Beyond accuracy: Fluency and complexity in the spoken output of bilingual and monolingual preadolescents

Magdalena Grose-Hodge*, Ewa Dabrowska, Dagmar Divjak

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

26 Downloads (Pure)


The language of Heritage Speakers (HS), or of early bilinguals of a minority language, is often seen as incomplete or less developed than that of Monolingual Speakers (MS). This study investigates whether 7- to 9-year-old English/Polish HS can be distinguished from MS in terms of linguistic skills when complexity and fluency are focal rather than accuracy. Data from 78 participants shows no significant differences between HS and MS in fluency on an overall measure in either of the languages, although HS produce more fillers and repetitions. On complexity measures, the results for English were similar across groups for Mean Length of T-Unit, but there was a statistically significant difference in Polish, with HS achieving higher values. On a more specific measure of syntactic complexity (Subordination Index), HS achieved higher scores in both languages. There were no significant differences for Lexical Diversity. Corresponding measures were positively correlated across languages, suggesting that the L1 does not impede L2 achievement. Overall, we observed substantial overlap between the groups, with the vast majority of HS falling within the MS norms and the MS falling within the HS norms. This emphasises the need to move away from the deficit approach towards HS.
Original languageEnglish
Pages (from-to)328-356
Number of pages29
JournalJournal of Monolingual and Bilingual Speech
Issue number3
Publication statusPublished - 30 Jan 2024


  • heritage language acquisition (HLA)
  • syntactic complexity
  • fluency
  • bilingualism
  • incomplete acquisition
  • deficit approach


Dive into the research topics of 'Beyond accuracy: Fluency and complexity in the spoken output of bilingual and monolingual preadolescents'. Together they form a unique fingerprint.

Cite this