普拉德─威利症候群之概念系統：中文的個案研究

全文

(1)國立台灣師範大學英語學系碩士. 論文. Master Thesis Graduate Institute of English National Taiwan Normal University. 普拉德─威利症候群之概念系統：中文的個案研究 A Conceptual System in Prader-Willi Syndrome: A Case Study in Chinese. 指導教授：詹曉蕙博士 Advisor: Dr. Shiao-hui Chan 研究生：楊禮旗 Student: Elvis Li-chi Yang. 中華民國一百年六月 June, 2011.

(2) 棄燕雀之小志慕鴻鵠以高翔 —南朝梁·丘遲. “Aspiration comes before achievements; implementation is the catalyst.” — Elvis Yang.

(3) es.10. ffi4&.-*t&f,*. q n c .H er $'e, x + + 4n-# .d {ft iE € *t*fE € FrT &# z-#x-A tftM^& @n.Zg mA ;. +_{. +w ^72 qnm_ 3_1_+ 4tF. ++.&<-a'jr. +E_H2 q.ysrya4 -#jtfE. E:. fi i%- Ail fert^A+ . 4 * aatE? z|z afiLZ?:sft_,. +E+&tft.:fr a*-E 7$'-r. tFlE+'A I - . J{fd Fl€ tF+,9 *,fFe#+&i+} r,J.#jrA jr H,++ rxte.fiF. :t,4 . *1il4b& X 4e-tr\,i€ri € fl 4+h * ffi1 H . +ii4, Ii&$AE n €+J€ lEffi2^1E41€.H. ". =. tilE),- rf,a*, rF+E*fr+ft#+&.AEAEt €BWfr+**i&1 li,e* ' _trlx€ +frt^&e+& . @rf;ffise . SeR.s.EES"i, FlF+Keft'|-F#lB ' t€4**+&'t1@^e|Efll,wF1i,?L+h*. +4L.f+$ftrr\, #lE . T di . 4+*ft'. r,ltp++tH " ++r4+i+-Lid# fr,t++*tgtF=* . p{ntF":: ( H_4 4,8HE+, y.TtA#_4 i€) : .'# jrA jr € +H}. gtflErga\. n Brefa\H. w aE-@r"r+__g!. o/. A iSrul-P+:. t*.M^wz, {'nq?,tg-ffii. (:,En+w&z). 677zrolsl. *. : l. +)tH€ nil EF tL&+fr tn'' -?,)R"rrsg,+-# rA Z d' - h eLEffi A f..W4]#.&_+ & tE€.r 4 rFrnfa\ffr 2. +*frE+tA+4'Z+'h\nl1nA. /oo +. 7. , 8 8.

(4) 摘要. 普拉德—威利症候群，主要與輕度和中度智能遲緩有關，是一種影響多重官能的先天性基因異常疾病，發生率是五千分之一到兩萬五千分之一。無論是非語言的認知功能，包含智力、記憶、視覺和聽覺的處理，抑或是語言方面如語音、音韻、構詞、句法、篇章，及語用上的缺陷，歷年來都有相當的研究與發表 (Butler et al, 2004; Cassidy et al, 2009; Conners et al, 2000)。在一次偶然的機會下，實驗者接觸到一位普拉德—威利症候群的病友(AH)，在和她的對談中，發現當她被問到晚餐吃甚麼菜時，她似乎只能回答「菜菜/青菜」這類的高階範疇概念 (superordinate)，至於下轄於「蔬菜(菜菜/青菜)」之基本層面概念(basic level)的詞彙就有命名上的困難。其無法表達基本層面概念詞彙遂成研究 AH 之概念系統(語義分類)的主要動機。在文獻中，相較於非語言認知以及語言上的功能，普拉德— 威利症候群之概念系統似乎鮮少有學者進行探究。因此本研究旨在探究普拉德— 威利症候群之概念系統，希望透過一連串的實驗作業了解此症候群之概念系統的運作模式，並試圖對上述 AH 所呈現出的語言障礙提出解釋。本實驗的受試者除了 AH 這名普拉德—威利症候群之病友(實際年齡二十歲；心智年齡九歲)之外，亦招募了二十六位心智年齡相仿的三年級小學生作為對照組 (平均實際年齡九歲兩個月)。本實驗由三個線上作業組成。第一個作業是非語言作業，受試者需在三張圖片裡，挑出一張不屬於同一類別的項目；如斑馬、綿羊，和冰箱（前兩者屬於動物類，而「冰箱」屬於餐具廚具類）。第二個作業是多模組(multi-modal)的語言作業，在播放預錄的高階範疇(superordinate category) 之音檔後（如「動物」），再以視覺呈現圖形（如「豹」），而受試者必須決定所呈現的圖檔是否屬於之前所播放的類別。在這項作業裡，實驗者操弄高階範疇之原型性(prototypicality)，包括核心(central)與邊緣(peripheral)兩大部分（例如「豹」與「蝸牛」在「動物」類裡分屬核心以及邊緣兩類）。第三個作業亦是一項多模 i.

(5) 組(multi-modal)的語言作業，在播放預錄的基本層面(basic level)詞彙之音檔後（如「貓」），再以視覺呈現圖形（如「貓」或「老鼠」），而受試者需在聽完音檔後分辨所呈現的圖形是否為如音檔所述之詞彙。在這項作業裡，實驗者操弄基本層面詞彙之語義相關性(semantic relatedness)，包括相關與不相關的項目（如「老鼠」與「貓」相關；「大象」與「貓」(較)不相關）。本實驗的三項作業以縱的面向以及橫的面向剖析 AH 的概念系統:其縱面顯示於作業一與二之高階範疇和基本層面詞彙之間的分類階層結構(taxonomical hierarchy)，探究 AH 的語言以及非語言的語義分類(semantic categorization)，並探討此分類階層是否完整；而其橫面顯示於作業二與三之核心與邊緣項目以及語義相關與語義(較)不相關項目。本研究有數個重要發現。第一，AH 在作業一、二上的高錯誤率以及過長的反應時間，顯示出她的分類階層結構不完整。後續的質性研究要求 AH 對高階範疇下定義，結果顯示她對於語義概念的抽象化及類化能力不佳。例如她知道狗是動物，卻無法說明動物類是甚麼；而她對「家具類」、「武器類」等高階範疇並非有完全的理解和認識。至於作業二、三所操弄的原型效應與語義相關效應，在線上的實驗程序中，對於對照組來說都有其效應產生，但對於 AH 而言，因為其錯誤率過高，沒有具代表性的數據可以顯示在她的概念系統也有原型效應的存在。儘管原型效應在線上的實驗程序中對 AH 而言不能得知，但隨後的質性研究證實了原型效應確實有其影響性；例如，她知道狗和蝸牛都是動物，但是在決定蝸牛是否為動物時其考慮時間比決定狗的時間長。至於其語義相關效應，線上實驗結果顯示，相較於對照組而言，此效應在 AH 的概念系統中顯得相當薄弱不明顯。總體而言，由本實驗的結果可以得知 AH 的分類階層結構不甚完整；此外原型效應在她的概念系統/分類能力具有影響力，但其語義相關效應的影響就不如正常受試者來得明顯。. ii.

(6) 本研究是國內第一個有關普拉德—威利症候群概念系統的個案研究；此研究結合了語言學，心理學以及認知科學等相關領域；希望藉由此研究的發現開啟日後對普拉德—威利症候群概念系統或語言障礙上的廣泛及深入探討，並於心理語言學及罕見疾病特殊教育上有所貢獻。. 關鍵字: 普拉德—威利症候群；概念系統；分類階層結構；語義結構；原型效應；語義相關效應. iii.

(7) ABSTRACT. Prader-Will Syndrome (PWS), mainly associated with mild to moderate mental retardation, is a multiple genetic anomaly disorder affecting multiple body systems, with an incidence estimated at 1:5000 to 1:25000. Deficits in non-language cognitive functions, including intelligence, memory, visual and auditory processing, and disruptions in language-specific functions, such as phonetics, phonology, morphology, syntax, discourse, and pragmatics have been investigated and reported (Butler et al, 2004; Cassidy et al, 2009; Conners et al, 2000). During a chat with a PWS patient (AH), it was observed that she had difficulties naming/specifying the types of vegetables (basic level exemplars) that she ate for dinner and could only give the superordinate term “VEGETABLE” as a response. Her disability of specifying the basic level exemplars under a superordinate category was a motivation for the conduction of a study about AH’s conceptual system (semantic categorization). The non-language (non-verbal) cognitive and language functions in PWS have received a certain level of attention, but there seems not much, if any, study dedicated to investigating the conceptual system in PWS individuals. Therefore, this study aims to empirically explore the operation of conceptual system in PWS and to give reasonable accounts of AH’s aforementioned language disruptions. A PWS patient/AH (Chronological age: 20/Mental age: 9) and twenty-six mental-age-matched normal controls (Mean Chronological age: 9;2) were recruited for the online experiment. The experiment consisted of three online tasks: Task One was the Exclusion Task, in which the subjects were presented with three coterminous items on the computer screen and had to exclude one item that did not belong to the same superordinate category as the other two. For example, in the list of “zebra”, “sheep” and “refrigerator”, the former two were within the category of “ANIMAL”, iv.

(8) and “refrigerator” was under “CUTLERY AND KITCHEN UTENSIL” and thus should be excluded. Task Two was the multi-modal Prototype Task. Subjects were first presented with an auditory superordinate category (e.g. “ANIMAL”), and then a picture (e.g. “leopard” or “snail”) was presented visually, which subjects had to judge if the picture belonged to the category they had just heard. The factor of prototypicality was manipulated so that the experimental trials included either central or peripheral exemplars of the named category (such as “leopard” as central and “snail” as peripheral exemplars for “ANIMAL”, respectively). Task Three was the multi-modal Semantic Relatedness Task. The subjects were first presented with an auditory basic level category, and then a visual picture was presented. They had to judge if the picture was the basic level category that they had just heard. The factor of semantic relatedness was manipulated in this task so that the experimental trials included both related and unrelated items (e.g. “rat” was related and “elephant” was unrelated to the basic level category, “cat”). The three experimental tasks were to explore the vertical and horizontal aspects of AH’s conceptual system. For the vertical part, Tasks One and Two examined the verbal and non-verbal semantic categorization of AH, and at the same time investigated the preservation (intactness) of taxonomical hierarchy between superordinate and basic level categories in AH. On the other hand, the horizontal aspect lay in the prototypicality (central vs. peripheral) of exemplars within/under superordinate categories and in the semantic relatedness (related vs. unrelated) between basic level categories. Several significant results were found. First, AH’s higher error rates and longer response times in Tasks One and Two revealed that her taxonomical hierarchical structure was incomplete/partially preserved, mainly due to her lack of knowledge in some superordinate categories, such as “FURNITURE”, and “WEAPON”. In a follow-up interview, AH’s obstructive definitions of the superordinate categories v.

(9) further evidenced her deficiencies in abstraction and generalization (e.g. she could name “dog” as an animal, but failed to define “ANIMAL”). Also, prototypicality was effective in the normal controls, but the high error rates in AH made her data not eligible for statistical analysis. However, the follow-up interview revealed that the prototypicality effect was indeed effective; for example, more time was consumed in her deciding if “snail” was “ANIMAL” than deciding if “dog” was “ANIMAL.” Semantic relatedness was also effective in the normal controls, but marginal (not as evident) in AH. In sum, compared to the normals, AH was reported to possess incomplete semantic structure (taxonomical hierarchy), to have prototypicality effects in her categorization, and to have marginal semantic relatedness effects in her recognition. This study, interfacing with linguistics, psychology and cognitive science, is the first case study in Taiwan that dealt with the conceptual system in Prader-Willi Syndrome. It is sincerely hoped that the results of this study can open up broader and deeper investigation of this disease, either about its conceptual system (semantic structure) or about its general language dysfunctions. It is also hoped that this study can make some contributions to psycholinguistics and special education on patients of this rare genetic disease.. Keywords: Prader-Willi Syndrome (PWS); conceptual system; taxonomical hierarchical structure; semantic structure; prototypicality effects, semantic relatedness effects. vi.

(10) ACKNOWLEDGEMENTS. I have accomplished one formidable task that I thought of as barely possible to complete. As to a graduate student with a background in linguistics only, it is “hard” to tackle the topic interfacing with linguistics, psychology, medicine and special education. However, this thesis was the study that testified that I was able to transcend my limitations and explore impossible possibilities. There have been two happiest moments of my life so far: One is when I was born in Taiwan and nurtured by my parents, and the other is when Prof. Shiao-hui Chan agreed to have me as her advisee. Prof. Chan was not only an incredibly “nice” mentor but also an academically “assiduous” advisor. She was quite demanding, always pinpointing potential flaws of my thesis in order to have me finish one decent work; she was quite liberal, always listening to my ideas and giving me more leeway to make my own decisions; she was quite patient and willing to help, always meeting and discussing with me even though she had a tight schedule: we had met over fifty times over the past one year (once/twice almost every week; each time was around two hours); she was quite encouraging, always invigorating me whenever I was on the suicidal verge. Without her demandingness, liberality, patience, willingness to help, and encouragement, I would hardly have completed my work. Therefore, you deserve my biggest thanks, Prof. Chan! Sincere gratitude is expressed to the two committee members, Prof. Doris Chun-yin Chen and Prof. Hintat Cheung, for providing their expertise to this thesis. Their critical and insightful comments have transformed the earlier versions of this thesis a decent work. Without their meticulous perusal and precious suggestions, there would have been more problems or defects in this thesis. Of course, none of the abovementioned professors is responsible in any way for the weaknesses of this thesis; vii.

(11) I shoulder full responsibility for all the potential errors and mistakes. Besides, I would like to thank Prof. Doris Chen for the introduction of me to Lydia Chang; without her introduction, the whole plan would have been just a plan. Deep indebtedness is owed to Taipei GuTing Elementary School, where the twenty-six normal control students were recruited. I would love to show my profoundest gratitude to an English teacher, Lydia Chang and a computer teacher, Tzu-ying Chang. Lydia Chang was the one who mainly helped me with the distribution of the consent sheet and with the manhunt; without her sincere support, the subject recruitment would have been more difficult. And Tzu-ying Chang was the one who helped me with all the technique and equipment arrangements, which made the entire experimental run smoothly and swimmingly. Also special thanks go to my well-trained and helpful assistants: Caroline Hong, Jeffery Huang, and Vincent Li. Without their assistance, collecting data within one or two days would have been castles in the air. I would like to thank the professors who helped me with the running of the pilot tests. Prof. Hugo Zeng (English Department) and Prof. Shiah-pey Sheu (Japanese Department) from Soochow University, and Prof. Vincent Wu-chang Chang, Prof. Doris Chun-yin Chen, and Prof. Jen-i Li from National Taiwan Normal University. I would also like to thanks other professors who taught me in NTNU: Prof. Shiao-hui Chan (neurolinguistics), Prof. Tammy Miao-Hsia Chang (academic writing, cognitive linguistics, and discourse analysis), Prof. Hsin-yi Chen (statistics), Prof. Miao-ling Hsieh (syntax I), Prof. Shu-kai Hsieh (computational linguistics), Prof. Tzyh-lai Huang (etymology), Prof. Jen-i Li (general linguistics and semantics), Prof. Hsueh-o Lin (pragmatics), Prof. Gerardo Fernández-Salgueiro (syntax II), Prof. Kwock-ping Tse (advanced phonetics and phonology), and Prof. Joy Jing-lan Wu (typology). Their industriousness toward lecturing and researching has inspired me so much and also viii.

(12) ushered me into the kaleidoscopic field of linguistics. I am particularly indebted to Prof. Tammy Chang. She helped me “a lot” with the revision of a conference paper regarding “Taiwanese Ability Verbs”, which enabled me to have a successful presentation at the conference in Hong Kong, and she also wrote me a letter of recommendation to help me obtain the financial support for the conference from NSC. Of course, I will always recall the happy moments that I studied in NTNU with my classmates and friends: Frank Chen, Carolin Kuo, Lucy Lee, Toro Lee, Vincent Li, Stella Liu, Sally Tsai, Yi Wang, Aaron Wu. Special gratitude goes to Muhan Wang, who helped me a lot throughout my three-year graduate school life and never lost her patience with my abundant questions. Finally my innermost gratitude is dedicated to my parents. Although they never have the slightest awareness of what I have been doing/what I have done during my graduate life, they always show their unconditional love and never-withering devotion to me. Come what may, they are always with me and around me, never pushing me into giving up my study and earning money to pay them back. Their unselfish dedication to me always spurs me to do good to society, to do well on study, and to do better for the people I love.. ix.

(13) TABLE OF CONTENTS CHINESE ABSTRACT ............................................................................................. i ENGLISH ABSTRACT ........................................................................................... iv ACKNOWLEDGEMENTS ..................................................................................... vii TABLE OF CONTENTS ...........................................................................................x LIST OF TABLES ................................................................................................. xiii LIST OF FIGURES .................................................................................................xiv Chapter One Introduction ....................................................................................1 1.1 Motivation ....................................................................................................1 1.2 Research Questions .......................................................................................5 1.3 Significance of the Study ..............................................................................6 1.4 Organization of the Thesis ............................................................................7 Chapter Two Literature Review ...........................................................................8 2.1 General Background of Prader-Willi Syndrome .............................................8 2.2 Non-verbal Cognitive Functions in PWS ..................................................... 10 2.2.1 Intelligence....................................................................................... 10 2.2.2 Memory ......................................................................................... 11 2.2.3 Visual and Auditory Processing……. ............................................... 12 2.2.4 Higher-order Processing… ............................................................... 12 2.3 Language Functions in PWS .....................................................................14 2.4 Conceptual System ..................................................................................... 17 2.4.1 Ontology (Ontological Types) .......................................................... 18 2.4.2 Concept, Categorization, and Fuzzy Boundaries ............................... 20 2.4.3 Internal Conceptual Structure and Prototype/Prototypicality Effects………………………………………………………………...23 2.4.4 Internal Structure of Hierarchies and Basic-level Characteristics and Effects ............................................................................................... 26 2.4.5 Semantic Similarity/Relatedness....................................................... 29 2.4.6 Taxonomic Hierarchy and Meronomic Hierarchy ............................. 31 2.5. Summary of Chapter Two .......................................................................... 32. x.

(14) Chapter Three. Methodology. .............................................................................. 33. 3.1 Subjects ....................................................................................................... 33 3.2 Design and Materials ................................................................................... 34 3.2.1 The Exclusion Task (Task 1) ........................................................... 35 3.2.2 The Prototype Task (Task 2) ........................................................... 36 3.2.3 The Semantic Relatedness Task (Task 3) ......................................... 40 3.2.4 Counterbalancing and Randomization of the Stimuli ........................ 45 3.2.5 Pilot Tests: Determination of Prototypicality and Semantic Relatedness ....................................................................................... 46 3.2.6 Follow-up Interviews with AH and a Normal Control ....................... 51 3.3 Procedure....................................................................................................52 3.4 Data Analysis .............................................................................................. 55 3.5 Summary of Chapter Three ......................................................................... 56 Chapter Four. Results .......................................................................................... 57. 4.1 Overall performance ................................................................................... 57 4.2 Prototypicality Effects................................................................................. 61 4.3 Semantic Relatedness Effects ...................................................................... 66 4.4 Qualitative Research on AH ........................................................................ 67 4.4.1 Detailed Analysis on AH’s Responses in Tasks 1-3 .......................... 68 4.4.2 Follow-up Interviews with AH and a Normal Control ....................... 74 4.5 Summary of Chapter Four ........................................................................... 94 Chapter Five. Discussion and Conclusion ........................................................... 96. 5.1 Conceptual System of the PWS Patient ........................................................ 96 5.1.1 Taxonomic Hierarchical Structure of AH ......................................... 98 5.1.2 Prototypicality Effects .................................................................... 103 5.1.3 Semantic Relatedness Effects ......................................................... 106 5.2 Limitations of the Present Study and Suggestions for Further Research ..... 110. xi.

(15) References .............................................................................................................. 112 Appendix I Prototypicality Pilot Test (Questionnaire A) ......................................... 118 Appendix II Prototypicality Pilot Test (Questionnaire B) ........................................ 125 Appendix III Semantic Related Object Naming Pilot Test (Questionnaire A).......... 132 Appendix IV Semantic Related Object Naming Pilot Test (Questionnaire B) ......... 134 Appendix V Semantic Relatedness Degree Rating Pilot Test (Questionnaire A) ..... 136 Appendix VI Semantic Relatedness Degree Rating Pilot Test (Questionnaire B) .... 145 Appendix VII Prototypicality and Semantic Relatedness ...................................... 155 Appendix VIII Stimuli Used in the Exclusion Task (Task 1)................................... 173 Appendix VIII Stimuli Used in the Prototype Task (Task 2) ................................... 175 Appendix VIII Stimuli Used in the Semantic Relatedness Task (Task 3) ................. 177 Appendix IX Consent Sheet ................................................................................... 179 Appendix X Photos of the Online Experimental Process ......................................... 180. xii.

(16) LIST OF TABLES Table 2.1 Lexical Hierarchies According to Two Variables of Hierarchical Relation…................................................................................................. 32 Table 3.1 A Summary of the Subjects ....................................................................... 34 Table 3.2 Stimuli Used in the Prototype Task (Twenty-four Central and Twenty-four Peripheral Objects with Averaged Points) ................................................. 38 Table 3.3 Stimuli Used in the Semantic Relatedness Task (Twenty-four Identical Objects vs. their Corresponding Related and Unrelated Objects with Averaged Points)....................................................................................... 41 Table 3.4 A Summary of the Three Experimental Tasks ........................................... 44 Table 3.5 A Summary of the Pilot Tests ....................................................................50 Table 3.6 A Counterbalance Table ............................................................................. 55 Table 4.1 Error Rates (%) and Response Times (ms) of AH and the Normal Controls ........................................................................... 58 Table 4.2 Z-scores of Error Rates (%) [ER_Z-Score] and Response Times (ms) [RT_Z-Score] in AH ................................................................................. 59 Table 4.3 The Range of Error Rates (%) and Response Times (ms) in the Normal ............................................................................................ 60 Table 4.4 The Error Rates of the Twenty-four Peripheral Objects in the Normal Controls vs. the Rate of Zeros Given by Adult Raters (marked as Zero Given) in the Pilot Test.............................................................................. 62 Table 4.5 Paired t-tests of Error Rates (%) and Response Times (ms) in Tasks 2 & 3 for the Normal Controls and AH ............................................................... 65 Table 4.6 Comparison between Expected and Responded Answers in Task 1 (13 errors) ................................................................................................. 68 Table 4.7 Twenty-four Central Objects and AH’s Responses in Task 2 (12 errors) ................................................................................................. 71 Table 4.8 Twenty-four Peripheral Objects and AH’s Responses in Task 2 (22 errors) ................................................................................................. 72 Table 4.9 Twenty-four Identical Objects and AH’s Responses in Task 3 (7 errors) ................................................................................................... 73 Table 4.10 Definitions of the Thirteen Superordinate Categories by AH and the Normal Control, PJ ................................................................................... 75 Table 4.11 AH’s Re-performance of Task 2.............................................................. 89 xiii.

(17) LIST OF FIGURES Figure 2.1 Taxonomic Hierarchy of Three Different Levels of Specificity... ………. 27 Figure 3.1 An Example Trial in the Exclusion Task….. ……………………………. 36 Figure 3.2 An Example Set in the Prototype Task. …………………………………. 40 Figure 3.3 An Example Set in the Semantic Relatedness Task….. …………………. 44. xiv.

(18) CHAPTER ONE INTRODUCTION. 1.1 Motivation Inspired by the conversation between a Prader-Willi Syndrome (PWS) patient and her elder sister, I decided to embark upon an odyssey of exploring the language deficits of the patient. What follows is an excerpt of their conversation. (The patient was named AH; and A was her older sister). A:. 你吃什麼?. ni chi shemo you eat what „What are you eating?‟ AH: 吃放(飯?)啊! chi fang (fan?) a eat put (rice?) exclamation mark (EM) „I am eating dinner!‟ [fan(rice) was the intended word, but it was imprecisely. A:. articulated due to the anomaly of her articulatory structure. fan(rice)=> fang(put): hypernasality]. 吃飯喔!?配菜喔!? 1 chi fan o pei eat rice EM match. cai vegetable. o EM. „Eating your dinner with some dishes (vegetables)!?‟ AH: 對呀! dui a yes EM „Yes!‟ 1. In Taiwanese/Chinese culture, “吃飯 (chi fan)” has a broad sense that is similar to “having a meal” in English. However, in this conversation, the idea of “吃飯 (chi fan)” was more restricted to “having boiled rice only”; this interpretation was triggered mainly owing to the influence of the contextual phrase “配菜 (pei cai)”. “配什麼菜 (pei shemo cai)” means “what kinds of dishes (e.g. vegetables, meat, or fish) that someone has “to go with boiled rice, which is usually unpalatable if eaten alone. As to a Chinese/Taiwanese, the concept of “吃飯 (chi fan)” is usually composed of “ having plain boiled rice”, “eating accompanying dishes”, and sometimes “having soup”. 1.

(19) A:. 配什麼菜? pei match. shemo what. cai vegetable. „What vegetables are you eating to go with the rice?‟ AH: 什麼?. A:. shemo what „What?‟ 配什麼菜?. pei shemo cai match what vegetable „What vegetables are you eating to go with the rice?‟ AH: 什麼(青浦?)菜?. A:. shemo (qingpu) cai what (qingpu?) vegetable „What vegetable?‟ 我說你配什麼菜? wo shuo. ni. pei. shemo. cai. I say you match what vegetable „I said, “What vegetables are you eating to go with the rice?”‟ [The older sister sounded impatient after asking the same question three times.] AH: 配菜喔? pei. A:. cai. o. match vegetable EM „What vegetables am I eating?‟ [AH confirmed the question that she had heard.] 嘿呀! hei a yep. „Yep!‟ [The implicature that “Thank God! You finally understood my question.” could be inferred from her tone.] AH: 配那個菜呀! pei match A:. nage that. cai a vegetable EM. „I am eating “THAT” vegetable!‟ 什麼菜呀? Shemo What. cai a vegetable EM. „What vegetable?‟ 2.

(20) AH: 那個那個，那個阿公種的菜呀! nage that A:. nage that. nage that. agong zhong grandpa grow. de de. cai a vegetable EM. „That, that, that vegetable that Grandpa grows!‟ 什麼菜呀?菜那麼多種，高麗菜，筍子也是菜呀! Shemo cai a what vegetable EM gaolicai sunzi. cai namo vegetable that ye shi. duo many cai. zhong kind a. (white) cabbage bamboo shoot also is vegetable EM „What vegetable? There are “SO” many kinds; (white) cabbage, or bamboo shoot is a vegetable, too.‟ AH: 高麗菜啊! gaolicai. A:. a. (white) cabbage EM „(White) cabbage!‟ 是哦?! shi o yes/really EM „Really?!‟. Based on the conversation above, it was noted that AH had some difficulties with general language abilities, including hypernasality, excessive fillers “that, that, and that… (那個那個，那個…)”, repetitiveness in speech, and some pragmatic cue insensitiveness. And word-retrieving difficulties could be obviously observed in this conversation, too. AH tended to use adjectives or relative clauses to describe the specific object(s) that she could not name/label. Her disabilities of specifying the basic level exemplars (i.e. she seemed only able to give the superordinate category “VEGETABLE” as a response instead of specifying what vegetables she ate) under a superordinate category (VEGETABLE) should be analyzed and investigated along three dimensions: whether she had no stored knowledge for “types” of vegetables at all (i.e. conceptual representation of the exemplars under “VEGETABLE”), or she could conceptualize vegetables/vegetable kinds but did not have a label for them (i.e. 3.

(21) linguistic representation). Alternatively, she might have both the conceptual and nominal knowledge, but just could not retrieve it on the spot (i.e. access problem), which led to her naming difficulties. In terms of Prader-Willi syndrome (hereafter PWS), it is a neuro-developmental or multisystem genetic disorder (cf. Prader, Labhart & Willi, 1956). Major characteristics associated with PWS include mental retardation, infant hypotonia, hypogonadism, short stature and typical face traits (small forehead, almond-shaped eyes, small mouth, and thin upper lip). Since PWS was recognized as a distinct genetic syndrome, many studies (Butler, Bittel, Kibiryeva, Talebizadeh & Thompson, 2004; Cassidy & Driscoll, 2009; Conners, Rosenquist, Atwell & Klinger, 2000) have been conducted, which focus particularly on medical, genetic and behavioral (cognitive/intellectual/psychological) aspects of the syndrome. A dimension that has received comparatively less attention is language functioning. Even though retarded developments and delayed abilities in language and speech have been reported (Å kefeldt, Å kefeldt & Gillberg, 1997; Branson, 1981; Defloor, Van Borsel & Curfs, 2000; Downey & Knutson, 1995; Edmonston, 1982; Hall & Smith, 1972; Kleppe, Katayama, Shipley & Foushee, 1990; Lewis, 2006; Lewis, Freebairn, Heeger & Cassidy, 2002; Lewis, Freebairn, Sieg & Cassidy, 2000), the number of papers on language functions in PWS is still a drop in the bucket. Among the available papers on language, most of the investigators focused on the presence of articulation problems in PWS individuals, sometimes extending their discussion to phonetic or phonological problems (Å kefeldt, Å kefeldt & Gillberg, 1997; Kleppe, Katayama, Shipley & Foushee, 1990; Lewis, 2006). Some papers also reported morphological or morphosyntactic disorders (Kleppe, Katayama, Shipley & Foushee, 1990; Weiss, Lillywhite & Gordon, 1980), but virtually there have been no studies investigating the conceptual system of PWS individuals. Therefore, the goal of this thesis is to explore 4.

(22) the vertical (superordinate vs. basic level categories) and horizontal (prototypical/central vs. peripheral items; semantic related vs. semantic unrelated items) aspects of the conceptual system in a PWS patient, which includes how well, compared with a group of normal controls, is the patient‟s taxonomical hierarchical structure (semantic structure) preserved, how prototypicality and semantic relatedness affect her categorization and recognition respectively.. 1.2 Research Questions According to Cruse (1986; 2004) and Rosch (1973; 1975), semantic categorization is a cover term that comprises many notions; prototypicality effects and basic level effects are two significant mechanisms that play a role in categorization (i.e. grouping together a multiplicity of instances/items/stuff into a unitary concept). Prototypicality is a within-structure (i.e. within a categorical structure, for example, a superordinate category) notion; members/items under a certain categorical structure are arranged around the prototype (the “best” exemplar) and ranked as from centrals to peripherals. By contrast, basic level effects is a within-hierarchy mechanism and proposed to be the most “inclusive” level, evidenced by it being the most frequently-mentioned, initially-used, easiest-visualized, and earliest-acquired. In addition to prototypicality and the hierarchical arrangements, conceptual system/structure is also characterized by an important feature: relatedness. As can be repeatedly seen in various semantic networks (e.g. Collins & Loftus, 1975; Sowa, 1987), some between-category and within-category concepts are closely linked to each other when they are similar in perception or are functionally related (e.g. doctor-nurse). Hence, this study is aimed to address the following questions:. 5.

(23) (1) Is the taxonomic hierarchical structure preserved in basic level categories and superordinate categories in PWS as in normals? (2) Does prototypicality within a semantic category play a role in PWS as it does in normals? (3) Does “semantic relatedness”, which plays a vital role in normal conceptual organization/system, also exert its influence within-categories in PWS?. 1.3 Significance of the Study Categorization of superordinate and basic level categories by normal, mental retarded (Down‟s syndrome, anoxia) and even autistic children has been heavily investigated (e.g. Schwartz, 1981; Sperber & McCauley, 1984; Tager-Flusberg, 1981; 1985a/b), but there seems no studies, to date, conducted to probe into the conceptual system (semantic organization) in PWS, not to mention any other one that was carried out in Chinese context. Consequently, the present study is initiated with the aim of investigating the taxonomical hierarchical structure (categorization of superordinate and basic level categories), effectiveness of prototypicality and semantic relatedness in the female case with PWS and her performance of a series of experimental tasks is compared with that of normal controls. It is hoped that, by means of both quantitative and qualitative methods, achieved is a better understanding of how semantic organization/conceptual system operates both in the case of PWS and in normal children within Chinese context. Besides, it is sincerely hoped that the research questions addressed and the experiment designed and conducted in this thesis will shed some light on this comparatively unplowed field, and also that the findings of the present study may encourage more investigators or authorities to pay attention to the issues of language education, speech pathology, and therapeutic care associated with PWS or other mental-retarded individuals. Lastly, it is hoped, with this study, to make 6.

(24) some contributions to the individuals challenged with rare genetic diseases, particularly with PWS.. 1.4 Organization of the Thesis This thesis is structured as follows. Chapter One is the introduction to the entire study. Chapter Two reviews and discusses several aspects of PWS, including their genetic and clinical features, non-verbal cognitive deficits, and abnormal speech and language patterns, followed by a number of semantic notions that are directly related to the experiment. Chapter Three elaborates on the research design of the present study. Chapter Four reports the findings of great significance in AH, compared with the normals. Chapter Five discusses and concludes the important findings, which might contribute to future explorations.. 7.

(25) CHAPTER TWO LITERATURE REVIEW. In this chapter, studies about Prader-Willi syndrome (PWS) are reviewed in Sections 2.1-2.3, and the semantic notions that are directly related to current experimental design are reviewed in Section 2.4. Section 2.5 summarizes this chapter.. 2.1 General Background of Prader-Willi Syndrome Prader-Willi Syndrome (PWS), initially reported by Prader, Labhart, and Willi in 1956, is a multiple anomaly disorder affecting multiple body systems and is mainly associated with mild to moderate retardation, with an incidence estimated at 1:5000 to 1:25000 (Cassidy & Driscoll, 2009; Holm, Sulzbacher & Pipes, 1981). PWS results from the loss of the function of genes that are inherited from the parents. Genes that function differently when they are inherited from the mother or the father are referred to as “imprinted”. Imprinted genes appear to be involved in important developmental processes (Nicholls, Knoll, Butler, Karam, & Lalande, 1989). PWS may result from one of the three mechanisms leading to the loss of function of imprinted genes on chromosome 15 in the q11-q13 region. First, the paternally contributed PWS region of chromosome 15 may be deleted (i.e. compared to the normal paternal and maternal gametes of chromosome 15, there is a sectional deletion), which accounts for approximately 70% of the cases of PWS. Secondly, maternal uniparental disomy (UPD) of chromosome 15 (i.e. the paternal and maternal gametal copies are abnormally identical) can cause PWS in 25-28%. Finally the imprinting defect caused by a familial translocation, methylation defect or other structural abnormality of chromosome 15 may result in PWS in 2-5% of individuals (Everman & Cassidy 2000; Nicholls, Knoll, Butler, Karam & Lalande, 1989). 8.

(26) Some common clinical phenotypic characteristics present in PWS are briefly reviewed as follows (Cassidy & Driscoll, 2009): Hypotonia and Abnormal Neurologic Function: Prenatally, hypotonia is manifested as decreased fetal movement and abnormal fetal position at delivery. In infancy, decreased movement and lethargy, such as weak cry, poor reflexes, and poor suck (which leads to early feeding difficulties and poor weight gain), are noticed. Hypotonia is one of the main factors that lead to language disorders, which are discussed in detail in Section 2.3 of this thesis. Hypogonadism: In both sexes, hypogonadism manifests as genital deficit throughout life. In males, the penis may be small; small or empty scrotums and cryptorchidism (one or both testes fail to descend normally) are also reported. In females, the labia and clitoris are generally hypoplastic (below the normal development/size; in an immature state), and menarche (first period) may occur as late as the 30s. Hyperphagia and Obesity: The hyperphagia is hypothalamic in origin, resulting in lack of satiety (food satiation). The insatiable appetite makes PWS individuals keep eating, which makes obesity unavoidable. Behavioral and psychiatric disturbances: Patients may have ill-temperedness, stubbornness, and compulsive-like behavior. Specifically, due to their insatiable appetite, they tend to lie about being hungry and beg for more food. If lying doesn‟t work, food stealing is another measure they may take. Developmental and Cognitive Delays: Gross motor and language milestone are delayed. Early milestones are reached on average at double the normal age (e.g. sitting at 12 months, walking at 24 months, and words at 2 years). Cognitive delays affect further language development, including poor articulation and having trouble in speech performance. The cognitive aspect will be discussed in detail in Section 2.2. 9.

(27) Besides the above-mentioned features, there are some other characteristics, such as short stature (an average adult height of 155 cm for males and 148 cm for females), hypopigmentation (pale skin color), sleep abnormalities (reduced REM latency, excessive daytime sleepiness), and high mortality (probably because of choking on gorged food, gastric necrosis resulting from binging). The Prader-Willi Syndrome Association (1980b) reported that many children with PWS have delayed motor skill development (e.g. sit at 12 months and walk at 30 months on average (Hall & Smith 1972)) and often have problems with balance, coordination, and large muscle strength. In a word, Prader-Willi syndrome is a complex multisystem genetic disorder, which is characterized with physical, behavioral, cognitive, and, more importantly, speech and language disorders.. 2.2 Non-verbal Cognitive Functions in PWS This section introduces characteristics in the non-verbal cognitive domains in individuals with PWS, including intelligence, memory, visual and auditory processing, and higher-order processing.. 2.2.1 Intelligence Depressed intellectual functioning was one of the defining characteristics of PWS (Dunn, 1968; Prader, Labhart & Willi, 1956; Zellweger, Schneider & Johannsson, 1968). PWS is usually associated with mental retardation (50-70 % of cases) or low-normal intelligence (Curfs & Fryns, 1992; Holm, 1981). From the 57 studies reviewed and summarized by Curfs and Fryns in 1992 on 575 individuals of different ages and of different nationalities, the distribution of IQ scores is as follows: Normal 4.9%, Borderline 27.8%, Mild Mental Retardation 34.4%, Moderate Mental Retardation 27.3%, Severe to Profound Retardation 5.6%. Gender was not a 10.

(28) significant factor that influenced the distribution of the IQ scores. In addition to the severity of mental retardation of PWS, stability of cognitive functioning has also been investigated. Previous studies demonstrated that IQ scores in PWS declined with age (Crnic, Sulzbacher, Snow & Holm, 1980; Dunn, 1968; Stein, Hutt, Spitz & Hollander, 1993); however, there was little subsequent study that could verify this claim. Therefore, the current consensus is that the cognitive trajectories in PWS appear to be stable over time as opposed to other genetic disorders affecting intelligence, such as Fragile X and Down syndromes (Dykens, Hodapp, Walsh & Nash, 1992; Greenswag, 1987).. 2.2.2 Memory It has been found that short-term memory is relatively weak in children with PWS (Warren & Hunt, 1981). In Warren & Hunt‟s study, children with PWS were matched in age and IQ with children without PWS but with mental retardation of unknown etiology. They were asked to perform a series of short-term memory tasks as well as a long term memory task of phonological code retrieval. The children with PWS performed as well as the ones without PWS on the long-term retrieval task, but they performed worse on the short-term memory tasks: they lost more information that had been learned over time. Consistent with the previous findings, Conners, Rosenquist, Atwell and Klinger (2000) also found that long-term memory is superior to short-term memory in the PWS group, but not in the control group (nine age- and IQ-matched adults without PWS).. 11.

(29) 2.2.3 Visual and Auditory Processing Visual processing is superior to auditory processing in children with PWS (see Baker & Leland, 1967; Conners, Rosenquist, Atwell & Klinger, 2000 for a different view). Cognitive skills, such as visual perception, organization, and puzzle-solving abilities, have been noted as relative advantages in some people with PWS (Dykens, 2002). Taylor and Caldwell (1983) suggested that PWS participants performed superiorly to those overall-IQ-matched obese control participants in picture completion, object assembly, and block design, on Wechsler Adult Intelligence Scale (WAIS). Curfs, Wiegers, Sommers, Borghgraef and Fryns (1991) also showed that some individuals with PWS possess the capability to recognize and evaluate figural relations greater than would be expected in a similar block design on the Wechsler Intelligence Scale for Children (WISC), as evidenced by 9 of 26 PWS children in their study. Corroborating previous findings, an event-related potential study by Stauder, Brinkman and Curfs (2002) revealed that although the short term memory is impaired in PWS patients, their visual modality is more preserved (i.e. less affected) than their auditory one.. 2.2.4 Higher-order Processing PWS patients show deficits in higher-order cognitive processing, such as abstract thinking, executive functions, and metacognitive abilities (Whitman & Thompson, 2006). Metacognition abilities are defined as “thinking about thinking” or “a higher level (usually a regulatory level) that controls or dominates thinking”. Simply put, it is concerned about the awareness or analysis of one's own learning or thinking processes or it is also paraphrased as the use of executive processes in overseeing and regulating cognitive processes. Many affected individuals show difficulties changing viewpoints (i.e. obstinacy, inflexibility or rigidity of thinking) even when they are proven wrong, 12.

(30) have problems generalizing (i.e. poor at generalizing from one situation to another to see cross-situational commonalities or similitude), and have impulsive behavior without elaborate (sophisticated or fully-planned) thinking (Sullivan & Tager-Flusberg, 2000; Tager-Flusberg & Sullivan, 2000). These abstract-thinking difficulties and metacognitive malfunctioning traits lead to academic and social functioning inabilities, which further impact on the mobilization of executive functions, the use of memory, visuomotor skills, language, and objective judgment among PWS individuals (Whitman & Thompson, 2006). In fact, both Holm (1981) and Sulzbacher, Crnic, and Snow (1981) suggested that, based on the results from independent parent surveys, reading development is close to normal, while math and number concepts are relatively weak in PWS patients. Dykens, Hodapp, Walsh and Nash (1992) also reported that reading and math are slightly differentiated, with math almost a year delayed of reading. As to reading specifically, PWS patients show relative strengths in written language skills, which include vocabulary knowledge and reading decoding. Good reading decoding ability may result from the visual spatial skills of individuals with PWS that have been reported as a relative strength (Dykens, Hodapp, Walsh, and Nash, 1992). However, some studies also show that reading can be weaker than math (Conners, Rosenquist, Atwell & Klinger, 2000). Roof, Stone, MacLean, Feurer, Thompson and Butler (2000) studied different genetic subtypes of PWS and found that patients with a deletion on average have significantly lower verbal IQ scores than those with uniparental disomy (UPD). Plus, for those with a deletion, performance IQ exceeds verbal IQ while for those with UPD the opposite pattern is manifested, further suggesting that those with a deletion may have relatively greater visual-perceptual-spatial abilities, while those with UPD may have relatively greater verbal skills. This could account for the reported phenomenon that those with UPD do substantially greater on reading and 13.

(31) spelling than do those with a deletion. As for math achievement, little difference was noted between the two genotype groups on arithmetic skills, with both uniformly impoverished (Butler, Bittel, Kibiryeva, Talebizadeh & Thompson, 2004; Whittington, Holland, Webb, Butler, Clarke & Boer, 2004).. 2.3 Language Functions in PWS Ingram (1959) classified the causes of child language disorders into “mental retardation”, “impaired hearing” ,“environmental causes”, “structural, neurological or neuromuscular abnormalities affecting the speech apparatus”, and “autism”. Mental retardation not only affects the general cognitive abilities but also delays (or holdbacks) the development of language in PWS (Munson-Davis, 1988). Besides, in light of Åkefeldt, Åkefeldt and Gillberg‟s (1997), mental retardation was regarded as an underlying factor for certain patterns of language disorder in PWS. The speech and language skills are widely disparate in the severity and type of deficits among PWS patients, ranging from being nonverbal to being those who acquire normal speech and language skills by adulthood. Despite the great variability in the speech and language skills, several common speech or language characteristics have been noted, which include poor/retarded speech-sound development, hampered oral motor skills, and language deficits. PWS individuals‟ speech is often associated with articulatory imprecision (i.e. some sounds are specifically difficult for PWS individuals to pronounce/articulate correctly or precisely; some of the sounds are reported to have been distorted or altered unintentionally while uttered), hypernasality or hyponasality (i.e. improperly or abnormally converting oral sounds to nasals or in an inverse manner: too much or not enough nasal sounds in speech), flat/changeless intonation patterns, an abnormal pitch, and a harsh voice quality. Prosody, or the melody of speech, may also be malfunctioned. Speech that is very slow and sometimes 14.

(32) unconfident, typical of a “flaccid dysarthria” (i.e. faulty speech production characterized by imprecise consonants and irregular articulation, due to motor difficulties resulting from hypotonic muscle tone), may also be reported (Lewis, 2006). Zellweger (1979) presented that articulation disturbances are prevalent among most children with PWS, and dysarthria triggers the incomplete intelligibility of their speech. Developmental apraxia of speech has also been observed with some children with PWS (Branson, 1981; Munson-Davis, 1988). Dysfluency is another problem that PWS patients may have. Kleppe, Katayama, Shipley and Foushee (1990) found that many of the 18 subjects in their study were dysfluent, with the percentage of fluency ranging from 66% to 99%, with a median of 91%, and that the primary types of dysfluencies were involved with interjections and revisions. In addition to these speech difficulties, individuals with PWS may have other language problems. Language problems include inadequacies in vocabulary, grammar, morphology, discourse (e.g. narrative abilities), and pragmatics (Lewis, 2006). In terms of speech perception and production, poor receptive and expressive abilities are frequently identified, as evidenced by age normative data with 90.5% presenting receptive language delays and 91.7% presenting expressive language delays (Lewis, Freebairn, Heeger & Cassidy, 2002). Expressive language skills are usually more hampered than receptive skills (Branson,1981; Kleppe, Katayama, Shipley & Foushee, 1990; Munson-Davis, 1988). Å kefeldt, Å kefeldt and Gillberg (1997) launched a comprehensive investigation on the speech and language skills in 11 PWS subjects and controls (matched with sex, age, BMI and IQ) and found that the differences in speech and language skills (e.g. phonology, grammar, and language comprehension) were marginally significant, but the voice quality, pitch level, resonance, and oral motor function were significantly more impaired in PWS than those in the control, which is in accordance with what the previous studies suggested. 15.

(33) As to language acquisition, a weak cry and early feeding difficulties are presented mostly because of hypotonia in infancy. Reduced babbling and early language delay are often noticed. Children with PWS are 18 months of age before they begin to verbally evidence a vocabulary, combine words, and develop early syntax while typically developing children have acquired at least two words in addition to “dada” and “mama” by 12 months of age. A substantial number of affected children are much later in acquiring speech; some may be as late as 6 years of age (Lewis 2006). According to Hall and Smith (1972), some children with PWS may begin uttering short sentences at approximately 42 months of age, while other studies suggested that some children may begin talking as early as 24 months (Hall & Smith 1972; Prader-Willi Syndrome Association, 1980b). As to word usage specifically, Branson (1981) reported that the word-recall difficulties create the repetitions, additions, and circumlocutions in the spontaneous speech of the subjects with PWS, which is one of the decisive factors in the presence of dysfluent behavior. With regard to content words and function words, content words were used more than function words from each subject‟s language sample, with content words ranging from 51% to 69% (M=58%), and function words from 31% to 49% (M=42%) (Kleppe, Katayama, Shipley & Foushee, 1990). The lopsided tendency towards content-word use can be accounted for by the deficiency of the “higher-order cognitive processing” in PWS (Whitman &Thompson, 2006). Defloor, Van Borsel and Curfs (2000) investigated speech fluency in 15 individuals with PWS (CA from 9;9 to 20;0 and total IQ from 40 to 94) and found that dysfluent speech behavior is a common symptom in PWS, with whole-word repetitions being seen more often on function words while part-word repetitions being observed more on lexical (content) words. Additionally, dysfluencies often affected monosyllabic words, with whole-word repetitions and prolongations occurring preeminently on monosyllabic 16.

(34) words. Defloor et al. (2000) also presented that the dysfluent speech behaviors associated with PWS do not conform completely to developmental stuttering despite some of the shared features observed between these two. PWS patients have been found to have great difficulty with narrative or story telling/retelling tasks (Lewis, Freebairn, Sieg & Cassidy, 2000), which may be due to specific deficits in “auditory short-term memory (Dykens, Hodapp, Walsh, and Nash, 1992)”, “linear or temporal order processing”, and “auditory verbal processing skills (Curfs, Wiegers, Sommers, Borghgraef & Fryns, 1991). Besides, poor pragmatic or social skills are often observed in individuals with PWS. For instance, Munson-Davis (1988) demonstrated that deficits of pragmatics and unawareness of social norms or etiquette in communicational interactions can be a problem in older children and adolescents. Disabilities of higher-order cognitive processing mentioned previously also trigger the ignorance of implicature (implied meaning) in conversations. Individuals with PWS thus cannot or hardly seize the connotation, infer the extra meanings, follow the speech acts, and read between the lines when they have conversational interactions with other people (Whitman & Thompson, 2006). Besides, the socio-pragmatic features, such as temper tantrums, stubbornness, difficulties in detecting social cues and poor social relationships also prevent PWS individuals from abiding by the Cooperative Principle in the conversation.. 2.4 Conceptual System Having reviewed studies about PWS (non-verbal and verbal aspects), the following sections will introduce basic notions relevant to the present research, including, but not limited to, Concept, Prototype Effects (Prototypicality Effects), Basic level Effects, and Semantic Relatedness Effects. 17.

(35) 2.4.1 Ontology (Ontological Types) Ontology, defined by Matthews (2007), is a philosophical branch regarding the “nature of existence”. The question whether and in what sense a language system or its elements exist is an ontological question concerned with language. Cruse (2004) stated that at all levels of specificity difference of quality (the most salient and important aspect of variation within descriptive meaning) can be observed and that hierarchies of semantic categories of varying ontological types are thought of or formed in human cognition. What follows is a set of ontological types at the highest level of generality exemplified by Cruse (2004): THING QUALITY QUANTITY PLACE TIME STATE PROCESS EVENT ACTION RELATION MANNER. At lower levels of generality (e.g. under the scope/category of “THING”), hierarchically organized sets of conceptual categories can be found, such as Living thing: animals, reptiles, crustaceans, plants… Animals: dogs, cats, snakes… Dogs: collies, Alsatians, Pekinese… Therefore, the above illustration demonstrated that ontological types within a taxonomic hierarchy are semantic categories of different levels of generality or specificity. For instance, there are numerous kinds of dogs under the semantic category of “DOG”: Collies, Alsatians, and Pekinese are the ontological types/semantic categories at the lowest level of generality or at the highest level of specificity. According to Jeng (2005), Roget’s Thesaurus of English Words and Phrases, published by Peter Mark Roget (1779-1869) in 1852, was probably the first thesaurus that arranged and classified English words according to the six major semantic categories. The six superordinate categories were “Abstract relations”, “Space”, 18.

(36) “Matter”, “Intellect”, “Volition” and “Emotion, religion and morality”, and these collectively were referred to as an ontological model or conceptual system of the knowledge of the world at that time. Owing to the old versions (there have been many altered versions published since the advent of Roget’s Thesaurus of English Words and Phrases in 1852) being too antiquated and démodé, Kipfer (2001) modified the thesaurus and extended it from the original six to the fifteen postulated superordinate categories: (1) The Body and the Senses (e.g. Birth, Energy, Health); (2) Feelings (e.g. Feeling, Pleasure, Sadness); (3) Place and Change of Place (e.g. Space, Location, Direction); (4) Measure and Shape (e.g. Quantity, Degree, Mean); (5) Living Things (e.g. Youth, Age, Life); (6) Natural Phenomena (e.g. Season, Rain, Weather); (7) Behavior and the Will (e.g. Behavior, Action, Willingness); (8) Language (e.g. Signs, Meaning, Speech); (9) Human Society and Institutions (e.g. Relationship by Blood, Ancestry, Marriage); (10) Values and Ideas (e.g. Ethics, Right, Wrong); (11) Arts (e.g. Dance, Music, Entertainer); (12) Occupations and Crafts (e.g. Occupation, Exertion, Worker); (13) Sports and Amusements (e.g. Baseball, Golf, Boxing); (14) The Mind and Ideas (e.g. Existence, State, Inclusion), and (15) Science and Technology (e.g. Heat, Cold, Fuel). The new version, Roget’s International Thesaurus (Kipfer 2001), compliant with the Western world view of the current time, is applauded to be a great achievement over the 1852‟s outdated version of six superordinate categories. Compared with four-level hierarchical structure in Roget’s Thesaurus (Lloyd, 1982), Roget’s International Thesaurus (Kipfer, 2001) adopted three-level hierarchical structure, instead. For example, the superordinate category “Place and Change of Place” dominates the head “179 Vehicle” among other heads, and under this head there is the basic level category “train” in boldface type with several subordinate categories “railroad train”, “passenger train” and “Amtrak”, etc. 19.

(37) Commenting on Rosch (1973) and Rosch and Mervis (1975), Lakoff (1987) stated, “Linguistic categories should be of the same type as other categories in our conceptual system. In particular, they should show prototype and basic level effects.” The prototype effects (prototypicality effects) are characterized by the psychological or linguistic phenomenon of asymmetry in which some members in a category are ranked as more prototypical or central members than others. As for the basic level effects, they equally demonstrate the phenomenon of asymmetry in which a certain level of words are more basic (inclusive) than the words of other levels. More specifically, the basic level categories are presumably more salient and they constitute the major parts of our knowledge representation. In a nutshell, prototype and basic level effects play a vital role in human cognition and they are two main mechanisms that are actuated as acquisition or categorization occurs in every encounter of new experience (see Sections 2.4.3 and 2.4.4 for more details on the prototype effects and basic level effects).. 2.4.2 Concept, Categorization, and Fuzzy Boundaries Based on Cruse (1986; 2004), concepts play a significant part in the efficient functioning of human cognition. They are well-organized as bundles of stored knowledge, which is gained and accumulated through a diversity of events, entities, situations, and so forth in our experience. If we were not capable of assigning various dimensions of our experience to stable categories, it would trigger unstable chaos either in language or in society; if we were unable to apply previously-learned/ encountered experience to any new experience that is gained from any new event, we would probably make the same mistake and eventually fail to learn from the new experience. The reason why we are able to acquire new ideas/lessons from each fresh, unique experience is that we put similar elements of experience into categories that 20.

(38) we can recognize as having occurred before, and then we can access stored knowledge about them (Cruse, 2004). Concepts, further explicated by Cruse (1986; 2004), mainly function to categorize experience and give access to stored knowledge regarding entities which fall into the categories. For example, when a young kid has never seen a dog before and then is told what he is seeing/watching is a dog, then he will recognize another dog a “DOG” next time when he sees a dog regardless of it being the same dog as he has seen before. It is because the kid is able to categorize what he has seen into the semantic category of “DOG”, and his stored knowledge in his semantic memory about dogs (e.g. what a dog looks/smells like; how a dog barks) will be activated when he sees a dog again next time. Be more specific, when the young kid sees a dog, say an Alsatian, and is told that that is a dog, the icon/concept of a dog will have been formed in the kid‟s cognition/mind/semantic memory, and when the kid sees a Pekinese (a different kind of dog from that of previous time) next time, the new experience is automatically categorized and the stored knowledge is retrieved to help the kid to recognize/understand what he sees is a “DOG” even though Alsatians are distinct from Pekinese. When it comes to concepts, categorization of concepts is an issue that necessitates elaborations. Tager-Flusberg (1985a) remarked that categorization is a central cognitive ability, through which we can group together a variety of instances into a unitary concept. According to Cruse (1986; 2004), the classical approach to categorization goes back to Aristotle. An Aristotelian definition of a category has a fixed and inflexible boundary. Aristotle defines a category in terms of a set of “necessary and sufficient criteria”, which is overthrown by a famous example “GAME” put forth by Wittgenstein (1972). He argued that it is scarcely possible to come up with a list of features that can be possessed by all games and jointly used to distinguish games from non-games. One might suggest the following as possible 21.

(39) criteria of “GAME” (exemplified by Cruse 2004): (a) involving winning and losing (b) involving more than one person (c) has arbitrary rules (d) done purely for enjoyment However, there are many counterexamples demonstrating that „necessary and sufficient criteria‟ is not a bullet-proof method of defining any category. For example, a basketball game is done/played not purely for enjoyment but mainly for competition, and solitaire is a game involved with only one person. By describing the instances of the category of “GAME”, Wittgenstein (1972) manifested a relationship of “family resemblance”: A, B, and C are the three members of a category or figuratively the three members of a family, and let‟s say B is more prototypical and bears more resemblance to their parents. A has the same mouth as B, and B has the same eyes as C, but A and C have no features in common. Even though it seems that there are no commonalities between A and C, they are still linked by a chain of an intermediate member (namely B) with whom they do share features. Wittgenstein‟s family resemblance unites the members of a category, links the central and peripheral members, and further consolidates the notion of prototype theory. Also, much empirical research on category structure shows that the boundaries of some natural categories are fuzzy and contextually flexible, which are subsequently referred to as “fuzzy boundaries”. For instance, Berlin and Kay (1969) found that the judgments of central examples of colors are relatively constant and reliable while the judgments of borderline instances (for example the ones between red and orange) showed less degree of agreement, constancy and reliability among a group of subjects or within a single subject on different occasions. Also, Labov (1973), on his experiment where the subjects were asked to name the line drawings, such as illustrating cups, mugs, 22.

(40) vases, bowls, and the like, discovered that the varied parameters, like ratio of height to width, curved or straight sides, presence or absence of a handle, indeed affect the judgments/assignations of certain items to a particular category. He also suggested that contextual conditions (background information) can alter subjects‟ judgments; for example, imagination of all the possible items with rice as contents leads to the category of “BOWL”, while a similar process of imagination with tea as contents triggers the formation of the category of “CUP”.. 2.4.3 Internal Conceptual Structure and Prototype/Prototypicality Effects The internal structure of a category (conceptual structure) is a systematical organization that is structured around the „best‟ examples or prototypes of the categories, and the other items are assimilated to a category according to whether they sufficiently resemble the prototype or not. In other words, there are central members, less central members, and borderline cases distributed based on how much they take after the prototypes within the internal structure of natural conceptual categories (Rosch, 1973; 1975; Rosch & Mervis, 1975; Rosch, Mervis, Gray, Johnson & Boyes-Braem, 1976). The most basic experimental technique utilized by Rosch is the elicitation of subjects‟ Goodness-of-exemplar (GOE) ratings. Subjects are asked to grade a certain member of a category on a rating scale from point 1 to 7, with 1 being a very good example and 7 being very bad example/not an example at all. For instance, potato is usually valued with point 1 as a very prototypical example while lemon is ranked as the least prototypical one (point 7) within the category of “VEGETABLE. However, Cruse (2004) argued that GOE is not always a reliable rating technique due to its high dependency on culture. For example, in a British context, date is typically rated from 3 to 5 on the GOE scale but rated as top 1 among a group of Jordanian participants (who are reported to be more familiar with the fruit 23.

(41) date), suggesting that “familiarity” is undoubtedly a factor that influences GOE scores. There is substantial evidence showing that prototypicality, as measured by GOE scores, has a strong correlation with some aspects of human cognitive behavior. The correlations are usually referred to as prototype effects/prototypicality effects. The prototype effects are characterized with the following features: “order of mention”, “overall frequency”, “order of acquisition”, “vocabulary learning”, “speed of verification”, and “priming”(Cruse, 1986; 1990; 1994; 2004). Order of mention is often tested when subjects are asked to list a series of members of a certain category under time pressure (online); subjects usually list the most prototypical example(s), which are ranked as the top 1 on the GOE scale. For example, if the category of “VEGETABLE” is given, subjects tend to prioritize the members, such as carrot or cabbage, and lemon (GOE: 7) is probably the last example subjects might come up with. Besides, overall frequency is also the evidence showing that GOE strongly correlates with the frequency of mention. For example, if the topic of a conversation among a group of people is about how “FRUIT” positively influences health, then the most frequently-used or -mentioned word must be the items that are valued with the first degree, and apples or oranges, which are regarded as the most representative examples of this category, will probably be the frequently-mentioned words/objects in the conversation. Order of acquisition and vocabulary learning are demonstrated by the phenomenon that children tend to acquire/pick up the prototypical members of categories. When it comes to “ANIMALS”, children are more inclined to list dogs or cats, and less likely to list peripheral members, like vicunas, which is less-frequently seen or heard in the course of children‟s acquisition. As to vocabulary learning, children learn new words more readily if they are provided with definitions 24.

(42) accompanied with prototypical instantiations (concrete examples) than if they are only instructed with an abstract definition that, even though, accurately covers the entire range of the word‟s meaning. For example, when a term “孝順 (filial piety/duty)” is taught, the abstract definition (i.e. to show filial obedience and devotion to parents) works less efficiently than the actual example or concrete action that good children always behave themselves and try not to let their parents worry about their (children‟s) health and safety. In this case, “being safe and sound” and “freeing parents from worries”, at least in traditional Chinese culture, serve as a prototypical representation of the concept/category of “FILIAL PIETY/DUTY”. Speed of verification is usually presented by psychological experiments in which subjects are required to respond as quickly as they can in a categorization task. Subjects are reported to produce or react faster if the tasks are involved with a prototypical member. Compared to the pair FRUIT: apple, subjects will react/respond slower when the pair FRUIT: olive is presented. Olive seems a less prototypical example of the category of “FRUIT”, which leads to a longer period of time for consideration and judgment among subjects. In terms of priming, subjects are shown a word on the screen, for example “doctor”, and then shown another word “nurse”. After the display of the two words, subjects are asked to press “Yes” if they think these two words are related; otherwise, they press “No” if they think the just-displayed pair is less related or without relations at all; for example, a doctor: teacher pair will probably consume subjects more time to decide than a doctor: nurse pair, which shows the prototypicality effects in pair-word reaction.. 25.