Differential brain mechanisms during reading human vs. machine translated fiction and news texts

Fa-Hsuan Lin,Hsin-Ju Lee,Jyh-Neng Yeh,Wen-Jui Kuo,Yun-Fei Liu,Claire H. C. Chang,Iiro P. Jaaskelainen

doi:10.1038/s41598-019-49632-w

Abstract

Few neuroimaigng studies on reading comprehension have been conducted under natural reading settings. In this study, we showed texts presented in a natural way during functional MRI (fMRI) measurements to reveal brain areas sensitive to reading comprehension. Specifically, this paradigm independently manipulated two holistic features of article style: text genre and translation style, a qualitative index of how typical word choices and arrangements are made in daily use of the language. Specifically, articles from The New York Times (news) and Reader’s Digest (fiction) translated from English to Mandarin Chinese either by human experts or machine (Google Translate) were used to investigate the correlation of brain activity across participants during article reading. We found that bi-hemispheric visual cortex, precuneus, and occipito-parietal junction show significantly correlated hemodynamics across participants regardless of translation style and article genre. Compared to machine translation, reading human expert translation elicited more reliable fMRI signals across participants at precuneus, potentially because narrative representations and contents can be coherently presented over tens of seconds. We also found significantly stronger inter-subject correlated fMRI signals at temporal poles and fusiform gyri in fiction reading than in news reading. This may be attributed to more stable empathy processing across participants in fiction reading. The degree of stability of brain responses across subjects at extra-linguistic areas was found correlated with subjective rating on the text fluency. The functional connectivity between these areas was modulated by text genre and translation style. Taken together, our imaging results suggested stable and selective neural substrates associated with comprehending holistic features of written narratives.

Highlights

Reading is to understand the meaning encoded in an article composed of a hierarchical structure of paragraphs, sentences, and words
We used the BOLD signal time courses recorded when participants read articles in a naturalistic manner for a few minutes to study the neural substrates underpinning reading comprehension. We used this approach to study how BOLD signals are modulated by translation style, which was tested by presenting articles translated from English to Mandarin Chinese by either human or machine, and article genre, which included first-person narrative fictions and news reports
Our analyses showed that bi-hemispheric visual cortex, precuneus, and occipito-parietal junction show significantly correlated BOLD dynamics across participants regardless of translation style and text genre

Summary

Introduction

Reading is to understand the meaning encoded in an article composed of a hierarchical structure of paragraphs, sentences, and words. Brain areas showing significantly different ISC values between reading of different literary (F vs N) and translation style (H vs M) were www.nature.com/scientificreports first chosen as seed regions-of-interest (ROI’s).

Results

Conclusion