Not this, but that.' Exploring disambiguation in the context of multilingual word learning
View/ Open
Date
31/07/2021Item status
Restricted AccessEmbargo end date
31/07/2022Author
Repnik, Katharina Maria
Metadata
Abstract
Growing up and living with more than one language modulates new language learning
in a variety of ways. Similarities and differences to monolinguals with regards to
languages learning in all its aspects were found. These range from phonology to
pragmatics. However, non-linguistic domains, such as cognition, can also be affected
by the multilingual experience.
The acquisition of words is an indispensable step in learning a language, be
that one’s first or subsequent languages. To learn a word, its word form must be
correctly mapped onto its meaning and then remembered. One ‘strategy’ children and
adults implicitly employ to map a new label to a new item or concept is referred to as
disambiguation. This mental process is analogous to the common process-of-elimination and based on the Disjunctive Syllogism. The idea is that a new label refers
to a novel object in the presence of familiar items creating the support for one-to-one
mappings between a word and its referent. Attaching a new label to a new label-unknown object rather than a familiar label-known object expedites vocabulary
growth, especially in children. Although this mental process can be found in adults
too, it was predominantly studied in the context of word learning in children, and more
recently multilingual children.
The interest in researching disambiguation in multilingual contexts arose as
researchers began to wonder if people, who grew up and were exposed to more than
one language, would exhibit the so-called disambiguation effect at all given that they
know multiple labels for the same objects. The formulation of many-to-one mappings
stands in opposition to one-to-one mappings mentioned above. However, does this
mean that multilingual children and adults never disambiguate? The answer to this
question cannot be binary, considering the myriad of direct and indirect effects
multilingualism has on word learning.
In this thesis, I explore the use of disambiguation in various multilingual contexts:
1. Disambiguation studies on children are usually conducted under the scope of
word learning, which consists of mapping and retention of words; however, only very few studies explicitly explore the relationship between the mapping
of a new word and the ability to retain it. In this first study, I explored whether
and how monolingual and multilingual children’s use of disambiguation as a
fast mapping constraint boosted their ability to retain this new mapping within
the same experiment. For this, I designed and adapted a looking-while-listening eye-tracking paradigm to allow for data collection within a young
population (aged 18 to 30 months old). I found that children from monolingual
and multilingual backgrounds are able to use disambiguation as mapping
strategy; however, they differ with regards to how disambiguation modulated
their ability to retain the new word. Whilst the use of disambiguation positively
affected monolingual children’s retention performance, children from
multilingual backgrounds did not display this boosting effect. Thus, it can be
said that disambiguation starts out as default mapping strategy that in some,
but not other, cases develops into a more consolidated learning/retention
strategy.
2. In a second study, I examined whether disambiguation as a mapping strategy
is prevalent in multilingual adults and how their multilingual experience may
modulate the processing speed of this process-of-elimination. Multilingualism
does affect a person’s learning not only directly but also indirectly. An
accumulating amount of research is currently being conducted on the mutual
effects of multilingualism and executive functioning with findings addressing
cognitive domains such as working memory and cognitive control. As these
findings allude to differences in cognitive processing, any task requiring the
contribution of these skills, such as the one in this study, one must control for
these variations. In this study, I explored the individual contributions of
multilingualism and executive functioning to a fast mapping and retention task.
Language experience variables strongly modulated the accuracy and
processing times of disambiguation and retention conditions, whereby speed-accuracy trade-offs were found. Furthermore, working memory and cognitive
control impacted the accuracy and reaction times. Those with better scores on
the working memory task also performed faster. Results highlight the requirement for regarding multilingualism as a multivariate spectrum in future
research rather than in dichotomous categories.
3. In a third study, I extended the context of disambiguation by not only
examining its use on mappings between words and objects but also on
mappings between factual information and objects. Factual information about
items in this study are, for instance, “this is the one I keep in the living room”.
Studies with young children have shown their tendency to transfer the use of
disambiguation onto other domains of language learning, such as assigning
new factual information to an unfamiliar object rather than a familiar name-known object. However, it is self-evident that adults may not show this transfer
of disambiguation due to their ability to draw from other forms of knowledge,
such as their cognitive, semantic, and pragmatic abilities, and life experience
in general. In this study, I examined how multilinguals adults’ linguistic
experience impacts on the use of disambiguation in factual contexts by looking
at eye-gaze data. Furthermore, I probed participants’ underlying decision-making processes by questioning their choices after specific trials. The results
showed that adults do not extend disambiguation to contexts other than labels,
but that language background did modulate those instances in which
disambiguation took place based on contrast.
In conclusion, it can be said that disambiguation is a sound strategy for making fast
decisions about labels and how they attach to meaning. However, the investigations in
this thesis also show that the context in which disambiguation is to be used is
paramount to its success or someone’s implicit decision to rely upon it. Lastly, this
thesis and its findings contribute to the modern view that multilingualism is a
continuum and should be researched and regarded from a holistic angle.