| 摘要: | 自然語言的不同層次均有其內在規則和結構，例如中文字形結構多由分布於單字左右兩側的意旁和音旁組成，語句結構則由符合語意和語法規則的字詞組成，因此不同部件在單字空間中的分配和不同字詞在句子中的前後位置，均有其相對應的、空間和時間的機率訊息。成人在閱讀母語文字的無數次經驗中，重複接觸到這些規則，透過腦中的視覺統計學習能力，擷取語言中所隱含的規則，藉以發展、建立閱讀所需的各項基本歷程，因此視覺統計學習和閱讀這兩種認知功能及其背後的神經機制，彼此間很可能有交互影響的密切關係。奠基於以上的理論假設，和本實驗室於中文字形辨識、語句理解所累積的研究成果，本三年計畫延續並擴展目前正在進行的兩年期科技部計畫，提出八個實驗，利用行為派典和腦造影技術，針對以下議題進行探討：一、延伸在目前計畫中所新發展出的視覺統計學習作業，以非語言刺激材料交錯呈現在時間和空間位置向度上的規則性，並測量受試者同時學習這兩種規則性的可能性和程度(實驗一)；二、當受試者經歷視覺刺激材料在時間和空間位置向度上的規則性時，需依賴統計學習能力擷取這些規則，本計畫以功能性核磁共振造影技術(實驗二)和腦磁波儀(實驗三)對其背後的神經訊號加以測量；三、檢驗以拼音文字為母語的受試者(實驗四)和以中文為母語的受試者(實驗五)在"對時間、空間位置規則性的統計學習能力"上的異同；四、以功能性核磁共振造影技術(實驗六)檢驗"對時間、空間位置規則性的敏感度"和進行中文字形音轉換時的大腦血氧變化是否相關，並以腦磁波儀(實驗七) 檢驗"對時間、空間位置規則性的敏感度"和進行中文語法處理時的腦磁波反應是否相關；五、探討"對時間、空間位置規則性的敏感度"對於母語為中文的讀者在學習法文初期時的貢獻(實驗八)。本計畫將利用目前進行中之科技部計畫所發展出的、針對時間和空間位置規則性進行統計學習的新穎作業，進一步澄清視覺統計學習能力是否為一跨規則種類的通用學習機制，或是會因時間、空間位置中的規則性而展現不同的特性；除了驗證視覺統計學習能力對於閱讀和其習得的貢獻，也將首度探討不同母語文字閱讀經驗對視覺統計學習能力之影響，為相關文獻中的創新嘗試。本計畫的成果將為視覺統計學習能力和閱讀能力間的交互影響關係提供來自行為和腦神經反應的實徵證據，有助於建立和這兩種能力相關的理論模型，也將幫助學者更了解人類的統計學習和語言能力，和其認知功能和生理機制的密切關聯。 ;Readers of specific writing systems presumably rely on their sensitivity to probabilistic statistics of different parts in words, and different words in sentences, to develop the knowledge of regularity “hidden” in the various levels of their native language after extensive exposure to the countless exemplars of written input. Through the approach that examines the relationship between individual differences in domain-general (learning) and domain-specific (reading) abilities, the present proposal aims to investigate the characteristics and neural correlates of statistical learning in the visual modality (VSL) of positional regularity in the temporal and spatial displays, and its reciprocal relationship with the functioning and acquisition, as well as the neuronal indices of processing efficiency, of reading.Based on our extensive research on Chinese character recognition and sentence comprehension and an ongoing two-year MOST project, we propose to conduct eight experiments in three years to investigate the following research questions: First, we develop a novel VSL task that measures concurrent learning of positional regularity in the temporal and spatial displays, and to delineate the capacity and specificity of this fundamental cognitive function (Exp. 1). Second, the neural correlates of VSL of temporal and spatial positional regularity are identified via neuroimaging tools with superior spatial and temporal resolution, namely, functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG), in Exp. 2 and 3, respectively. Third, we explore whether VSL of temporal and spatial positional regularity would be modulated by long-term literacy experience of reading alphabetic or logographic orthography by testing foreign and Chinese participants in Exp. 4 and 5, respectively. Fourth, to further support the link between VSL and reading behaviors, and to determine whether VSL in the temporal and spatial displays is specifically related to reading Chinese sentences and characters, respectively, we correlate these two kinds of VSL with brain activation identified in a task that particularly demands the knowledge of orthography-to-phonology conversion of Chinese characters via fMRI (Exp. 6), and with neuromagnetic signals identified in a task that particularly demands the knowledge of syntactic dependency of Chinese relative clauses via MEG (Exp. 7). Fifth, we determine whether the VSL ability of temporal and spatial positional regularity would have different degrees of predicting power to the improvement of French word recognition in beginning learners of French who are native readers of Chinese after one year of learning time (Exp. 8).Through the proposed experiments, we will not only generate new understanding of the sub-components of VSL and how they influence literacy processing/acquisition, we will also provide critical evidence for the neural bases of such links, which are rarely explored in the literature. In addition, we will conduct the unprecedented study to examine whether long-term experience of alphabetic or logographic orthography would modulate VSL of temporal and spatial positional regularity in return. The empirical findings will elucidate the bi-directional relationship between VSL and reading, and inform relevant theories in cognitive functions and in language processing. They also point to constructive directions for practical teaching strategies that take the characteristics of native and foreign language into account. |
