文言文寓言之篇章理解：關注寓意明顯與隱微的眼動證據

Comprehending Classical Chinese Fables: Eye-tracking Evidence on the Processing of Explicit or Implicit Morals

108國語文課綱要求學生具備閱讀白話文與文言文這兩種不同文體之能力，然閱讀文言文的困難，包含字詞解釋以及篇章主旨的理解，為此，本研究以真實的文言文寓言為材料，檢視成熟讀者是否能跨越文言文字詞解釋的困難，正確地擷取篇章主旨。寓言是文言文的入門教材，能讓學生連結古人的生活經驗，深化人與人之間的關係，以及待人處世的道理，寓言文章獨特地連結敘事為主的故事和說理來表達文章主旨的寓意，寓意的寫作技巧包含明顯的直接書寫或隱微的含蓄象徵。本研究採用眼球追蹤技術記錄大學生讀者閱讀這兩種類型寓言文章的歷程，對比在故事及寓意注意力的分布，也分析字詞解釋、文章整合及推論理解三種層次的閱讀理解表現。研究顯示，當閱讀「寓意明顯文章」時，寓意段落的重讀時間比故事段落短，可能因為寓意明顯，讓讀者直接達成內在連貫性的標準，獲取篇章主旨。但是，閱讀「寓意隱微文章」時，寓意段落的重讀時間比故事長，顯示讀者需要花較多時間推論隱微的寓意，才能達成連貫性的標準，獲取篇章主旨。讀者花時間關注的段落，顯示其理解表現較佳。本研究提供教師文言文教學的注意事項，建議引導學生注意文章段落的屬性，把各句意義與段落意涵形成連貫性的理解，亦可提供適性化的篇章理解教學。

Introduction

The curriculum guidelines for 12-year basic education require students to comprehend various literary genres in both vernacular and classical Chinese. However, students often fail to reach a deep understanding of classical Chinese. Previous studies have shown that although word decoding is the main difficulty in understanding classical Chinese, students can employ top-down reading strategies, such as causal and context inference, to extract the gist of the texts. Given these considerations, this study introduced classical Chinese fables as reading materials because fables are unique in their structures. A fable consists of two distinctive text types: the story and the moral. The story serves as the surface structure or vehicle, and the moral, presented in an expository form, serves as the main topic of the fable. Readers may infer the gist by constructing meaning from the story’s framework or extracting the main idea from the moral paragraph. By using fables, this study aimed to use eye-tracking technology to closely investigate how readers process the story and moral paragraphs when reading classical Chinese fables. Previous studies showed that the moral paragraph plays an important role in comprehending fables. Without the explicit moral paragraph, it is difficult to infer the gist of the fables. However, two written skills are used for the moral: one explicitly states the main concept, while the other expresses it more implicitly. Both are commonly seen in classical Chinese fables. It is worth investigating the reading process in these two types of fables. According to the construction-integration model, readers need to construct the propositions and integrate them with prior knowledge to reach the coherence of texts. The factors that drive readers to form a coherent mental representation of a text are closely related to their standard of coherence— the level of coherence they aim to maintain while reading. When coherence between paragraphs is strong, readers can meet their standard of coherence without additional processing strategies, passively accepting the content without increasing reading time. However, when coherence between paragraphs is weak, readers must actively process the text until coherence is established. Although this process demands greater cognitive resources and time for inference, it ultimately enhances comprehension. In other words, actively activating information and inference-making during reading affects readers’ memory representations. Given that reading is a dynamic, interactive process that involves cognitive processing to establish coherence between paragraphs, this study examines how readers create coherence between the story and moral paragraphs in classical Chinese fables to provide practical instructional recommendations for effectively teaching fables. Specifically, we investigate whether university students can extract the main idea of fables despite encountering challenging vocabulary. To further explore the cognitive mechanisms underlying this process, eye-tracking technology was used to examine whether students focus more on the story or moral paragraph while inferring the fable’s main idea.

Method

Sixty-two graduate and undergraduate students were recruited, except those who majored in classical Chinese. Four fables were used, two with explicit and two with implicit morals. Each fable consisted of two paragraphs: one presenting the story and the other the moral. In explicit moral fables, the moral paragraph clearly and directly states the gist of the story. In contrast, the implicit-moral fables employed metaphors in the moral paragraph, using indirect analogies to connect the moral to the story’s events. The post-comprehension test collected data to measure three types of questions: decoding, integrating, and inference. Decoding questions measured readers’ understanding of vocabulary and word meanings. Integration questions required readers to reconstruct events and recall details from the texts. Inference questions evaluated readers’ ability to derive the fables’ central messages and underlying meanings. Each text included 12 questions, with four items per question type, resulting in 48 test items. The reading ability test was derived from the College Entrance Examination Center. A total of 20 items were selected to measure participants’ decoding, integration, inference, and Chinese literary and cultural knowledge.

The EyeLink 1000 was used to collect eye movement. The area of interest in this study was sentence level. Each sentence was labeled as either a story or a moral. Two eye-movement measures were used: the first-pass reading time (FPRT) and rereading time (RRT). The first-pass reading time was the fixation duration when encountering the sentence before leaving, reflecting the early process, such as meaning construction and syntactic analysis. The rereading time was summed up with the fixation duration except for FPRT, reflecting late processes, such as integration. All the data was analyzed using linear mixed models.

Results

The reading comprehension results indicated that readers scored higher on integration and inference questions than on decoding questions. A significant interaction effect between the fable and question types revealed three key findings. First, readers performed similarly in both explicit- and implicit-moral fables for decoding questions. Second, readers scored higher in explicit-moral fables for integration questions than in implicit-moral fables. Third, readers scored higher in implicit-moral fables for inference questions than in explicit-moral fables. The results of FPRT showed that readers spent more FPRT reading explicit-moral fables or moral paragraphs than reading implicit-moral fables or story paragraphs. The results of RRT revealed a significant interaction effect. The simple main effects showed that when reading implicit-moral fables, readers spent more RRT on the moral paragraph than on the story paragraph, whereas this pattern was reversed for explicit-moral fables. Furthermore, readers spent more time on the story paragraph in explicit-moral fables than in implicit-moral fables, while they allocated more time to the moral paragraph in implicit-moral fables than in explicit-moral fables.

Discussion and Implication

The current study investigated how readers allocate their attention to the story and moral paragraphs when reading explicit and implicit moral fables and examine their reading comprehension. Results showed that readers obtained higher integration scores in explicit-moral fables and inference scores in implicit-moral fables. The eye-movement patterns can explain these results. Specifically, the results of the RRT showed that when reading explicit-moral fables, readers spent more RRT on the story than on the moral paragraph. These suggest that the explicit presentation of the moral enhances semantic overlap between paragraphs, creating strong coherence. As a result, readers can passively accept the moral without additional inference while dedicating more time to integrating the story’s contents to construct global coherence between the story and moral paragraphs. In contrast, when reading implicit-moral fables, readers spent more RRT on the moral paragraph than the story paragraph. This reflects the lower semantic overlap and weaker coherence between paragraphs, requiring readers to invest additional time in the moral paragraph to infer and integrate the symbolic metaphors with the story. That is, readers need to actively engage in inference to achieve coherence in understanding the moral. These findings suggest a distinction between the passive reception of meaning and the active pursuit of meaning across different paragraphs. Overall, this study contributes to our understanding of how readers comprehend classical Chinese fables and provides insights for educators and teachers. When students have a basic vocabulary level of classical Chinese, teachers can focus on enhancing reading comprehension. Genre types, paragraph features, and text coherence are critical to providing adaptive instruction.