原型理論

原型理論（粵拼：jyun4 jing4 lei5 leon6）係語言學同其他認知科學領域講到嘅一個概念，主張人腦決定「呢件物件，係咪屬於呢個類別」嗰陣，靠嘅唔係精確嘅定義，而係睇件物件「有幾似心目中ＸＸ嘅原型」^[1]。

原型理論會作出一個預測：是但攞一件物件，人腦會對「呢件物件有幾似典型嘅ＸＸ類物件」作出某啲判斷，雖然唔同個體做嘅判斷可能會有啲唔同；有心理學家就嘗試喺 1971 年嗰陣做實驗，驗證呢套諗法，（簡化講）例如叫啲人諗傢俬嘅例子嗰時，佢哋硬係比較大機率會諗起梳化，多過會諗起衣櫃—「梳化喺啲人心目中係比較典型嘅傢俬」。而有好多語言學家就攞住呢個概念，主張人對腦入面嗰啲字詞嘅概念，係靠住對每隻字詞都畀返個「典型例子」形成嘅。

認知心理學方面嘅知識，會幫到讀者理解篇文以下落嚟講嘅內容。

基礎概念

原型理論嘅諗頭始於 1970 年代嘅認知科學，例如當時有心理學家就噉講（以下段嘢經過咗重述）^[2]^[3]^{[註 1]}：

「

粵文翻譯：傳統上啲人分析將物件分類（呢樣舉動）嗰陣，都會將係咪屬呢個呢個類別當做一係有一係冇嘅現象。即係，好多哲學、心理學、語言學同人類學上嘅文獻都假設類別係指一件物件係咪有齊晒某拃特徵，有齊晒ＸＸ應有嗰拃特徵嘅物件，就算係屬於ＸＸ呢個類別。相對嚟講，最近有人拗^{[註 2]}... 話有一啲自然類別係模擬^{[歐 1]}嘅—意思係指一件物件唔係『一係屬呢個類別一係唔屬』，而係『有幾屬呢個類別』。

」

喺原型理論當中，一個人腦裏面會裝住一大拃類別。對於每一個類別，佢都可以搵一隻字詞嚟指個類別，而當佢要決定「眼前呢件物件，屬唔屬ＸＸ類別」嗰時，靠嘅唔係「記住ＸＸ嗰拃定義上有嘅特徵」得知件物件算唔算係ＸＸ，而係個腦入面會對每個類別都畀返個「典型例子」—呢啲噉嘅「典型例子」就係所謂嘅原型^{[歐 2]}，再比較吓眼前件物件有幾似個原型，有咁上下似嘅話眼前物件就算係屬於ＸＸ^[4]。

一個類別有咩特性，可以視乎語境而有異。舉個簡單例子，餐點呢個字詞假如放喺粵式茶樓語境當中，好可能會包括叉燒包或者燒賣等嘅成員，但同一個字詞如果擺喺例如四川菜館或者醫院飯堂，佢包含嘅成員以及有嘅特徵就可能會好唔同^[5]^[6]。

出名類別

動物物種

原型理論相關嘅文獻成日都會用「人腦點樣將唔同嘅動物分類」嚟做例子。首先，假設觀察者並唔係生物學家，唔識點樣睇遺傳特徵嚟將動物分類。

而家觀察者要思考雀鳥呢一隻字詞^[7]^[8]。人腦能夠理解一種比較「理性」嘅諗嘢方法，就係個腦記住一拃「一件物件要算係雀鳥，所需要達致嘅條件」，呢啲條件可能包括^{[註 3]}：

有兩隻翼；
有羽毛；
身體大致呈 S 字噉形；
由蛋生出嚟嘅；

呀噉。一個噉諗嘢嘅人腦攞住一件物件，會客觀噉檢驗件物件係咪合乎晒呢啲條件，而且對於每件合乎晒啲條件嘅嘢—包括天鵝、烏鴉、火雞、企鵝等—全部都一視同仁。但現實並唔係噉：研究表明，啲人係就係有能力理解點樣客觀噉將一件物件分類做雀鳥，但佢哋一見到雀鳥或者雀仔等嘅字眼嗰陣，會即時聯想起曉飛呢樣嘢^{[註 4]}，而如果有件物件唔具有呢樣特徵、但的確屬於雀鳥嘅（例如企鵝），啲人對於件物件嘅反應會稍為慢咗啲，而啲人對於有呢樣特徵又屬於雀鳥嘅物件（例如天鵝）反應硬係會快些少。一般認為，噉係反映咗人腦當中有個雀鳥嘅原型，而呢個原型係具有曉飛呢樣特徵嘅^[9]。

就算將雀鳥換做第啲動物分類，上述講嘅嘢都仲會成立。

顏色判斷

原型理論嘅另一個出名例子係人腦對色彩嘅命名。好似紅或者黃等描述顏色嘅字詞，查實有高度嘅彈性^[8]。首先好明顯嘅係，世上嘅主流語言都會有一系列嘅字詞嚟指唔同嘅顏色，例如粵語有

紅（hung4）
黃（wong4）
綠（luk6）

等嘅字詞，彷彿講到好似唔同色之間有明確分界噉。但係實證嘅心理實驗表明咗，唔同顏色之間根本冇明確嘅類別可以分到—由下圖嘅顏色輪最「紅」嗰笪抽一隻色嚟睇，多數人都會將攞咗出嚟嗰隻色描述做紅，但如果由最「紅」嗰笪同最「紫」嗰笪中間攞一隻色嚟睇，就會出現

「Err... 呢隻色係紅？或者可能係紫？」
50% 嘅人話隻色係紅，另外嗰 50% 嘅就話隻色係紫。

噉嘅情況。「色類」之間冇明確嘅分界線，只有「幾似心目中典型嘅紅色」或者「幾似心目中典型嘅紫色」等嘅判斷^[10]。

批評聲音

蝙蝠雖然同雀鳥有某啲共通點，但就算唔係讀生物學嘅人，都唔會當佢係「唔似原型雀鳥嘅雀鳥」。

喺廿世紀後期，原型理論取得咗重大嘅成功，但係都有唔少人批評佢嗰套諗法。

唔少研究者都直覺上認為，至少一部份嘅類別係理應有清楚明確邊界嘅，而唔係好似原型理論所講嘅噉「條邊界含糊」嘅。例如係雀鳥噉：根據古典嗰套原型理論，具有個類別嘅某啲特徵嘅嘢，有可能被視為「類別當中唔似原型嘅成員」而打前嘅研究已知有翼曉飛係雀鳥呢個類別嘅重要特徵，之但係具有呢啲特徵嘅蝙蝠就被視為「一定唔係雀鳥」嘅物體——噉就表示，喺最少某一啲情況下，類別之間嘅邊界係明確、非黑即白嘅^[11]。

原型理論亦都被指係難以解釋指抽象概念嘅字詞。好似傢俬、餐點、雀鳥同色彩等嘅字詞，都係指緊可以直接感知（睇得到或者摸得到呀噉）嘅物體或者特徵；相比之下，指行為嘅字詞（例如欺騙以及攻擊）或者指抽象邏輯概念（例如因為以及同埋）就被指係比較難用原型理論嗰套思路嚟想像^[12]。

亦有人批評原型理論相關研究嘅研究方法。有語言學家指出，廿世紀嘅原型理論研究者做咗一啲實驗，喺實驗過程當中似乎曾經叫受試者「評吓呢嚿物體有幾似某個某個意思」^[13]，而呢點有引導受試者之嫌——研究者開埋口叫受試者講有幾似，噉受試者就「梗係會答某某物體有幾似某個類別嘅成員」，形成咗一種情況，就係研究者畀嘅指示似係暗中引導受試者畀研究者想要嘅答覆。呢種實驗有可能高估咗原型理論「受實驗證據支持」嘅程度^[14]。

因此，好多語言學家都主張，話原型理論頂晒攏都只能夠解釋一部份字詞嘅意思，做唔到完全噉解釋所有嘅字詞語義現象。不過，原型理論依然係對語義學造成咗巨大嘅衝擊。

睇埋

註釋

↑ 順帶一提，喺英文行話當中，prototypical 係形容詞，指似原型嘅噉解。
↑ 留意呢段字出自一份喺 1975 年出版嘅文。
↑ 講緊由一般人嘅角度睇。一般人唔係雀鳥方面嘅專家。
↑ 有關聯想呢樣嘢要點驗證，可以睇吓促發效應嘅嘢。

詞表

篇文用咗嘅詞彙嘅英文版：

↑ analog
↑ prototype

參攷

篇文引述咗嘅文獻或者網頁：

↑ Coseriu, E., Willems, K and Leuschner, T, (2000) Structural Semantics and 'Cognitive' Semantics, in Logos and Language.
↑ Rosch, E. (1975). Cognitive representations of semantic categories. Journal of experimental psychology: General, 104(3), 192.
↑ 原版段英文係噉嘅："when describing categories analytically, most traditions of thought have treated category membership as a digital, all-or-none phenomenon. That is, much work in philosophy, psychology, linguistics, and anthropology assumes that categories are logical bounded entities, membership in which is defined by an item's possession of a simple set of criterial features, in which all instances possessing the criterial attributes have a full and equal degree of membership. In contrast, it has recently been argued ... that some natural categories are analog and must be represented logically in a manner which reflects their analog structure."
↑ Coleman, L., & Kay, P. (1981). Prototype semantics: The English word lie. Language, 57(1), 26-44.
↑ Khalidi, Muhammad Ali. 1995. Two concepts of concept, Mind and Language 10: 402-422.
↑ Riemer, N. (2010). Introducing semantics. Cambridge University Press. 7.1.4，"members of the category MEAL will have very different attributes if the context is a hospital, a wedding banquet, a camping trip or the family dinner table."
↑ Ibbotson, P., & Tomasello, M. (2009). Prototype constructions in early language acquisition. Language and Cognition, 1(1), 59-85.
↑ ^8.0 ^8.1 Geeraerts, D. (2006). Prototype theory. Cognitive linguistics: Basic readings, 34, 141-165.
↑ Riemer, N. (2010). Introducing semantics. Cambridge University Press，6.1.3 佢呢度仲有提到，要求受試者列出雀鳥嘅例子嗰陣，佢哋比較大機會列出似原型嘅例子，而且佢哋有可能講出企鵝嚴格嚟講係一種雀鳥噉嘅句子。
↑ Kamp, H., & Partee, B. (1995). Prototype theory and compositionality. Cognition, 57(2), 129-191.
↑ Cruse, D. A. 1990. Prototype theory and lexical semantics, in Tsohatzidis, S. L. (ed.) Meaning and prototypes: totypes: studies in linguistic categorization. London: Routledge, 382-402.
↑ Lehrer, Adrienne. 1990. Prototype theory and its implications for lexical analysis, in Tsohatzidis, S. L. (ed.) Meaning and prototypes: studies in linguistic categorization. London: Routledge, 368-381.
↑ Rosch, E. and Mervis, C. 1975. Family resemblances: studies in the internal structure of Categories, Cognitive Psychology 7: 573-605.
↑ Riemer, N. (2010). Introducing semantics. Cambridge University Press. 7.1.4.5

[4] 順帶一提，喺英文行話當中，prototypical 係形容詞，指似原型嘅噉解。

[5] 留意呢段字出自一份喺 1975 年出版嘅文。

[13] 講緊由一般人嘅角度睇。一般人唔係雀鳥方面嘅專家。

[14] 有關聯想呢樣嘢要點驗證，可以睇吓促發效應嘅嘢。

[6] ↑ analog

[7] rototype

[1] Coseriu, E., Willems, K and Leuschner, T, (2000) Structural Semantics and 'Cognitive' Semantics, in Logos and Language.

[2] Rosch, E. (1975). Cognitive representations of semantic categories. Journal of experimental psychology: General, 104(3), 192.

[3] 原版段英文係噉嘅："when describing categories analytically, most traditions of thought have treated category membership as a digital, all-or-none phenomenon. That is, much work in philosophy, psychology, linguistics, and anthropology assumes that categories are logical bounded entities, membership in which is defined by an item's possession of a simple set of criterial features, in which all instances possessing the criterial attributes have a full and equal degree of membership. In contrast, it has recently been argued ... that some natural categories are analog and must be represented logically in a manner which reflects their analog structure."

[8] Coleman, L., & Kay, P. (1981). Prototype semantics: The English word lie. Language, 57(1), 26-44.

[9] Khalidi, Muhammad Ali. 1995. Two concepts of concept, Mind and Language 10: 402-422.

[10] Riemer, N. (2010). Introducing semantics. Cambridge University Press. 7.1.4，"members of the category MEAL will have very different attributes if the context is a hospital, a wedding banquet, a camping trip or the family dinner table."

[11] Ibbotson, P., & Tomasello, M. (2009). Prototype constructions in early language acquisition. Language and Cognition, 1(1), 59-85.

[geer2006-12] 8.0 ^8.1 Geeraerts, D. (2006). Prototype theory. Cognitive linguistics: Basic readings, 34, 141-165.

[15] Riemer, N. (2010). Introducing semantics. Cambridge University Press，6.1.3 佢呢度仲有提到，要求受試者列出雀鳥嘅例子嗰陣，佢哋比較大機會列出似原型嘅例子，而且佢哋有可能講出企鵝嚴格嚟講係一種雀鳥噉嘅句子。

[16] Kamp, H., & Partee, B. (1995). Prototype theory and compositionality. Cognition, 57(2), 129-191.

[17] Cruse, D. A. 1990. Prototype theory and lexical semantics, in Tsohatzidis, S. L. (ed.) Meaning and prototypes: totypes: studies in linguistic categorization. London: Routledge, 382-402.

[18] Lehrer, Adrienne. 1990. Prototype theory and its implications for lexical analysis, in Tsohatzidis, S. L. (ed.) Meaning and prototypes: studies in linguistic categorization. London: Routledge, 368-381.

[19] Rosch, E. and Mervis, C. 1975. Family resemblances: studies in the internal structure of Categories, Cognitive Psychology 7: 573-605.

[20] Riemer, N. (2010). Introducing semantics. Cambridge University Press. 7.1.4.5

[1]

[2]

[3]

[註 1]

[註 2]

[歐 1]

[歐 2]

[4]

[5]

[6]

[7]

[8]

[註 3]

[註 4]

[9]

[10]

[11]

[12]

[13]

[14]