老化對處理非熟悉隱喻的影響

Abstract

在此研究中，我們使用事件相關電位來探討正常老化是否影響熟悉度不同之隱喻的理解。隱喻是語言中常見的表達形式，文獻中普遍指出，理解隱喻所牽涉的認知歷程包含了抽象思考、推論、由既有知識中形塑新的概念……等。更甚者，隱喻實為我們理解並感知這個世界的重要基礎。然而，上述各項認知功能，已在文獻中顯示會受到正常老化的影響，老人的抽象思考、推論表現，往往不及於年輕人。關於老化是否影響隱喻理解，在文獻上尚未取得定論，一方面由於研究的數量少，另外也因為各實驗間的控制變項皆不同，尚難以相互比較。在年輕人的研究中，我們知道「熟悉度」是隱喻處理的研究中重要的變項，熟悉與非熟悉的隱喻將引發不同的處理歷程。因此，在我們的實驗當中，操弄了這個重要變項──隱喻的熟悉程度。另外，根據之前的研究指出，老化對語言處理的影響會因個人認知功能不同而有所差異，所以我們進一步測量幾個老人基本的認知功能，以利更細緻的分析。實驗中，受試者閱讀三種語意情境(condition)的句子：(1)表達具體動作的句子(literal sentence) (2)熟悉的隱喻 (3)非熟悉的隱喻。接著，受試者會再看到一個與句中動詞字面意相關或不相關的探子 (probe)，受試者必須做語意關聯度判斷作業，判斷探子是否與整句話語意相關。我們在受試者閱讀句子的同時記錄他們的腦波，發現老人在閱讀非熟悉隱喻的句子時引發的N400效果，比文獻中的年輕人慢且小 (450-550ms)，同時，老人未引發年輕人上看到的整合歷程，也就是LPC (late positive component)效果。進一步與個人認知功能做分析，發現老人在理解非熟悉隱喻時引發的N400效果，在刺激出現後的450-750毫秒，與受試者的語言流暢度測驗（verbal fluency）成績呈現正相關。以語言流暢度測驗成績的中位數作為標準，我們將受試者分為高語言流暢度與低語言流暢度兩大組別。實驗結果顯示，高語言流暢度的老年受試者在非熟悉隱喻的情境底下引發較持續的N400效果(450-750ms)；而熟悉的隱喻所引起的效果較為短暫(450-550ms)，在統計上屬邊緣顯著(marginally significant; p=.07)，且只集中在四顆電極上。相反地，低語言流暢度的老年受試者在理解隱喻時，並未引發任何腦波上的效果。儘管兩組受試者在理解的歷程或使用的神經機制(mechanism)上有差異，他們在隱喻改寫測驗上的表現並無差異。綜觀以上，我們的實驗結果顯示，隱喻的熟悉度確實影響老年人對隱喻的處理歷程，老年人整體在理解非熟悉隱喻時，引發的效果比年輕人慢且小，不過，擁有較好語言流暢度能力者，較能抵抗老化所帶來的影響，在處理非熟悉度的隱喻時，能尋求更多認知神經資源來幫助理解隱喻。而最後在隱喻改寫測驗上的高正確率，表示無論語言流暢度高低，所有老年人在充足的時間思考下，仍然能理解隱喻。由於在我們線上（online）實驗刺激出現後的800毫秒之內，未能觀察到LPC (Late positive component) 的效果，因此無法斷定老年人究竟在何時達到隱喻的整合性理解歷程，此部分尚待更多研究加以釐清。

Whether aging affects comprehension of metaphors has not been systematically investigated. This study assessed event-related potentials (ERPs) from 24 healthy older adults while they read literal, familiar metaphorical, and unfamiliar metaphorical usages of sentence-final action verbs, and judged, 800ms post verb onset, if a probe that was either literally related or unrelated to the verb was related to the sentential message. Unlike younger adults in the literature that showed a larger N400 response followed by a late positive component (LPC) to unfamiliar metaphors relative to the literal condition, older adults as a group showed a negativity to unfamiliar metaphors only in a later and more restricted time window (450-550ms) and no LPC effect. Further analysis suggested that individual variations of older adults’ brain responses could be accounted for by their verbal fluency scores. Older adults with higher verbal fluency elicited a sustained negativity to unfamiliar metaphors relative to the literal condition (450-750ms). However, older adults with lower verbal fluency scores did not show any reliable effects in their brain responses. Despite these differences in online measures, older adults in general were quite accurate (99.05% in familiar metaphors and 94.32% in unfamiliar metaphors) in a subsequent offline metaphor-paraphrasing task, indicating that, given time, older adults could successfully obtain the figurative readings for metaphors. These results suggest that metaphor familiarity modulates older adults’ online processing of metaphors in that age-related differences are mainly significant in unfamiliar metaphors. Older adults in general elicit smaller and later N400 effect and no LPC relative to younger adults in literature. In addition, individual cognitive differences affect older adults’ online processing of metaphors, with high-functioning older adults elicit a prolonged negativity in unfamiliar metaphors, suggesting higher functioning older adults are better able to recruit additional cognitive-neural resources to aid unfamiliar metaphor interpretations.