| 摘要: | 在過去以語言為刺激材料的短期記憶研究裡，語音被認為扮演一個重要的角色，而只有當語音沒有辦法幫助受試者記憶的時候，才會觀察到字形對於短期記憶的影響。相較於拼音文字，中文字的字音及字形之間的對應較不直接。利用此中文字的特性，可能較容易展現出字形對於短期記憶的影響。本研究設計四個行為實驗，企圖展現字形對於短期記憶的影響，並且更進一步地想區分具有語言成份的字形及沒有語言成份的視覺訊息對於語言使用者短期記憶的不同影響。實驗一與實驗二獨立操弄中文字的語音與字形相似度，以探討當受試者可以利用語音記住刺激材料的時候，是否仍可觀察到字形對於短期記憶的影響。實驗結果展現出語音相似度的影響，說明了語音對於短期記憶的重要性。更重要的是，我們也發現當語音可以幫助受試者記刺激材料時，仍看到字形相似度的影響。當受試者在記憶實驗一中形似音似的刺激材料時，由於作業的高難度，使他們發展出只專注於中文字左邊偏旁而忽略其一致的右邊偏旁的策略，以獲得較好的記憶表現。當我們在實驗二中重新選擇刺激材料，使這樣的記憶策略不再有效之後，受試者的記憶表現顯示：不管在語音可否幫助受試者記憶的情況下，字形對於短期記憶都會造成影響。由實驗一與實驗二的結果，得知即使語音在短期記憶中扮演一個主要的角色，字形亦有其貢獻。實驗三利用操弄同樣的偏旁但出現在文字中相同以及不同的位置，以探討偏旁的位置是否對於展現出字形相似度在短期記憶中的效果有影響。其結果顯示：只有當同樣的偏旁出現在相同的位置時，受試者的短期記憶才有較差的表現，表示偏旁的位置對於展現字形相似度的效果是重要的。實驗四試著探討不同於字形、純粹是知覺層次的視覺相似度是否亦會對於中文使用者的記憶產生混淆而造成再認表現較差的情形。實驗結果顯示：視覺相似度並沒有對中文受試者的短期記憶造成影響。此外，我們亦請一組中文受試者以及一組不認識中文字的外籍受試者針對實驗三與實驗四的中文刺激材料做七點量表的視覺相似度評分。這兩組受試者對於中文字的視覺相似度的評分是相似的。當將視覺相似度的評分結果與中文受試者的短期記憶表現做比較後，我們更進一步確認：中文字的視覺相似度與記憶這些文字的表現並沒有直接相對應的關係。總結本研究，我們觀察到字形的表徵對於短期記憶的影響。此結果支持一個從心理語言學的角度出發、和長期記憶中的語言表徵有密切關係的短期記憶系統，亦即在以語言為刺激材料的短期記憶中，不僅只有語音表徵，文字的其他訊息（例如字形、字義）也同時對短期記憶有貢獻。According to previous theories, phonology plays an important role in verbal short-term memory (VSTM). Meanwhile, orthography plays a role only when phonology is not available or less useful. Comparing to alphabetic languages, the correspondence between phonology and orthography is relatively opaque in Chinese. Therefore, the contribution of orthography to VSTM may be more likely to be revealed from Chinese characters. In the thesis, four experiments were conducted. The first two experiments manipulated the phonological and orthographic similarity among Chinese characters independently and attempted to determine whether the effect of orthography would be observed, even when the phonological information is helpful for the participants to maintain the memory items. The results showed the robust phonological similarity effect in the two experiments, which indicated that phonology plays a dominant role in VSTM. More important, the orthographic similarity effect was consistently observed when phonology of the stimuli was helpful to retain verbal materials. In addition, once the possible strategy of focusing on one of the radicals within a Chinese character while ignoring the other radical in Experiment 1 was prevented, Experiment 2 demonstrated the orthographic contribution to VSTM in both the phonologically similar and dissimilar characters. The findings of first two experiments suggested that although phonology is a dominant component for VSTM, orthography is also crucial to retain verbal stimuli. Experiment 3 manipulated the identity and positional information of the radicals within Chinese characters and investigated its impacts on VSTM. The results showed that both the identity and the positional information of a radical are important to determine the orthographic similarity among items in VSTM. Finally, Experiment 4 was designed to examine whether the non-linguistic visual similarity of Chinese characters would affect VSTM just like the linguistic orthographic similarity. The results did not reveal the effect of visual similarity. In addition, when comparing the visual similarity ratings from both the native Chinese and the foreign participants, who do not have linguistic knowledge of Chinese writing system, and the native Chinese participants’ VSTM performance, the results further confirmed that non-linguistic visual similarity did not influence the retention performance. Different from the physical perception (e.g., the similarity of stroke patterns), Chinese orthography is appeared at an abstract linguistic representation which would only have influence on the language users. Overall, the current findings supported the role of orthographic representations in VSTM, and were consistent with the psycholinguistic approach of VSTM. |
