Zhou Zhimin XX4W85UK Xin Gao Assignment 2 PDF

Preview

Summary

This article starts with the definition comparison of Chinese character recognition with English word recognition to develop the rationale of adopting the Laberge-Samuels model to analyse the cognitive process Chinese character recognition. Subsequently, these Chinese character recognition difficult...

Description

ZHOU Zhimin

14336079

Chinese character recognition for English-speaking learners of Chinese as a Foreign Language: difficulty analysis

Course: MA in TCSOL

Module: XX4W85 - Understanding L2 Chinese Learning & the Learner

Marking tutor: Dr Xin Gao

Word Count: 4062

Table of contents

1. Introduction……………………………………………………………………………….…...1

2. The definition of Chinese character recognition…………………………….…..…1

3. Difficulty analysis from orthographic perspective of Chinese language.…2 3.1 Representational unit of Chinese characters…………………………………....2 3.2 Orthographic depth of Chinese characters………………………………….…...3

4. Difficulty analysis from cognitive process perspective………...…...4 4.1 Visual perception process of Chinees characters………………………….….…5 4.2 Meaning activation process of Chinese characters………………………..….…7

5. Conclusion…………………………………………………………………………………..…10

References…………………………………………………………………………………….……12

1. Introduction It is reported that different levels of English-speaking learners of Chinese as a foreign language (CFL) all experience difficulties in Chinese character recognition (Everson, 1998; Shen, 2005; Hu, 2010; Sung and Wu, 2011). Shen (2010a) claims that CFL students struggle to differentiate Chinese characters in same or similar structures. Everson (2011) reveals that English learners feel difficult to work out meaning of Chinese characters. Hu (2010) claims anglophone learners feel memorization Chinese characters is a challenging task.

This article starts with the definition comparison of Chinese character recognition with English word recognition to develop the rationale of adopting the Laberge-Samuels model to analyse the cognitive process Chinese character recognition. Subsequently, these

Chinese

character

recognition

difficulties

are

discussed

in

details

from

orthographic perspective of Chinese language in the aspects of complex hierarchical structure, orthographic rules and orthographic depth. The Laberge-Samuels model is then used to examine these difficulties during the cognitive process of CFL learners. Finally, some suggestions are made from the perspectives of orthographic knowledge and cognitive skills to tackle these difficulties.

2. The definition of Chinese character recognition Everson (1998) describes Chinese character recognition as elicitation pronunciation and lexical meaning from the written form of Chinese characters. Shen (2005) explicates it as a process how phonetic, graphic and semantic properties are activated to lead to character recognition when a character is visually presented. Regarding word recognition in English, Neuhaus, Roldan, Boulware-Gooden and Swank (2006) state it as a unitization process of phonological, orthographic, and semantic representations of a word. Grabe (2009, p.101) gives the definition in a dynamic way as “an interactive process of accumulating phonological, orthographic, and semantic (and possibly syntactic

and

morphological)

information

(through

increasing

activation).”

By

comparing the definitions of Chinese character recognition with the definitions of word recognition in English, three commonalities can be exhibited. Firstly, both include three of the same linguistic constituents: phonological element, orthographic element and semantic element. Secondly, an activation process is mentioned in both definitions. Thirdly, the three linguistic constituents are all engaged in this process. The only distinction in this comparison is the different forms of orthographic element: Chinese character and English word. This is caused by drastically different orthographic systems of these two languages. Although Chinese character and English word look so different,

1

they are still the written form or print of these two languages. Hence, it is reasonable to apply English word processing theory to Chinese character recognition.

3. Difficulty analysis from orthographic perspective of Chinese language 3.1 Representational unit of Chinese characters The complex hierarchical structure of Chinese characters brings one of major challenges to learners (Shen, 2005). First of all, it is essential to probe how Chinese characters construct with their constituents. There are three levels of hierarchical structure in Chinese writing systems: strokes, radicals, characters.

Structurally, strokes form

radicals, and radicals form characters in diverse configurations. A character is comprised of 1 to 30 strokes from a total of 28 distinctive types of strokes (Huang and Liao, 1981, cited in Shen, 2005). Based on different character pools and decomposition principles, the total number of radicles varies from 567 to 648 (Fu, 1985; Han 1994; Li et al., 1988, cited in Li and Chen, 1997) *. Strokes constitute an integral character or a radical; two or more radicals can be further combined into a compound character.

This hierarchical structure of Chinese character is confined in a square shape, so it is difficult for learners to find the starting point of the whole structure, especially when there are a lot of strokes in one character (Hoosain and Siok, 1996 cited in Law, Ki, Chung, Ko, and LAM, 1998). For example, it is very challenging to work out the first stroke to write in the characters of 戴 , 禽 and 奏 without good orthographic knowledge. Moreover, the square shapes that Chinese characters are confined in has the same size. Hence, the more strokes in a character, the more visual information contains in the same-size square shape. It can be easily seen by looking at two characters like 工 and 舞.

Lin and Akamatsu (1997) claims the characters that have higher stroke density

demand more attention from readers.

Peng, Li and Yang (1997) complement that besides the complex hierarchical structure, orthographic rules play a pivotal role in Chinese character construction.

That is, the

orthographic rules should be followed when components are used to develop a genuine and legitimate Chinese character. A series of experiments conducted by Peng, et al. (1997) prove radical position is one of the orthographic rules that has a profound impact on character identification and the radical positions in characters are not arbitrary. It is found in horizontal structure characters, some radicals only exist on the left side, such as 亻 and 王; some merely show up on the right side, such as 尔 and 反 , while some appear on either side, such as 工 and 口 (Li and Chen, 1997). Xiao (2009) adds that violation of fixed radical positioning results in erroneous variations or non-

2

characters. It is asserted that configurational structure of character is another key orthographic rule since random combination of strokes and radicals cannot form a right character (Peng, et al., 1997; Yeh, Li, Takeuchi, Sun, and Liu, 2003;). Xiao (2009) expresses the same opinion by treating character configuration as one of key elements of orthographic feature. For example, the character 资 consists of three radicals: 冫, 欠, 贝 . There are more than 20 different ways to put these three radicals together, while there is only one way to make an acceptable character. Therefore, Yeh, et al. (2003) categorize configurational structure of character as “linguistic” property instead of “figural” property.

On top of the complex hierarchical structure and orthographic rules, the subtle differences among similarly shaped characters make Chinese character recognition even harder for learners, especially beginner learners due to their unfamiliarity of strokes-based writing (Shen, 2013). Law, et al. (1998) mention that some characters look very similar and they have just one different stroke. This phenomenon can be well illustrated by 孑 and 子 . If the horizontal stroke 一 in 子 is replaced by an upward slant stroke ㇀ , it becomes another character 孑 . Law, et al. (1998) also take another fact into account that a slight variation in how strokes intercept each other can produce totally different but very lookalike characters. For example, in the character 工 , the vertical stroke does not cross the upper horizontal stroke; while if it does, it becomes another character 土 .

It is also noticed that the comparative length of strokes also

needs to be considered to differentiate similar characters. For instance, the upper stroke in the character 土 is shorter than the bottom stroke; but other way around, it generates another character 士.

3.2 Orthographic depth of Chinese characters Learning Chinese characters is reported as an exceptionally difficult task for anglophone learners as some orthographic knowledge of Chinese is a prerequisite to work out the relationships among pronunciation, shape and meaning of Chinese character (Everson, 2011). In Chinese orthography, Chinese characters map onto morphemes instead of phonemes, so there is no direct correspondence between phonology and writing system (Shen, 2005; Xiao 2009; Everson, 2009). Sung and Wu (2011) makes further explanation that in view of being logographic in nature, Chinese character represents a meaning with little access to its pronunciation. In relation to word recognition attributed to phonological representation, “orthographic depth” is an index to indicate the extent to which orthography can regularly and systematically represent letter-to-phoneme correspondence (Liberman, Liberman, Mattingly and Shankweiler, 1980, cited in

3

Everson, 1998). Chinese is considered as a deep orthographic system, while English is somewhat in between (Everson, 1998; Wong 1998; Shen, 2005).

Wang (1998) summarise that representational units and orthographic depth are two main differences between Chinese and English writing systems. The same conclusion is also drawn by Shen (2005). It is hard for English-speaking CFL learners to grasp a quick and precise visual perception of Chinese characters which have three-tier structures in a square shape with diverse constituents and subtle intra-character differences. Therefore, Lin and Childs (2010) stress the importance of sufficient figural memory for visual perception of Chinese characters. Moreover, unsystematic and irregular script-to-sound correspondence adds extra hindrance to the final access to the semantic meaning. Accordingly, Shen (2005) highlights the orthographic knowledge underpins the various strategies during cognitive process of characters. Because Chinese writing system is considerably different from English in nature, Englishspeaking CFL learners cannot transfer all their English reading strategies to Chinese characters learning, which makes Chinese character recognition even more challenging for them (Everson, 2009).

There also exist several other difficulties in Chinese character recognition. There are many Chinese characters sharing the same sound but having different meanings; consequently, the high number of homophones adds difficulty in accurate recognition of Chinese characters (DeFrancis, 1989; Shen 2010b; Sung and Wu, 2011). Shen (2005, 2008) indicates that as a result of multiple meanings of many Chinese characters, learners feel struggled to decide the right meaning of Chinese characters during reading.

Unlike English, all Chinese characters are equally spaced. Accordingly, the

lack of visible boundaries between Chinese words is a significant obstacle for Englishspeaking learners in comprehending words and processing text (Everson, 2011). However, this article will only focus on the difficulties of Chinese character recognition caused by differences in representational units and orthographic depth.

4. Difficulty analysis from cognitive process perspective Among so many word processing theories, the Laberge and Samuels model (hereinafter referred to as the LS model) with its regular revisions has a long and useful life (Samuels, 1994; Samuels, 2006). It has still been referred in books and journals and is also used as foundation of contemporary reading theories, such as Dual Coding Theory. (Sadoski and Paivio, 2007; Sadoski, Mctigue, and Paivio, 2012). As discussed in section 2, due to the commonalities between Chinese character recognition and English word recognition, it seems rational to use word processing

4

theory to analyse Chinese character recognition. The LS model will be used here to investigate the above-mentioned difficulties from cognitive process perspective.

4.1 Visual perception process of Chinees characters Visual perception process is the start of word recognition process. Model of Grapheme Learning by LaBerge and Samuels (1974) breaks the visual perception process into four hierarchical stages. In stage 1, features of the smallest element in the written form are initially processed (LaBerge and Samuels, 1974). In an English word, they are features of the shape of a letter, such as lines, intersections, angles, curvatures along with relative positional information such as left, right, up, down (Rumelhart, 1970). Selectivity is needed in this stage to choose a subset of appropriate features from all features to discriminate letters successfully; selectivity is defined as one characteristic of internal attention in the LS model (Laberge and Samuels, 1974). To illustrate, backward hook can be selected to discriminate letter “i” and “j”. In stage 2, a letter code is constructed based on the relevant analysed features (ibid.). This is followed by stage 3 where letter codes activate spelling-pattern codes and in stage 4 spellingpattern codes lead to word codes (ibid.). During this process, alertness, as another defined characteristic of internal attention in the LS model is also required (ibid.). Alertness can also be named as sensitivity that is used in vigilance tasks (Laberge and Samuels, 1974; Samuels, 1994). Cognitive psychologists Posner and Boies (1971) define alertness as “the ability to develop and maintain an optimal sensitivity to external stimulation” (p. 391). The alertness to positional information can be named as positional sensitivity and the alertness to internal structure in Chinese character can be termed as intra-character sensitivity.

Stroke is the smallest element in Chinese character, so the visual perception process in Chinese character starts from processing stroke features. Stroke features are detected, analysed and synthesized into stroke codes. By associating with related features, the activation passes up to the radical level until finally reaches character level. Coincidentally, Taft, Zhu and Peng (1999) delineate this hierarchical activation process of Chinese character visual perception in the same way without using Model of Grapheme Learning.

Now the impact of orthographic difference on visual perception

process of Chinees characters will be examined in these 4 stages. In stage 1, stroke features are detected as individual visual shapes, such as lines, hooks, turning points, dots. Many of them do not exist in English letters due to different writing systems. Lin and Akamatsu (1997) declare Chinese characters are distinguished by more visually distinctive features than English letters are. The unfamiliarity of stroke features results in difficulty in selecting a unique set of codes to activate a stroke code

5

in stage 2. The research conducted by Shen (2010a) discloses English-speaking CFL learners face great difficulty in distinguishing strokes due to their insensitivity to subtle differences between similar strokes. When strokes synthesise radicals in stage 3, positional information play a much more important role than when letters activate spelling patters in English. Some experiments implemented by Taft, et al. (1999) prove the significance of positional sensitivity in Chinese character recognition. Letters are arranged from left to right in a linear shape to form spelling patterns in English words. As strokes are assembled in various ways to structure a radical or an integral character, certain stroke positional constraints need to be adhered to (Sung and Wu, 2011). A typical example is three different characters: 人, 八, 入; they are composed of two same strokes arranged in different positions (Li and Chen, 1997).

More importance is attached to positional information as it also exists in stage 4 where radical codes are processed into character codes. The transposed radicals research by Taft et al. (1999) uncovers the significance of radical positional information in Chinese character recognition. The studies conducted by Fang and Wu (1989) and Taft and Zhu (1997) draw the same conclusion. For instance, 杏 and 呆 are composed of two identical radicals 口 and 木 in two different positions (ibid.). Li and Chen (1997) cite another example with characters of 柑 and 某 . This positional information is reflected in the above-mentioned orthographic rules in Chinese characters construction.

The positional information in both stroke and radical level calls for sensitivity development of Anglophone learners. Intra-character sensitivity is one of the cognitive skills required due to unique orthographic features of Chinese (McLaughlin, 1987, cited in Xiao, 2009). Likewise, Lin and Childs (2010, p. 258) name it as “sensitivity to visual structures”. Shen (2013) manifests the similar opinion that in order to identify the shape and the structure of characters precisely, an intra-character sensitivity need to be developed, especially when some Chinese characters look very similar.

Regarding intra-character sensitivity, Yeh, et al. (2003) reveal one interesting fact that English-speaking learners who simply view Chinese characters a visual pattern or an object before they start to learn Chinese. This point is reiterated by Shen (2013) that these learners often tend to encode Chinese character structure with their familiar codes, such as English letters; Shen continues the explanation with two examples: the character 印 is recalled as “EP” by English-speaking learners and the character 糊 as “KON”. This difficulty in mastering precise visual perception of Chinese characters can be explained by language transfer theory. Koda (2008, cited in Shen 2013) interprets

6

that learners use linguistic resources and cognitive patterns available from their first language to learn a second language. Yeh et al. (2003) allege that these leaners tend to activate the pattern recognition process when they see Chinese characters, which is helpful to extract simple or larger structural components but causes ignorance of stroke details and small feature differences of smaller components.

Finally, it is worth to state that Chinese orthographic system requests more visual coding processing than alphabetic system due to lack of the regularity in sound-symbol correspondence (Wang, 1998). Considering the lack of a phonological assembly path that is available in English, Chinese characters need to be visually accessed first in order to obtain their pronunciation (Everson, 1998). Hence, visual perception plays a predominant role in Chinese character recognition.

4.2 Meaning activation process of Chinese characters There are various routes in the LS model via which a word in print takes the process towards its meaning. Only three processing routes will be discussed here, because not every route in the LS model can be applied to Chinese character meaning activation due to the unique linguistic features of Chinese.

In the first route, the graphemic stimulus is coded into a visual word code which activates the phonological code; subsequently, the phonological code activates the meaning code (LaBerge and Samuels, 1974). Findings by Tan, Hoosain and Siok (1996) prove the existence of this route...