Automated Measures of Syntactic Complexity in Natural Speech Production: Older and Younger Adults as a Case Study

Galit Agmon, Sameer Pradhan, Sharon Ash, Naomi Nevler, Mark Liberman, Murray Grossman, Sunghye Cho

Research output: Contribution to journalArticlepeer-review

Abstract

Purpose: Multiple methods have been suggested for quantifying syntactic coplexity in speech. We compared eight automated syntactic complexity metrto determine which best captured verified syntactic differences between and young adults. Method: We used natural speech samples produced in a picture descriptask by younger (n = 76, ages 18–22 years) and older (n = 36, ages 53–89 healthy participants, manually transcribed and segmented into sentences. Wmanually verified that older participants produced fewer complex structures. developed a metric of syntactic complexity using automatically extracted synttic structures as features in a multidimensional metric. We compared our mto seven other metrics: Yngve score, Frazier score, Frazier–Roark score, develmental level, syntactic frequency, mean dependency distance, and sentenlength. We examined the success of each metric in identifying the age gusing logistic regression models. We repeated the analysis with automatic trscription and segmentation using an automatic speech recognition (ASR) systeResults: Our multidimensional metric was successful in predicting age gro(area under the curve [AUC] = 0.87), and it performed better than the other rics. High AUCs were also achieved by the Yngve score (0.84) and sentlength (0.84). However, in a fully automated pipeline with ASR, the performaof these two metrics dropped (to 0.73 and 0.46, respectively), while the pemance of the multidimensional metric remained relatively high (0.81). Conclusions: Syntactic complexity in spontaneous speech can be quantified directly assessing syntactic structures and considering them in a multivariamanner. It can be derived automatically, saving considerable time and efcompared to manually analyzing large-scale corpora, while maintaining hiface validity and robustness.

Original languageEnglish
Pages (from-to)545-561
Number of pages17
JournalJournal of Speech, Language, and Hearing Research
Volume67
Issue number2
DOIs
StatePublished - 12 Feb 2024
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 2024 American Speech-Language-Hearing Association.

Fingerprint

Dive into the research topics of 'Automated Measures of Syntactic Complexity in Natural Speech Production: Older and Younger Adults as a Case Study'. Together they form a unique fingerprint.

Cite this