本研究是探討中文雙字詞的語意透明度是否影響詞彙在心理辭典中的表徵方式。實驗一、二操弄詞彙的語意透明度、字頻及詞頻，以詞彙判斷作業進行。結果發現顯著的語意透明度效果，語意不透明的詞（成語詞）之反應時間和錯誤率均較語意透明的詞（組合詞）少；詞頻效果也達顯著，高頻詞的反應時間和錯誤率均較低頻詞少。而字頻與語意透明度間有交互作用存在，組合詞有顯著的字頻效果，成語詞則不受字頻影響。實驗三、四則利用眼動追蹤技術，探討不同語意透明度、字頻、詞頻的詞彙在語句中的閱讀處理歷程。結果發現，在有關凝視時間的指標上，均出現顯著的語意透明度效果和詞頻效果。成語詞的字頻效果除了出現於第一次凝視時間指標外，在其餘指標上均無字頻效果。組合詞的字頻效果則出現於第一次凝視時間及總凝視時間上，而第一次經過的其餘眼動指標也有存在字頻效果的趨勢。本篇研究以Taft annd Zhu(1995)的多層次交互激發表徵模式為架構，提出一中文成語詞與組合詞的詞彙表徵模型。此模式顯示組合詞在詞素層次具有詞素表徵，在詞彙層次具有全詞表徵。而成語詞在詞素層次無對應的詞素表徵，是以全詞方式表徵於詞彙層次中。 關鍵字：眼動、語意透明度、心理辭典 / The present study aims to investigate the semantic transparency of Chinese di-syllabic words whether influences the representation in mental lexicon. Two sets of experimental tasks were adopted. In the first set (experiment 1 and 2), the target words were varied in terms of semantic transparency, character frequency and word frequency. Subjects were asked to deicide whether the target word presented is a word or not. A two-way ANOVA indicated main effects for semantic transparency that idiomatic words were responded faster than compositional words, for word frequency that high frequency words were responded faster than low frequency words. There were also significant interactions between character frequency and semantic transparency. In the compositional condition, the results clearly showed a significant character frequency effect, but in the idiomatic condition, the results showed no character frequency effect. In the second set (experiment 3 and 4), eye movements of subjects were monitored as they read sentences containing target words that varied on terms of semantic transparency, character frequency and word frequency. There were significant semantic transparency effect and word frequency effect in the fixation duration measures. Except the first fixation duration, there was no character frequency effect with idiomatic words in other measures. The results showed a significant character frequency effect with compositional words in the first fixation duration (FFD) and total viewing duration(TVD), and showed a marginal significant character frequency effect in other first pass measures. The present study proposed a model of mental lexicon based on the Interactive-activation Model (Taft & Zhu, 1995). In our model, the compositional words have morpheme and whole word representation, and the idiomatic words are processed as single units, have whole word representation but no morpheme representation. Key words：eye movement、semantic transparency、mental lexicon

本論文運用文字探勘相關技術，嘗試自動計算法條間的相似度，輔助專家從公司法眾多法條中整理出規則，建立法條之間的關聯，使整個法典並不是獨立的法條條號與法條內容的集合，而是在法條之間透過語意的方式連結成網路，並從分析與解釋關聯的過程中，探討文字探勘技術運用於法律條文上所遭受之困難及限制，以供後續欲從事相關研究之參考。 本論文的研究結果，從積極面來看，除了可以建立如何運用文字探勘於輔助法律知識擷取的方法之外，另一方面，從消極面來看，倘若研究結果顯示，文字探勘技術並不完全適用於法律條文的知識擷取上，那麼對於從事類似研究的專業人員而言，本論文所提出的結論與建議，亦可作為改善相關技術的重要參考。 / This thesis tries to use text mining technique to calculate, compare and analyze the correlation of legal codes. And based on the well-known defined legal concept and knowledge, it also tries to help explain and evaluate the relations above using the result of automatic calculation. Furthermore, this thesis also wishes to contribute on how to apply information technology effectively onto legal knowledge domain. If the research reveals the positive result, it could be used for knowledge build-up on how to utilize text mining technology onto legal domain. However, if the study shows that text mining doesn’t apparently apply to knowledge extracting of legal domain, then the conclusion and suggestion from this thesis could also be regarded as a important reference to other professionals in the similar research fields.

法律明定使用者可以對別人的創作內容有一定程度的合理使用(Fair Use)範圍，像是以教學為用途而可影印書籍部份內容。然而合理使用的規範判定在電腦上難以實作，現有的DRM (Digital Rights Management, 著作權管理)系統甚少實現這樣的理想。 本研究在現有的ODRL2語言上架上使用本體論(Ontology)語言來加強本身所缺乏的語意，讓本體論與規則語言(Rule Language)結合，來達到規範(Policy)標示的能力，並且使用現有的推論引擎(Reasoning Engine)，拿到規範正確執行的能力。最後在帶有規範執行的機制底下，標示合理使用的基本規範，並加強ODRL的標示能力去逼近合理使用的精神。 / United States copyright laws grants users the rights to make “fair use” of copyrighted works, e.g. copying part of a book for the use of education purpose. Current digital rights management (DRM) systems are hard to enforce fair use doctrine for two reasons. First is that the current XML-based rights expression language (REL) are unable to describe the rights of fair use; and second is that the architecture of DRM systems interferes with fair use. This approach proposed a rights expression language based on ODRL 2.0 with Semantic Web technologies to get the ability to describe fair use policies. In addition, we design a fair use procedure mechanism for managing and enforcing fair use policies defined by trusted third parties. Both REL improvement and new procedure are done to approximate fair use.

客語的「有」字就其語法功能而言可以作動詞、助動詞、及動詞補語用。就其意義上而言，「有」標示著領屬義、強調義、及結果完成義。在這些共時歧異的功能意義背後，實存在著一個核心意義---領有。本論文旨在用語法化理論將客語「有」字之共時分歧用法作一合理之解釋，而提出三個主張：第一、將存在的概念提升至更深層，將之視為領有義的預設。第二、將領有義分析為更基本的基模（source schemas）。第三、利用這些基本的基模來解釋「有」字的各種延伸意義。透過這樣的角度分析客語「有」字句，將可以更合理的解釋「有」的共時歧異現象。 / Hakka Yu 'to own, to possess' has various grammatical functions. It can function as a full-fledged verb, appear in a serial verb construction, serve as an auxiliary, and act as a verbal complement. With the structural complexity, Hakka Yu denotes possession, emphatic assertion and the existence of the resultative state of being. The divergent syntactic behaviors as well as the different functions of Hakka Yu do not share the same surface form by chance. The coincidence suggests that there must be an extant semantic core that links together all the seemingly unrelated functions exhibited by various syntactic structures of Yu. This thesis, aiming at linking the diversity of usage of Yu, claims that the structural divergence of Yu corresponds to the "cline of grammaticality," which synchronically refers to a continuum at one hand of which is a fuller lexical form, and at the opposite end, a reduced grammatical form. Three major claims are therefore advanced to account for the synchronic differences of Hakka Yu.

以本論文研究不同價格呈現方式下，參考價格對消費者行為的影響，所採用的自變數為參考價格的高低、參考價格的表達語意與折扣幅度的大小。而欲衡量的變數是消費者對促銷的可信度、認知的價格標準移動程度、因促銷所知覺的節省程度及搜尋其它零售點所能產生的利益。 經過因子設計的實驗後，發現： 一、消費者所認知的最可能市價、最高市價、最低市價，在看過有外部參考價格的廣告後，不一定是以相同的方向作移動，且各別的移動百分比也不相同，所以必須各別來探討其效果。 二、出售價格的高低 (折扣幅度的大小) 會影響消費者對促銷所知覺的可信度的水準及消 費者認知最可能市價的移動程度，但不會造「最高」價格認知與「最低」價格認知移動程度的差異，且在節省知覺及知覺的搜尋利益上也無明顯效果。而且發現以下數點： (一)高出售價格的可信度大於低出售價格的可信度 (二)高出售價格造成的最可能市價上移程度大於低出售價格造成的最可能市價上移程度。 三、參考價格的高低會影響消費者對促銷所知覺的可信度、認知的最高市價、認知的最低市價、節省知覺，但卻無法影響認知中最可能市價及搜尋利益上的差異。並且發現以下數點： (一)以內部參考價作參考價及合理高參考價格的可信度大於誇大參考價格的可信度。 (二)以內部參考價作參考價與合理高參考價格造成最高市價認知上移程度小於誇大參考價格造成上移程度。 (三)以內部參考價作參考價格造成的最低市價認知上移程度大於誇大參考價格造成上移程度。(四)以內部參考價與合理高參考價格造成的節省知覺大於誇大參考價格所造成的節省知覺。 四、「過去售價」或「目前市價」來表達參考價格，只有在認知最高市價的移動比率上有所不同。且以「目前市價」造成的上移程度較大。 五、同時以「過去售價」表達誇大參考價格及「目前市價」表達合理高參考價格，比單一方式呈現單一參考價格而言 (一)認知的最低市價向上移動百分比最大 (二)知覺的搜尋利益最小，消費者最不願再尋求其它零售點。 也就是廠商若同時以過去售價及目前市價來表達價格標,且所標示價格要偏高,此時所產生的促銷效果較好,消費者較不願再尋求其他零售點進行交易。 六、加上以｢過去售價｣表達比實際出售價還低的參考價格，在促銷效果上並不比單一方式呈現參考價格或其他呈現方式來得好。

本研究從認知語言學的觀點出發，以具有類似意義之日語複合動詞「～nuku」、「～kiru」、「～toosu」為研究對象，透過複合動詞「～nuku」、「～kiru」、「～toosu」所具有之多重意思，以及其具有之中心基模，來探討其語意之異同點。 過去的研究主要是以複合動詞之前項動詞的分類為基準，來探討具有類似意義之複合動詞之異同點，然而卻無法完全說明造成其異同點之關鍵原因。本研究著重於複合動詞具有之「中心基模」與「語意擴張構造」，並主張具有類似意義之複合動詞其最大的不同，主要即來自於各個複合動詞本身具有之不同的「中心基模」與「語意擴張構造」。 透過此分析方法可以得知，複合動詞「～nuku」、「～kiru」、「～toosu」其各自的「中心基模」分別如下： 「～nuku」為「克服行動過程中所具有之障礙」 「～kiru」為「由於某動作之影響，造成某對象消失之結果」 「～toosu」為「通過某路徑之起點與終點」 由以上所述之不同「中心基模」來探討可知，複合動詞「～nuku」、「～kiru」、「～toosu」雖然具有類似意義，但在「時間範圍」、「量範圍」、「程度範圍」、「認識範圍」、「有無意志性」、「有無障礙」、「有無目的」、「積極度」、「理解度」等項目中，仍可發現其不同之處。

隨著網路的快速成長，資料探勘(Data Mining)及文字探勘(Text Mining)所須分析的資料集越來越龐大，透過單一機器執行資料探勘分析受限於記憶體大小及其計算能力，不僅運算時間大幅增加，分析資料集的檔案大小也因而受到限制；語意註解萃取出文件的重要內容，凸顯主題加強資料探勘及文字探勘的效果，而資料探勘、文字探勘和語意註解背後都牽涉到大規模的資料處理，透過雲端運算的技術使負載平衡，將運算工作分散至運算叢集中的每一台電腦，不僅加快運算和儲存的速度，更可降低整體的風險。 本研究使用Hadoop軟體實作雲端文字探勘平台，用於分散式文字探勘及結果分析，採用涵蓋21578篇新聞文件的路透社資料集(Reuters 21578)進行實證分析，依照Mod Apte切分法分為訓練資料集及測試資料集用以進行文件分類，文件分類的步驟分為數個部分，分別為進行資料格式轉換的資料前置處理、針對文件內容加註更詳盡的連結及描述的語意註解、用以產生分類預測模型的分類器(簡單貝氏分類器、餘集簡單貝氏分類器)與評估文件分類結果的評估器；路透社資料集經過去除停用字、附加語意註解資料及文本詞彙長度統計分類，再進行簡單貝氏分類器及餘集簡單貝氏分類器的訓練，比較測試資料集的分類正確率作為文件分類實證結果。 本研究根據實驗結果發現，探討去除停用字、語意註解、文件分類演算法及文本詞彙長度對於文件分類正確率的影響：(1)去除停用字使出現頻率高的停用字對於分類預測產生負面影響；(2)語意註解作為詮釋資料的取得方式，可增加文件分類的效果；(3)餘集簡單貝氏分類器，可用以減少偏斜資料對於分類預測結果的誤判；(4)文本詞彙長度較長的文章則會某種程度主導分類預測結果，造成誤判的產生，降低分類正確率；透過上述各影響因子的調整使文件分類的結果得到改善，使得文件分類正確率獲得較佳的效果。 本研究提出之系統以雲端運算環境運行文件分類演算法，使得大型資料集得以更為迅速取得分析結果，使用語意註解作為詮釋資料的來源，使得文件分類模型產生過程中有更多資訊可分析，使得機器判斷的正確程度獲得改善，亦可將文件轉換為語意網文件，供語意網搜尋引擎查詢檢索，未來應加入Twitter或Facebook等擁有大量非結構化資料的網站之資料，使本平台得以分析更大規模的資料，並且考慮資料集類別分佈的集中程度對分類正確率的影響程度，同時應實作效果更佳的分類演算法，進而改善系統整體的結果。 / Nowadays, businesses perform data mining and text mining need to handle large scale dataset. The computational resources of servers are often limited and lack of efficient to compute analytical jobs. But if they could run their data mining jobs under cloud computing clusters, they are able to get results very quickly on a large dataset without "out of memory" problems. In this paper, a series of experiments are conducted to measure and analyze the accuracy of the classification algorithms implemented on Hadoop using Reuters-21578 dataset; the process of text mining consisted of four stages: (1)data preprocessing, (2)semantic annotation, (3)classifier, (4)evaluator. Reuters-21578 had divided into training set and testing set based on Mod Apte Split, processed by stopwords removal, appended semantic annotations as metadata and splitted into several subsets according to different document sizes. Experiments outlined several issues that will need to be considered when conducting text mining. According to the experiment results, the researcher found that stopwords removal, semantic annotation, different classification algorithms and different document sizes could improve the classification accuracy. First, stopwords removal avoids common words from becoming noises that will do harm to classification result. Second, semantic annotation as the extra information could improve the result. Third, complementary naive bayes algorithm could solve the decision boundary problem which naive bayesian cannot handle. Fourth, long documents could dominate the classification results. Sixth, the class imbalance problem could cause a drop of classification accuracy. Text mining result could be improved by adjusting the parameters found above.

字彙在英語教學上佔有重要地位。因為字彙中的形式、語意和功能用法非屬於一對一之關係，所以常會造成學習者使用上的困擾，進而阻礙溝通。此篇論文主要在探討學習者對於語意之使用情形，希望可以找出以中文為母語的學習者在形式-語意連結上的共通點和所遵循的規則，以及連結背後之影響因素。 在方法上，因為強調學習者的真實使用情形，故以task填寫的方式進行語料蒐集。以同時程的研究方法來尋求同樣母語背景學習者之interlanguage共同特色。動詞因為在句子中佔有重要地位，故此研究以動詞為例來深入做探討。而選用bring, get, give, make, take等五個動詞，乃因為其出現頻率高、搭配性高，且部分語意互通之特性。task的設計則是藉助英語為母語者之語料庫，從中蒐集高頻率的語意和搭配詞等，加上國立編譯館前四冊課本之資料，來融入task的設計。受試者為台北地區中等程度學校之高三學生。採用英語在校成績、英語自我評鑑、英語程度測驗等標準，將受試者做英語能力上的分類，以便更清楚區分不同造成語意使用表現之影響變因。個人資料表也提供了受試者的基本英語背景、對英語字彙學習之看法，供做背景與表現之研究。 研究結果顯示，在受試者當中存在一個對該五個動詞使用上的偏好順序。此結果證明同樣母語的英語學習者的確有共享之interlanguage知識特點。從四個task的表現，可看出搭配詞的使用、上下文的語意推測是學習者表現不佳之處。而英語程度較高、來自社會組、來自較高排名之學校的受試者，會有較佳的語意使用表現。另一個結果顯示，出現頻率、核心語意和母語轉借為影響形式-語意連結的三個重要因素。而學習者因為程度不同，也會有不同之遵循規則。如程度較高的學習者偏重在英語的知識上，出現頻率和核心語意會互相競爭，成為影響學習者語意正確使用的最重要因素。程度較低的學習者則偏重在中文的知識上，即把母語知識轉借到英語的使用上。而且不同程度學習者表現出英語學習發展上不同的階段和歷程，程度越高的學習者，其英語上的知識越接近以英語為母語者。此外，在此研究中，開始學習英語的年紀和學習時間長短並非主要影響學習者語意使用表現之因素。喜愛英語及對自身英語有較高自信的學習者相對的有較佳的語意使用表現。而在英語字彙學習上，學習者一般仍注意字彙本身的形式，如拼字、發音，多於字彙之豐富語意及使用。雖然較多的學習者在用英語表達時，偏好先在中文中找到合適的字再轉換成相當的英文接在英語中找到合適的字來表達的學習者有較好之語意使用表現。 最後，把研究分析所得之結果，運用在教學活動上。透過語料庫的使用和語意網路圖的呈現，期望能進一步釐清語意之間的關係，也加強學習者對語意的正確使用。 / As Wilkins (1972) states that "Without grammar very little can be conveyed, without vocabulary nothing can be conveyed" (p.111), vocabulary plays an important role in the learning of a foreign language. However, the form, meaning, and function of vocabulary is not of a one-to-one correspondence. The mismapping often leads to misunderstanding and blocks communication. The study tries to explore how English learners in Taiwan manipulate the form-meaning mapping, and what the factors are to influence the mapping. From the aspects of lexical meaning, frequency, collocation, and L1 transfer, verbs are given as an example to be investigated from both the on-line corpus of native speakers, BOE, and the corpus of Chinese learners of English, obtained from four well-designed tasks. The five verbs selected are common and frequently used, but the usages are confusing for English learners. To investigate the verbs can help learners clarify the usage of the basic verbs in English. Subjects are 3rd graders from senior high schools of medial in the ranking of academic reputation in Taipei. Based on three criteria, the subjects are classified into three different levels to see the influence of variables. In addition, a subject background sheet provides the subjects' background of English and attitudes towards English vocabulary learning. From the analyses, we find that there is indeed a preference order of the five verbs among subjects with the same mother tongue. Among the tasks, collocation and the meaning inferring from contexts are difficult for the subjects to deal with. Subjects of higher proficiency level, from social science group, or from school of a higher academic reputation have a better manipulation of the five verbs. When doing the form-meaning mapping, three factors, that is, core meaning, frequency, and L1 transfer, do influence the performance. Learners with higher proficiency levels rely much on L2 knowledge, thus core meaning and frequency have a stronger preference over L1 transfer in the form-meaning mapping process. Learners with lower proficiency levels usually make use of L1 knowledge to map the L2 form with L1 meaning, thus there is more negative transfer from L1. The performance of the three proficiency levels represents the different developmental stages of interlanguage. The higher proficiency level one hs, the more his knowledge is close to the target language (L2). In the research, we also find that an earlier age and a longer period of learning do not necessarily ensure a better English ability. Furthermore, subjects who have more interest and stronger motivation to learn English generally have a higher proficiency level. In vocabulary learning, subjects often pay attention to form rather than usage and function of vocabulary. As for the process of self-expressing in English, subjects tend to make sure the words wanted in Chinese and then translate them into English. However, subjects who directly look for words wanted in English have higher proficiency levels and better performance in English. After the analyses, the results are applied to teaching activities. With the use of corpora and semantic networks, further clarification of the relations among word meanings, and enhancement of the correct use of word meanings are expected.

本研究旨在探究有關「思考｣ 的中文辭彙，藉以一窺人類思考過程的全貌，並進一步透過隱喻與轉喻等認知語意機制 (參看Kövecses and Radden 1998, Lakoff 1993, Ruiz de Mendoza 2000, Diez Velasco’s 2003等人之論著)，分析人類思考過程的中文辭彙體現。 首先，我們修改認知心理學中所提的訊息處理模型 (參看Palmer and Kimchi 1986, Sternberg 1969, and Neisser 1967等人之論著)，根據修改過後的模型，得到人類思考過程的中文辭彙體現可分為七種類型。在進一步分析後發現，有關「思考｣ 的中文辭彙所牽涉的認知語意機制包括Lakoff (1993)提出的隱喻類型，Kövecses & Radden (1998)及Diez Velasco’s (2003)提出的轉喻類型，及Ruiz de Mendoza (2003)所提出來的隱喻和轉喻的交互模式，而本研究另外發現需透過三重轉喻及兩個新的隱喻和轉喻的交互模式才能完整建立起人類思考過程的中文辭彙體現的字面語意與其象徵語意之間的連結。 / In this study, we intend to claim that the backstage cognition of thinking process, through linguistic manifestation, can be explicitly detected with six dynamic continuing stages. In addition, each stage reifies itself with linguistic expressions of concrete observable parts or organs of body, which in turn demonstrate that lexical manifestation of thinking is heavily based on such mechanisms as metaphor and metonymy. (cf. Kövecses and Radden 1998, Lakoff 1993, Ruiz de Mendoza 2000, Diez Velasco’s 2003). First of all, we refine the information processing model proposed in cognitive psychology (cf. Palmer and Kimchi 1986, Sternberg 1969, and Neisser 1967). Based on the refined model, the Mandarin lexical reification of thinking is claimed to fall into six stages corresponding to six types. A special case linking two or more stages is observed as the seventh type. Then, we probe into the linguistic data from the seven types of thinking. The analysis discovers that a full explanation of those linguistic data from the six stages of the thinking process requires the activation of some types of metaphor. We have also found metonymy provides mental access to thinking activities. As to the interaction of metonymy, the three types of double metonymy are not sufficient to account for the lexical manifestation of thinking. Triple metonymy also needs to be employed to explain the meaning derivations of the data. Finally, it is found that the three interaction models of metaphor and metonymy are insufficient to cover all the possible interactions of cognitive mechanisms behind the lexical manifestation of thinking in Mandarin. This study reveals two new interaction models not mentioned by Ruiz de Mendoza (2000).

本論文探討顏色詞「黑/烏」與「白」在台灣華語、台灣客語以及台灣閩南語中語意延伸之比較。黑與白是自然界中最原始的顏色，而在Berlin and Kay (1969) 的顏色發展歷程中處於第一階段。本文觀察顏色詞黑與白背後的認知機制，並藉此將語料分成原始意義、轉喻延伸、隱喻延伸和特殊例子這些類別來詳述。本研究發現有些隱喻延伸普遍地存在三個語言當中，有些則是某個語言特有的，而這些語言特定的顏色詞通常蘊含文化根源、發展成構式或從外語借用而來。 本文還運用Kövecses (2005) 隱喻中的文化因素來審視顏色詞黑與白。研究發現就跨語言層面而言，台灣華語發展出最多的隱喻延伸，台灣閩南語次之，而台灣客語最少。這顯示了台灣華語顏色詞的蓬勃發展跟該語言的文化、經濟或科技等活躍發展密切相關。此外，本文發現顏色詞黑大多為負面意義，而顏色詞白則多為非負面意義。顏色詞黑與白也發展出許多語意延伸的對立，例如神秘與明確、惡與善、違法與合法。 / This thesis explores the similarities and differences regarding color terms of black and white among TM, TH and TSM. Black and white are the two most fundamental colors in the natural world as designated at stage one in Berlin and Kay’s (1969) evolutionary sequence. We classify data into cognitive categories of prototypical meaning, metonymic extensions, metaphorical extensions and idiosyncratic examples, observe their corresponding lexicalization types (Packard 2000) and use Kövecses’ (2005) cultural factors in metaphor to examine them. The three languages all contain abundant data of prototypical meanings which refer to the physiologically visual color, black and white, the representative colors of human conceptual universals about night and day (Wierzbicka 1996, Goddard 1998). Via different ICM mechanisms (Kövecses and Radden 1998), the three languages develop diverse metonymic extensions. Besides, through the general metaphor ABSTRACT QUALITY IS PHYSICAL QUALITY (Goatly 2011), numerous attributes are conceptualized through the visual color black or white in metaphorical extensions. Some metaphorical extensions are universal across the three languages; some are language-specific because of cultural roots, fixed constructions or language borrowing. Idiosyncratic examples include arbitrary items and proverbial expressions. We also use Kövecses’ (2005) cultural factors in metaphor to observe the cross-cultural and within-culture dimensions. For the cross-cultural dimension, in terms of the source domain of color black or white among the three languages, TM develops the most prolific metaphorical extensions of color terms, TH has the least versatile generations, and the development of TSM color terms is in between. The complexity of TM color terms is closely related to the vivacity of cultural, economic or technological developments in that language. From the most primitive opposite of darkness and brightness, color terms of black and white also distinctively contrast with each other. The negative meanings of color terms black are proportionally dominant whereas the meanings of color terms white are mostly non-negative. Furthermore, color terms black and white reveal some evident oppositions in the metaphorical extensions such as mystery and clarity, viciousness and innocence and illegality and legality.