EEG Correlates and Behavioral Markers of Non-adjacent Dependency Learning During Passive Listening in Russian-speaking Adults and Children

In the process of language acquisition children encounter the need not only to understand and remember new words, but also to generalize various rules that regulate words’ ability to occur together, i.e., grammatical rules. There are rules associated with adjacent words (e.g., associate must be followed by with ) and those establishing dependencies between distant elements (e.g., he and she require verbs in Present Simple to end with -s : she know s ). It is believed that children can acquire non-adjacent dependencies (NAD) during passive listening until they are 4 years old, while adult speakers cannot recognize the presence of these rules unless they are presented with an explicit task to find some patterns in the experimental audio sequence (so-called active listening). 

The aim of our study is to explore electrophysiological correlates of NAD acquisition in Russian-speaking adults and 2-to-4-year-old children using electroencephalography (EEG).

In our EEG experiment adults and children are listening to the phrases of unfamiliar to them Lithuanian language, and these phrases follow two NAD rules that are similar to pronoun-verb agreement in Russian (e.g., tu ker-ta — ‘you work’, jie ker-ti   — ‘he works’). The second syllable of trisyllabic sequences (phrases) is always one of 12 verb roots, while the first syllable tu (pronoun) requires the third syllable (the verb inflection) to be ta , and the pronoun jie must agree with verb via the inflection ti . However, there are ‘correct’ and ‘deviant’ phrases in the experimental sequence, the former follow these grammatical rules, while the latter violate them (e.g., tu ker-ti — ‘you works’, jie ker-ta  — ‘he work’). Participants also listen to grammatically ‘correct’ phrases with the last syllable of higher than regular pitch, so-called ‘phonetic deviants’. Adult participants are tested in two paradigms, one of which has a fixed number of ‘correct’ and ‘deviant’ phrases (six ‘correct’ ones and one ‘deviant’, either grammatical or phonetic) and with pseudorandom order of phrase presentation (with 80% of ‘correct’ phrases and 10% of each ‘deviant’ type). Children, on the other hand, are tested in the second paradigm only. Therefore, firstly, by comparing electrophysiological activity of the brain in response to ‘correct’ and ‘deviant’ phrases, we will see whether participants have acquired NAD rules, and, secondly, what EEG-correlates are associated with reaction to NAD rule violations.

Based on previous research, we hypothesize that adult Russian speakers will not demonstrate any electrophysiological correlates of NAD rules acquisition, while children will probably succeed in this task. When we compare participants’ reactions to ‘correct’ and ‘deviant’ verb inflections, we will be able to identify EEG-correlates related to NAD rules acquisition. Moreover, we expect to see the mismatch negativity (changes in brain activity with negative polarity as a reaction to a deviant stimulus ~100 ms after its presentation) in response to ‘phonetic deviants’.