
The interaction between language and working memory: a systematic review of fMRI studies in the past two decades
Author(s) -
Zoha Deldar,
Carlos GeversMontoro,
Ali Khatibi,
Ladan Ghazi Saidi
Publication year - 2020
Publication title -
aims neuroscience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.257
H-Index - 12
eISSN - 2373-7972
pISSN - 2373-8006
DOI - 10.3934/neuroscience.2021001
Subject(s) - working memory , functional magnetic resonance imaging , psychology , cognition , baddeley's model of working memory , cognitive psychology , short term memory , neuroscience
Language processing involves other cognitive domains, including Working Memory (WM). Much detail about the neural correlates of language and WM interaction remains unclear. This review summarizes the evidence for the interaction between WM and language obtained via functional Magnetic Resonance Imaging (fMRI) in the past two decades. The search was limited to PubMed, Google Scholar, Science direct and Neurosynth for working memory, language, fMRI, neuroimaging, cognition, attention, network, connectome keywords. The exclusion criteria consisted of studies including children, older adults, bilingual or multilingual population, clinical cases, music, sign language, speech, motor processing, review papers, meta-analyses, electroencephalography/event-related potential, and positron emission tomography. A total of 20 articles were included and discussed in four categories: language comprehension, language production, syntax, and networks. Studies on neural correlates of WM and language interaction are rare. Language tasks that involve WM activate common neural systems. Activated areas can be associated with cognitive concepts proposed by Baddeley and Hitch (1974), including the phonological loop of WM (mainly Broca and Wernicke's areas), other prefrontal cortex and right hemispheric regions linked to the visuospatial sketchpad. There is a clear, dynamic interaction between language and WM, reflected in the involvement of subcortical structures, particularly the basal ganglia (caudate), and of widespread right hemispheric regions. WM involvement is levered by cognitive demand in response to task complexity. High WM capacity readers draw upon buffer memory systems in midline cortical areas to decrease the WM demands for efficiency. Different dynamic networks are involved in WM and language interaction in response to the task in hand for an ultimate brain function efficiency, modulated by language modality and attention.