanted to test plural marking on nouns
only or on other structures (such as French determiners which do bear number
information on the vowel). In any case the cross-linguistic data are not
really discussed, while the authors focus rather on two German experiments for
singular markedness interpretation, which they argue correlate without
statistical support or any convincing illustration (i.e., a scatter plot that
distributes the data widely across the figure) or clear discussion of their
impact.

EVALUATION

The papers in these volumes are directed at researchers and graduate students
in language acquisition and processing. The methodological approaches and
theoretical assumptions are quite varied in the papers, thus making the papers
extremely varied in their scope and coherence with the rest of the volume.
Because of their short length, some background might be necessary to be able
to appreciate their contents. They are all written as ‘stand-alone’ papers,
and are thus of value as short scientific papers, especially in a research
seminar type setting. However, undergraduates could also benefit from these
papers, especially if put in the context of other readings with theoretical
and methodological grounding for them. I often use BUCLD proceedings as short
discussion papers in seminars or even exam papers to help students develop
their ability to comment on scientific articles in domains or populations that
are relevant to their work (in my case, future SLPs). 

Having read a number of previous BUCLD proceedings, I can state that the
quality of the papers has generally been maintained or has risen over time, at
least in terms of presentation, however, as usual, there is high variability
in the quality of data analysis or even scientific discussion. This is mainly
due to a traditional lack of editorial oversight. As I have mentioned before,
the most appealing quality of these papers is that they report on very recent
research, which is often not yet available elsewhere and can be cutting-edge.
This can also have some downsides: Chapter quality is quite variable, with
some research still ongoing, some methodologies questionable, or theoretical
assumptions not explicit. In particular, some statistical analyses are
appropriately developed and presented while others are inadequate (for
example, the recent move to mixed-models in R (REF) is quite evident, but
model presentation often lacks basic details such as AIC information, while
other researchers perform t-tests on frequency data, which is inappropriate).
Some analyses (whether on response, reaction-time or even ERP data) do not
check for group interaction of effects, before breaking down analyses into
different partitions (Nieuwenhuis et al., 2011). Some papers don't even bother
to check for group effects, and some authors apparently do not spellcheck
their papers. I have made this comment before about the BUCLD proceedings
(https://linguistlist.org/issues/24/24-55.html). I think it is high time that
an editorial review is performed on papers before they are published. This
practice has been established in much smaller student-run conferences
elsewhere, e.g., VOCUM (http://vocum.ca/en/), and allows for some level of
peer review. Because of the new policy allowing all presenters to submit
papers, I fear that flawed papers might become more common in this
publication. However, I must state that in general, the quality of the papers
is quite high and they can present insightful and inspirational research. 

REFERENCES

Baayen, Robert Harald, Doug J. Davidson & Douglas M. Bates. 2008.
Mixed-Effects Modeling with Crossed Random Effects for Subjects and Items.
Journal of Memory and Language 59. 390-412.

Franck, Julie, Stephany Cronel-Ohayon, Laurence Chillier, Ulrich H.
Frauenfelder, Cornelia Hamann, Luigi Rizzi & Pascal Zesiger. 2004. Normal and
Pathological Development of Subject-Verb Agreement in Speech Production: A
Study on French Children. Journal of Neurolinguistics 17(2-3). 147-80.

Grüter, Theres & Hannah Rohde. L2 Processing Is Affected by Rage: Evidence
from Reference Resolution. In 12th conference on Generative Approaches to
Second Language Acquisition (GASLA). University of Florida, FL, 2013.

Marcinek, Bradley T., Karsten Steinhauer, Phaedra Royle & John E. Drury. 2013.
Syntactic Violations for Content Verses Function Words in Reading: Erp
Evidence. In Society for the Neurobiology of Language. San Diego, CA.
http://bradmarcinek.net/files/Marcinek_Steinhauer_Royle_Drury_SNL2013.pdf

Nieuwenhuis, Sander, Birte U Forstmann & Eric-Jan Wagenmakers. 2011. Erroneous
Analyses of Interactions in Neuroscience: A Problem of Significance. Nature
Neuroscience 14. 1105-07.

Pourquié, Marie & Phaedra Royle. (in preparation). Argument Structure and Verb
Inflection in French Sli. 

Slabakova, Roumyana. (2008). Meaning in the Second Language. Berlin: Mouton de
Gruyter.

Sorace, Antonella. (2000). Gradients in Auxiliary Selection with Intransitive
Verbs. Language 76(4). 859-890. 

Steinhauer, Karsten & John E. Drury. (2012). On the early left-anterior
negativity (ELAN) in syntax studies. Brain and Language 120(2). 135-162.


ABOUT THE REVIEWER

Phaedra Royle holds a Ph.D and is a full professor at the School of Speech
Language Pathology and Audiology at the Université de Montréal, and is a
member of the Centre for Research on Brain, Language and Music (CRBLM). Her
interests lie in psycholinguistics, neurolinguistics, language disorders,
language acquisition, morphology, morpho-phonology and morpho-syntax, and
processing of complex noun-phrases in French populations with and without
learning challenges (SLI, Cochlear implants, Bilingualism, Ageing).





------------------------------------------------------------------------------

*****************    LINGUIST List Support    *****************
Please support the LL editors and operation with a donation at:

              The IU Foundation Crowd Funding site:
       https://iufoundation.fundly.com/the-linguist-list

               The LINGUIST List FundDrive Page:
            http://funddrive.linguistlist.org/donate/
 


----------------------------------------------------------
LINGUIST List: Vol-29-2596	
----------------------------------------------------------