Project team: Catherine Wang, Kate Nation
Â
Lexical ambiguity is ubiquitous; it enriches our language and presents challenges for readers. According to the good-enough theory, readers process language partially rather than in full detail. Hence, lexical ambiguity might bias how individuals process and interpret language. While the influence of lexical ambiguity has been extensively studied in English, it is also substantial in Chinese. Unlike English sentences, there are no spaces to indicate word boundaries in Chinese written sentences. Instead, individuals need to segment characters into words by themselves. One form of ambiguous segmentation is the three-character Chinese string (ABC), comprising two two-character words (AB and BC), and the middle character(B) is always a part of both words. During sentence processing, readers need to exclusively assign each character to one word and decide whether the middle character(B) should be segmented with character A or C, which could introduce interpretation errors.
This experiment investigated how readers select and activate the contextually relevant meanings of such ambiguous input and what factors influence this process. Specifically, it studied the effect of contextual diversity (CD) and the position of words on how readers read and interpret ambiguous sentences. Ambiguous strings are categorised into two conditions based on SUBTLEX-CH-WD corpus, "highCD-lowCD" versus "lowCD-highCD", depending on the relative contextual diversity of the two possible word segmentations. Data will be collected via the online platform Gorilla (https://gorilla.sc). Participants will read the text silently and then make TRUE/FALSE judgements of statements that follow each text. These statements refer to either the AB or the BC interpretation of the three-character string, and we refer to the statement as 'congruent' if correct and 'incongruent' if incorrect, based on the context information. We measured global reading times of ambiguous sentences, accuracy and global reading times of comprehension questions to illustrate how lexical ambiguity influences readers' natural processing and interpretation of sentences.
Our experiment showed that CD and the position of words influenced individuals' processing and interpretation of sentences, demonstrating a similar good-enough reading effect in Chinese. Thus, investigating lexical ambiguity in Chinese will produce new data to extend models of reading to non-English contexts and produce more comprehensive models of reading in other language families.
