The Attention System Of Language

Published Date: 02 Nov 2017

Abstract

The interrelationships between attention and language are addressed both in the field of psychology and linguistics. The American cognitive scientist Leonard Talmy’s ongoing Attention System of Language is a pioneering project to systematically outline the attentional factors at various linguistic levels. Currently, the examples in Talmy’s framework are mainly taken from spoken English. The present study aims to examine how the 11 linguistic attentional factors in Category A (Factors involving properties of the morpheme), which is the most mature and developed one among the nine categories of the approximately 100 factors thus far identified, can be applied to written Chinese. The two methods used are introspection and empirical survey. 100 literate native Chinese speakers are invited to participate in a survey containing 11 qusetions. Statistical results show that while the intuitive judgments by the paticipants are deviating in Q4 (Factor Ab2), Q5 (Factor Ad1), and Q7 (Factor Ab3), the frequency distributions of the answers to the rest of the eight questions are in agreement with the assumptions by the attentional factors. Possible motivations for the deviations are disccused in detail. This study provides cross-linguistic and cross-modality empirical data for Talmy’s attention system and is the first step of establishing a complete attention system for written Chinese.

Key Words: Attention; language; written modality; Chinese

1. Introduction

A thousand psychologists may have a thousand definitions for attention. What they do agree on, however, is the limitedness of attention capacity and the selectiveness of attention processes (Deutsch & Deutsch, 1963; Dukas, 2002; Duncan, 1998; Koch & Tsuchiya, 2006).

Attention, according to the Father of American Psychology and the Founder of Western Scientific Psyschology William James (1890), "is the taking possession by the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought" (James 1890: 256). It is "the allocation of cognitive resources among ongoing processes" (Anderson 2004: 422) and arises out of our limited capacity for processing information (Strayer & Drews 2007: 129). The attention mechanism is crucial for the survival of animals and human beings in the complex world (Kudson 2007: 58). By selective attention, the load on the limited-capacity cognitive systems is reduced and the irrelevant information for the stimulus stream is filtered out (Downing 2000: 467), which "plays a very important function in ensuring the mental resources are not unnecessarily expended" (Alais, Morrone & Burr 2006: 1339).

Several studies have been conducted concerning the interrelationships between attention and language in the field of psychology and linguistics, in particular cognitive linguistics, which is aimed at ascertaining "the global integrated system of conceptual structuring in language" (Talmy 2011: 623). While some psychological studies (e.g. Fischler, 1998; Lambrecht, 1994; Robinson, 1995 & Tomlin, 1995) examine attention as an object external to language, cognitive linguists including Langacker (1984, 1987, 2008), Slobin (1996, 1997, 2003), Talmy (1978a, 1978b, 1996, 2000), and Tyler & Evans (2003) hold a different perspective on the role of attention in language, namely, "language itself acts as an attention-directing mechanism" (Taube-Schiff 2004: 20).

The present study, taking Talmy’s ongoing linguistic attention system as the basic theoretical framework, belongs to the larger domain of cognitive linguistics, which adopts a conceptual appoach to language, in differentiation from the formal and psychological approaches. This conceptual approach "is concerned with the patterns in which and the processes by which conceptual content is organized in language" (Talmy 2000: 2). The focal point in this particular study is the human cognitive mechanism of attention in relation to grammatical and conceptual structures in language.

To be specific, the three central questions that guide the current thesis are:

Firstly, Talmy’s Attention System of Language is mainly based on English, without mentioning Chinese, a typical representative of the Sino-Tibetan language family. The linguistic attentional factors are universal. But since Talmy’s attention framework is still in its infancy, the factors haven’t yet received substantial support from a wide range of languages, especially languages that belong to various languages families thus far classified by typological studies. Finding out how the factors of linguistic attention perform differently in different languages can provide new insights into the nature of linguistic attention. The present study is interested in mapping the Attention System of Language from English into Chinese. Thus the first research question is formulated as follows:

RQ 1: How are the factors of the Attention System of Language manifested in Mandarin Chinese?

Secondly, Talmy discusses the linguistic attentional factors in spoken and sign languages (see Talmy, 2010a, 2010b), leaving the written modality yet unexplored. This neglect of written language in linguistic research can be traced back to "the Father of Modern Linguistics" Ferdinand de Saussure’s Cours de Linguistique Générale (1916). Saussure’s "structuralism" is sometimes also referred to as "phonocentrism", which takes spoken language as symbols of meaning, and further, written language merely as symbols of spoken language, i.e., the written modality is the symbols of symbols, thus putting the written subordinate and inferior to the spoken. This idea has been challenged by some scholars (e.g. Nunberg, 1990). In particular, hieroglyphic writing systems feature a much looser connection among the written symbols, meaning and pronunciation. Hereby the second research question is formulated as:

RQ2: How can Talmy’s attentional factors be applied to written language, in particular written Chinese?

Thirdly, Talmy’s focus is on auditory perception (see Talmy, 2010a, 2010b), with no detailed comparison with visual perception, which is the major way for human beings to access information in the world. When people communicate with one another orally, auditory attention is in play. However, when people read, visual attention is what guides us to certain parts of a message. The shapes of characters or letters, the font sizes, the colors of the reading materials or even the use of punctuation marks can jointly direct our visual attention (see Lampert, 2009). Though Talmy explicitly mentions the role of the volume, pitch and timber of speaking in leading hearers’ attention, he hasn’t yet systematically talked about how these auditory factors correspond to visual factors (see Talmy, 2010a, 2010b). Thus the third research question is:

RQ3: What extra-linguistic visual attentional factors can be identified in written Chinese?

Thus, the current study aims to examine the cross-lanugage, cross-modality, and cross-sensory-system applicability of the Attention System of Language proposed by Talmy with empirical data from written Mandarian Chinese.

The present thesis investigates the implementaion of linguistic attentional factors in written Mandarin Chinese. The attentional factors under dicussion are limited to the 11 factors in Category A (Factors involving properties of the morpheme), which is by far the most mature and fully developed category in Talmy’s Attention System of Language (Lampert 2012: personal communication). As Talmy’s attentional framework is yet to be completed, few systematic studies have been published in the Attention System of Language, apart from the German scholar Martina Lampert’s Attention and Recombinance (2009). This systematic study of English morphological compositionality from the innovative perspective of linguistic attentional factors using the EMO Network which contains 561 words and phrases that have EMO as base provides the most detailed analyses of the attentional factors with English examples up to date.

Referring to Lampert (2009), a WATER Network inspired by Chen (2010) containing 399 Chinese characters with the WATER radical (水，冫and氵) is established according to A New Century Chinese-English Dictionary (2003) edited by Hui Yu. (It should be noted that the radical灬 , though bearing seemingly resemblance to the water radical and has been mistakenly called "four drops of water", is more related to heating and fire, e.g. 煮 and 蒸, rather than water.) The two lexical networks, from which most of the words/phrases or characters illustrating the factors are taken, are attached in Appendices A2 and A3. Comparisons between English and Chinese are more feasible when the examined objects are controlled within two lexical networks.

Due to the theoretical nature of the research questions and the current lack of empirical data in the study of linguistic attentional factors, the two methods used in the present study are intropsection and empirical survey. These two methods are justified in terms of both necessity and feasibility.

As early as in 1890, William James states in his The Principles of Psychology, "introspective observation is what we have to rely on first and foremost and always" (James 1890: 185). Talmy (2000) writes that "the use of introspecton must be recognized as an appropriate and arguably necessary methodology in cognitive science, together with the other generally accepted methodologies" (Talmy 2000: 6). A critical review of the existing literature on the topic of attention and language, an evaluation of their relevance to the study, and an in-depth analysis of the results and implications all require introspective thoughts.

Along with introspection, empirical survey in the form of questionnaire is the complementary research method employed. The survey is conducted over a one-month period from February 1st, 2013 to February 28th, 2013. 11 multiple-choice questions are designed respectively in correspondence to the 11 factors in Category A (the complete questionnaire and the translated version are attached in the Appendices A4). All the characters under examination are taken from the WATER Network. After a pilot study with 25 people, 100 native Chinese speakers with a high level of literacy have participated in the survey. The collected data are then analysed with the statistical software SPSS. This empirical research pioneers the attention system study in Mandarin Chinese.

Results show that the majority of the frequency distribution charts of the answers agree with the assumptions of Talmy’s attentional factors, though some are deviating. Tapping into the causes of these deviations may reveal which of the linguistic attentional factors are language-dependent and which are independent. The present thesis provides empirical data from written Mandarin Chinese for Talmy’s Attention System of Language and examines the cross-language, cross-modality and cross-sensory-system applicability of the linguistic attentional factors.

The whole thesis is divided into seven chapters. Following Chapter One Introduction is an inclusive review of past studies under the topic of attention and language, with special focus on Talmy’s Attention System of Language and Lampert’s systematic application of this linguistic attentional framework. The three research questions guiding the present study (namely, 1) how the factors of the Attention System of Language are manifested in Mandarin Chinese; 2) how Talmy’s attentional factors can be applied to written language, in particular written Chinese; and 3) what extra-linguistic visual attentional factors can be identified in written Chinese) are explicated in Chapter Three. Chapter Four explains why and how the methods of introspection and empirical survey are adopted in this study. The core of the thesis Chapter Five analyzes the eleven factors of Category A one by one, first with theoretical reasoning and then with empirical data. The significance and limitations of the present study, as well as the directions for future studies are discussed in depth in Chapter Six. The thesis closes with the conclusive Chapter Seven. For better readability, each chapter begins with an introduction and ends with a summary.

2. Literature Review

2.1 Introduction

2.2 Outside the Attention System of Language

In recent decades, a lot of studies have been done concerning the interrelationships between attention and language (e.g. Eviatar, 1998; Fischler, 1998; Myachykov, Garrod & Scheepers, 2009; Nozari & Dell, 2012: Segalowitz, 2001). Some studies focus on one particular aspect of language in relation to attention, for instance, sentence form (Lambrecht, 1994), word order (Nappa et al., 2004; Tomlin, 1995), or grammar (Myachykov, Tomlin & Posner, 2005; Taube-Schiff & Segalowitz, 2004a, 2005a; Tomlin, 1997), while some other studies are particularly interested in the role of attention in first- or second- language acquisition and production (Al-Hejin, 2005; Chung & Segalowitz, 2003, 2004; Duncan, 2010; Leow, 1997; Levelt, 1989, 1995, 1999; Meuter & Allport, 1999; Robinson, 1995, 2001, 2003; Schmidt, 1993, 2001; Segalowitz, 2000; Segalowitz & Frenkiel-Fishman, 2005; Taube-Schiff, 2004; Taube-Schiff & Segalowitz, 2004b, 2005b; Tomlin & Villa, 1994). The above studies could be summarized into the following table chronologically, from which the development of researches into attention & language could be seen clearly, in spite of its partial inclusiveness of existing studies. Several important studies are reviewed in detail afterwards.

Table 01: Summary of past studies in attention & language

No.

Author(s)

Title

Year

01

Levelt

Speaking: from intention to articulation

1989

02

Schmidt

Awareness and second language learning

1993

03

Lambrecht

Information structure and sentence form: a theory of topic, focus, and the mental represenations of discourse referents

1994

04

Tomlin & Villa

Attention in cognitive science and second language acquisition

1994

05

Levelt

The ability to speak: from intentions to spoken words

1995

06

Robinson

Attention, memory and the "noticing" hypothesis

1995

07

Tomlin

Focal attention, voice, and word order

1995

08

Leow

Attention, awareness, and foreign language behavior

1997

09

Tomlin

Mapping conceptual representations into linguistic representations: the role of attention in grammar

1997

10

Eviater

Attention as a psychological entity and its effects on language and communication

1998

11

Fischler

Attention and language

1998

12

Levelt

Language production: a blueprint of the speaker

1999

13

Meuter & Allport

Bilingual language switching in naming: asymmetrical costs of language selection

1999

14

Segalowitz

Automaticity and attentional skill in fluent performance

2000

15

Robinson

Cognition and second language instruction

2001

16

Schmidt

Attention

2001

17

Segalowitz

On the evolving connections between psychology and linguistics

2001

18

Chung & Segalowitz

A language-specific form of attention underlies L2-proficiency

2003

19

Robinson

Attention and memory during SLA

2003

20

Chung & Segalowitz

Language-specific grammatical attention in second language proficiency

2004

21

Nappa et al.

Paying attention to attention: perceptual priming effects on word order

2004

22

Taube-Schiff

Studies in linguistic attention control

2004

23

Taube-Schiff & Segalowitz

Linguistic attention control: attention-shifting governed by grammatical elements of language

2004a

24

Taube-Schiff & Segalowitz

Within-language attention control and second language processing

2004b

25

Al-Hejin

Attention and awareness: evidence from cognitive and second language acquisition research

2005

26

Segalowitz & Frenkiel-Fishman

Attention control and ability level in a complex cognitive skill: attention shifiting and second-language proficiency

2005

27

Taube-Schiff & Segalowitz

Linguistic attention control: attention shifting governed by grammaticized elements of language

2005a

28

Taube-Schiff & Segalowitz

Within-language attention control in second language processing

2005b

29

Myachykov, Tomlin & Posner

Attention and empirical studies of grammar

2005

30

Myachykov, Garrod & Scheepers

Attention and syntax in sentence production: a critical review

2009

31

Duncan

The effects of aging and bilingualism on language-specific attention control

2010

32

Nozari & Dell

Feature migration in time: reflection of selective attention on speech errors

2012

Schmidt (1990) proposes a Noticing Hypothesis that "SLA is largely driven by what learners pay attention to and notice in target language input and what they understand the significance of noticed input to be" (Schmidt 1990: 135). Schmidt (2001) expands this hypothesis and reviews studies that stress the role of attention in cognitive accounts of second language development. These studies regard attenton to input as "essential for storage and a necessary precursor to hypothesis formation and testing" (Schmidt 2001: 8). Two popular ideas held in common by the reviewed studies in Schmidt’s 2001 paper are: 1) "L2 learners process target language input in ways that are determined by general cognitive factors including perceptual salience, frequency, the continuity of elements, and other factors that determine whether or not attention is drawn to them"; and 2) "attention is what allows speakers to become aware of a mismatch or gap between what they produce and what proficient target language speakers produce" (Schmidt 2001: 15-16). What Schmidt focuses on is how voluntary attention of the L2 language learners enhances the efficiency of learning.

Myachykov, Garrod & Scheepers (2009) holds that the perceptual information gained through the visual system of human beings must be filtered by the attention system first, so as to "sort information for processing as more or less noticeable, important, or relevant", thus resulting in "linguistic processing associated with a relative degree of salience" (Myachykov, Garrod & Scheepers 2009: 2). The assumptions are that 1) "operations originating from distinct attentional networks have their counterparts in linguistic behavior or even overall organization of the language faculty" (Myachykov, Garrod & Scheepers 2009: 3) and that 2) "attentional processing of the cognized world may somehow be reflected in the organization of both production and comprehension of sentences" (Myachykov, Garrod & Scheepers 2009: 6). Even though this study takes a Chomskyan perspective and believes in the existence of an independent faculty for language in the human brain, it’s worth noting that this psychological research assumes a biological basis for linguistic attention.

Levelt (1989) states that "using a particular language requires the speaker to think of particular conceptual features" (Levelt 1989: 71) and that in language learning the child is sure to realize that "the language requires him to attend to certain perceptual or conceptual featuers when he encodes a message" (Levelt 1989: 104-105). Levelt emphasizes the particularity of attention-directing mechanisms of a language, which reflects the living environments of its speakers and maintains that language shapes our perception of the world in a subtle way.

Langacker (1987) takes attention to be "intrinsically associated with the intensity or energy level of cognitive processes, which translates experientially into greater prominence or salience" (Langacker 1987: 115-116). He suggests that language can guide speakers to make particular focal adjustment, which includes three aspects: 1) selection, 2) perspective, and 3) abstraction. Further, he explains that selection determines "which facets of a scene are being dealt with"; perspective is concerned with "the position from which a scene is viewed, with consequences for the relative prominence of its participants"; abstraction is related to "the level of specificity at which a situation is portrayed" (Langacker 1987: 117). In addition, Langacker (1987) states that "the prevelance of figure/ground organization in conceptual structure entails its importance for semantic and grammatical structure as well" (Langacker 1987: 120).

Langacker (2008) writes that

through linguistic expressions, we access particular portions of our conceptual universes. The dimension of construal referred to here as focusing includes the selection of conceptual content for linguistic presentation, as well as its arrangement into what can be broadly described (metaphorically) as foreground vs. background (Langacker 2008: 57).

Another notion related to attention that is introduced by Langacker (2008) is "prominence". It is a conceptual phenomenon that inheres "in our apprehension of the world, not in the world per se", which means "how prominent a particular entity is—whether it functions as profile, trajector, landmark, or none of the above—depends on the construal imposed by the linguistic elements employed, in accordance with their conventional semantic values" (Langacker 2008: 72-73).

Talmy (2000) points out that distribution of attention "consists of various patterns of different strengths with which one’s attention is directed over a referent object or scene in accordance with the specifications of closed class forms" (Talmy 2000: 76). Distribution of attention, along with configurational structure, location of perspective point, and force dynamics, constitute the main schematic systems (Talmy 2000: 12). The schematic system of attention covers three different patterns of attentional distribution—1) the "center-periphery" pattern (which includes "Figure-Ground" Organization), 2) the "level of attention" (in which "linguistic forms direct one’s greatest attention either to a level of the component element making up some aspect of a referent situation, or to the level of the whole that encompasses those components"), and 3) the "windowing" of attention, in which "linguistic forms can differentially direct greatest attention to, or withdraw attention from, particular portions of a referent situaton" (Talmy 2000: 13).

There are three factors that govern the distribution of attention over a referent scene in the attentional system, namely, 1) the strength of attention (ranging from faint to intense), 2) pattern of attention ("by which attentions of different strengths are combined and arranged in particular patterns"), and 3) mapping of attention ("by which the particular parts of an attentional pattern are mapped onto particular regions of the referent scence") (Talmy 2000: 76-77). In his later Attention System of Language, Talmy elaborates on these factors in more detail and the current thesis is mainly based on this ongoing study which will be introduced in the following section.

Slobin (1996, 1997, and 2003) proposes that certain specific grammatical elements of language are responsible for the attention-directing mechanism. Slobin (1996), drawing on studies by Humboldt (1836), Whorf (1940, 1942) and Boas (1938), discusses linguistic relativity in view of linguistic attention and concludes that,

The language or languages that we learn in childhood are not neutral coding systems of an objective reality. Rather, each one is a subjective orientation to the world of human experience, and this orientaion affects the ways in which we think while we are speaking (Slobin 1996: 91).

Slobin (2003) holds that a certain utterance is "never a direct reflection of ‘objective’ or perceived reality or of an inevitable and universal mental representation of a situation" (Slobin 2003: 157), for the fact that "the same situation can be described in different ways", either in any given language or across languages (Slobin 2003: 157).

These cognitive linguists share the perspective that "language itself acts as an attention-directing mechanism" (Taube-Schiff 2004: 20), which is realized in two identified ways: 1) "language affects the way in which one attends to the world by shaping one’s mental representations and communicative interactions"; and 2) "grammaticized elements of language will direct an individual’s attention to relationships that exist within the linguistic message" (Taube-Schiff 2004: 23-24).

2.3 Within the Attention System of Language

Talmy may not be the first cognitive scientist in the world who has noticed the inter-connectedness between attention and language, but he deserves to be credited as the first in the world to have set out "to outline the fundamental attentional system of language" (Talmy 2010a: 1). His upcoming book the Attention System of Language (MIT press) "will be an unprecedented investigation into the general relationship of attention and language" (Lampert 2009: ix). This project "may be the first with the aim of developing a systematic framework within which to place all such prior findings—— together with a number of new findings—— about linguistic attention" and "perhaps the first to recognize that the linguistic phenomena across this whole range do all pertain to the same single cognitive system of attention" (Talmy 2010b: 4).

In a speech situation, a hearer may attend to the linguistic expression produced by a speaker, to the conceptual content represented by that expression, and to the context at hand. But not all of this material appears uniformly in the foreground of the hearer’s attention. Rather, various portions or aspects of the expression, content, and context have different degrees of salience. Such differences are only partly due to any intrinsically greater interest of certain elements over others. More fundamentally, language has an extensive system that assigns different degrees of salience to the parts of an expression or of its reference or of the context. As for the speech participants, prototypically: the speaker employs this system in formulating an expression; the hearer, largely on the basis of such formulations, allocates her attention in a particular pattern over the material of these domains (Talmy 2010a: 1).

Talmy has listed attention among schematic systems of language "that structure conceptual content and context", along with "configurational structure", "perspective point", "force dynamics" and "cognitive state" (Talmy 2010b: 1). He takes attention as synonymous with consciousness and holds both as gradient (Talmy, 2010b).

Talmy’s attention system of language is aimed at providing "a comprehensive account of how language, at all of its levels, inheres a complex system of attention-directing mechanisms and of how these general attention-directing may affect language" (Lampert 2009: 389). Up till now, nearly 100 linguistic attentional factors have been identified, each of which "involves a particular linguistic mechanism that increases or decreases attention on a certain type of linguistic entity" (Talmy 2010a: 1). The formulation principles of the factors are 1) "greater vs. lesser attention"; 2) "mention of obtent and mechanism"; 3) "contrast types"; and 4) "degree of dividedness" (Talmy 2010b: 5). Firstly, the formulated factors are not on an absolute scale, but instead in contrast with each other in terms of greater or lesser attention; secondly, if a factor governs an attentional object’s salience, then this obtent is mentioned first and the mechanism comes last; thirdly, there are three main contrast types for the factors—a polar or "whether" factor, a circumstantial or "where" factor, and an alternant or "which" factor; fourthly, the above mentioned contrast types can in turn fall into three degrees of dividedness, namely, gradient-dichotomous-hierarchical (Talmy 2010b: 5).

All in all, the currently identified factors fall into 9 categories (the structure of the system from Category A to I is listed in Figure 1 below and the complete attentional framework is attached in Appendix A1 "the Attention System of Language") and under each of these categories, there are sub-categories. The current study focuses on Category A (Factors Involving Properties of the Morpheme).

The nine

categories

in Talmy’s

Attention System of Language

B

C

D

E

F

G

H

I

A

Factors involving properties of the morpheme

Factors involving morphology and syntax

Factors involving forms that affect attention outside themselves

Phonological factors

Factors involving properties of concepts

Factors involving relations among subsystems of language

Factors involving the occurrence of representation

Factors involving temporal progression

Factors involving properties of the speech context

Figure 01: The Nine Categories in Talmy’s Attention System of Language

(based on Talmy 2010a, see A1 in Appendices)

It should be pointed out that these attentional factors, instead of always being static and working alone, are in constant dynamic interaction with one another, thus resulting in "gradation in strength of attention through factor combination", "reinforcement of an attentional pattern through factor convergence", or "override" or "competition" from "factor conflict" (Talmy 2010a: 22-24). It is important not to take attention in the attention system of language as identical to general notions of "attention" in daily use or in other fields of academic studies (e.g. psychology), as "the attentional properties found in language appear to have both commonalities and differences with attentional properties in other cognitive systems" (Talmy 2010b: 3).

2.4 After the Attention System of Language

With Talmy’s linguistic attention factor model as a basic theoretical framework, Martina Lampert’s Attention and recombinance: a cognitive-semantic investigation into morphological compositionality in English (2009) is "the first major application of Len Talmy’s research into attention in language" (Lampert 2009: ix). It is aimed at systematically applying "Leonard Talmy’s ground-breaking ‘factor model’ of linguistic attention to a substantive body of semantically coherent, and in part just recently emerging, authentic language data and not to some selected ‘best examples’ only" (Lampert 2009: 389).

The first chapter is titled "Zeroing in: attention in language". From the second to the sixth chapter, the book is divided into two parts, with the first part (Chapters 2-5) giving "a comprehensive and meticulous summary of Talmy’s cognitive approach to language" (which includes "the main principles inspiring his theory of language", "his Overlapping System of Attention", "his system of basic factors that set strength of attention", and "methodology") and the second part (Chapter 6) presenting "attentional analyses that subsequently integrate more factors needed to account for the increasing complexity of the linguistic target items" (with "a bottom-up, data-driven perspective") (Marchetti 2010: 1). The last chapter (Chapter 7) is named "Converging models: linguistic factors that set strength of attention and binding by synchrony".

The "attentional analyses of individual instances" throughout the book are "drawn from a cross-section of the recently emerging lexical network which has emotion as its morphological (and conceptual) base" (Lampert 2009: 162). Attentional aspects of recombinance on the sub-morphemic level, the level of single morphological constituents (e.g., emotion, -age, and –ship), the "bi-constituential" level (e.g., non-emotion, disemote, and emoticon), the poly-constituential level (e.g., emotion-feeling, EQ, and overemotionalization), the sub-clausal constituential level (e.g., emotional language, emotion in language, and expressing emotion through language), further, the syntactic level, and at last, the discoursal level, have been examined (Lampert 2009: 170-374). Following the analyses are "several case studies manifesting systemic interaction of attention factors (alternatively recombining, converging or competing), to yield patterns of differential salience effects such as gradation of strength, reinforcement or conflict" (Lampert 2009: 162).

Substantive examples from the EMO Network in Lampert’s study will be used in the current thesis, in comparison with characters from the WATER Network.

2.5 Summary

3. Research Questions

3.1 Introduction

3.2 Applicability across Languages

Of central concern in this study is how the Attention System of Language proposed by Talmy can be adapted to written Chinese. It is necessary to identify which of the factors are relevant to the written language and which of them are applicable to Chinese. The current study will identify the Chinese-specific attentional factors based on Len Talmy’s factor model and comparisons will be made between English and Chinese. Not all the identified factors are equally applicable to the languages in the world. Some factors are language-specific. For example, factor Ca1 (a form affecting attention on an outside constituent, e.g. a topic marker, at least in the form of simplex morphemes), "is operative in some languages, e.g., Tamil, and not in others, e.g., English" (Talmy 2010a: 25); factor Db1 (degree of stress on a constituent), "functions extensively in some languages, e.g., English, but only modestly in others, e.g., French" (Talmy 2010a: 25). The current study will examine the 11 factors in Category A (Factors involving properties of the morpheme) only.

The basis for comparing English and Chinese attention systems is to identify the corresponding linguistic units in the two languages. In English, the smallest unit that combines sound with meaning is acknowledged to be a morpheme. Finding its counterpart in Chinese has been a hot issue among Chinese linguists.

Since Ma Shi Wen Tong (马氏文通) by Ma Jianzhong (马建忠) in the late 1800s，a Ci (词) (usually made up of two characters) had been taken as the smallest sound meaning combinatory unit. A new school of Chinese linguistics since the 1980s represented by Xu Tongqiang (徐通锵) and Pan Wenguo (潘文国)，however，is against this view and has criticized it for neglecting the special features of the Chinese writing system. It is held that "a sinogram is a three-in-one basic structural unit of form, sound and meaning, which is applicable in both written and spoken Chinese, and therefore is the foundation of any Chinese study" (Xu 2008: 1). This latter view has many supporters for it has recognized the unique status of Chinese written characters which have psychological reality in native speakers. Numerous brain imaging studies have been done to find out the neural mechanisms in character processing (Biederman & Tsao, 1979; Booth et al., 2006; Chou et al., 2012; Guo & Burgund, 2010; Perfetti & Liu, 2005; Sun, 2011; Tan et al. 2005; Wu et al., 2012).

However, a recently emerging view is that sound and meaning combination in Chinese doesn’t stop at the character level. Rather, it goes one step further down into the radical, a component of a character. This third view takes a radical as an unalienable part of a word with identifiable meaning and function and criticizes those linguists who do not assign due linguistic status to radicals for the mere reason that "they are part of the orthography, and orthography is viewed as no more than a record of language, not the language itself" (Chen 2010: 113).

Unlike alphabetic words, Chinese characters are ideographic symbols and have no grapheme-phoneme conversion rules (Hoosain & Shiu 1989: 706). The visual form of a character only provides very limited information about the sound of the character. Simple characters with similar shapes have very different sounds, for example, 田/tian2 (tián), 由/you2 (yóu), 甲/jia3 (jiǎ), and 申/shen1 (shēn) have similar looks but completely different pronunciations (Shan et al. 2010: 148). Each Chinese character is composed of different numbers of strokes, "ranging from one to more than 20" (Shan et al. 2010: 148). Strokes are combined to form radicals, which in turn constitute characters. Many radicals are simple characters themselves, having their own meanings and pronunciations, whereas others are bound forms, always acting as components of words and must be combined with other radicals to form characters (Shan et al. 2010: 148). This is similar to the distinction between free morphemes and bound morphemes in English.

The present study agrees with the third view and regards a radical as the smallest unit that has meaning distinguishing value, therefore putting Chinese radicals and English morphemes at the same level. In most contemporary Chinese dictionaries, there are usually 188 radicals listed as in current use. Based on Chen (2010) (examining the metaphorical and metonymic connections among the characters within the WATER network), this project will choose the WATER radical, which has three alloforms (namely, 1) 水 meaning "water"， 2) 冫meaning "two drops of water"，and 3) 氵meaning "three drops of water" ) for an in-depth analysis of the attentional effects of these radicals. A comprehensive WATER network consisting of 399 Chinese characters is established based on Hui Yu’s A New Century Chinese-English Dictionary (2003). Afterwards, comparisons regarding the linguistic attentional factors under Category A of Len Talmy’s attention system are made between the WATER network and Lampert’s (2009) EMO network of 561 words, so as to identify commonalities and differences between Chinese and English.

In a word, linguistic attentional properties in English and Chinese are examined so as to find out which of the existing factors under Category A are English- or Chinese- specific and which are shared.

3.3 Applicability across Language Modalities

This project will extend the existing Attention System of Language proposed by Talmy from the spoken modality to also cover the written modality. Talmy (2010a) has talked about differences between spoken language (English in particular) and sign language (specifically ASL-American sign language), but he hasn’t yet touched upon written language. Tina Lampert’s "Attentional profiles of parenthetical constructions: some thoughts on a cognitive-semantic analysis of written language" (2011) has given insights on "major differences as well as correspondences in the ‘attentional behavior’ between spoken and written parenthetical constructions". Given the importance of the Chinese orthographic system in the development of the Chinese language, it is necessary to examine whether the attentional factors can be applied to the written modality, especially to written Chinese.

The neglect of written modality in linguistic studies can be traced far back to the Father of Modern Linguistics F. de Saussure’s Cours de Linguistique Générale (1916). Saussure himself holds a paradoxical attitude towards the status of written language. At first, he states that written texts cannot be ignored in linguistic research:

Languages are mostly known to us only through writing. Even in the case of our native language, the written form constantly intrudes. In the case of languages spoken in remote parts, it is even more necessary to have recourse to written evidence. The same is true for obvious reasons in the case of languages now dead (Saussure 2001 [1916]: 24).

Later, however, he claims that written languages are only representations of spoken languages and are subordinate to the spoken:

A language and its written form constitute two separate systems of signs. The sole reason for the existence of the latter is to represent the former. The object of study in linguistics is not a combination of the written word and the spoken word. The spoken word alone constitutes that object. But the written word is so intimately connected with the spoken word it represents that it manages to usurp the principal role. As much or even more importance is given to this representation of the vocal sign as to the vocal sign itself, it is rather as if people believed that in order to find out what a person looks like it is better to study his photograph than his face (Saussure 2001 [1916]: 24-25).

Saussure (1916) lists four reasons for the ‘fake’ prestige of written language. Firstly, "the written form of a word strikes us as a permanent, solid object and hence more fitting than its sound to act as a linguistic unit persisting through time" (Saussure 2001 [1916]: 26). It is believed that the connection between a word and its written form is superficially imposed. Secondly, "for most people, visual impressions are clearer and more lasting than auditory impressions" (Saussure 2001 [1916]: 26). Therefore people prefer the visual and gradually the written image usurps the position of the sound. Thirdly, "a literary language enhances even more the unwarranted importance accorded to writing" (Saussure 2001 [1916]: 26). Literary language is disseminated through dictionaries, grammar books and school courses, thus giving an "unreal" impression that the written is more important than the spoken. Finally, "where there is any discrepancy between a language and its spelling, the conflict is always difficult to resolve for anyone other than a linguist" (Saussure 2001 [1916]: 26). Most of the time, the written form wins, for the mere reason that resorting to written records offers an easier solution.

Substantial subsequent studies in liguistics follow Saussure’s assignment of the roles of the written and the spoken languages and almost unanimously avoid giving written language serious consideration. However, it should be noted that Saussure divides writing into two systems: the ideographic (e.g. Chinese), "in which a word is represented by some uniquely distinctive sign which has nothing to do with the sounds involved" (Saussure 2001 [1916]: 26), and the phonetic (e.g. English), which is "intended to represent the sequence of sounds as they occur in the word" (Saussure 2001 [1916]: 26) and he clarifies that "our survey here will be restricted to the phonetic system of writing, and in particular to the system in use today, of which the prototype is the Greek alphabet" (Saussure 2001 [1916]: 27). This obviously implies that when it comes to studies concerning Chinese, the ideographic writing system should be given special treatment.

Before Saussure, philosophers including Aristotle, Rousseau and Hegel also hold that the writing is "sign of a sign", with the latter "sign" referring to the phonetic. 51 years after Saussure’s Course in General Linguistics, the French philosophier Jacques Derrida systematically criticized the former’s phonocentrism in De La Grammatologie (1967):

Saussure confronts the system of the spoken language with the system of phonetic (and even alphabetic) writing as though with the telos of writing. This teleplogy leads to the interpretation of all eurptions of the nonphonetic within writing as transitory crisis and accident of passage, and it is right to consider this teleology to be a Western ethonocentrism, a premathematical primitivism, and a performlist intuitionism. Even if this teleology responds to some absolute necessity, it should be problematized as such (Derrida 1997 [1967]: 40).

For long, Western tradition of linguistic studies have mainly focused on the spoken language and have failed to give the written modality due attention. Chinese and English belong to two different language families—the Sino-Tibetan and the Indo-European respectively. In writing, Chinese adopts a hieroglyphic ideographic writing system, while English uses grapheme combinations based on an alphabetic system to transcribe pronunciation. The written Chinese forms have a closer link to meaning than the English alphabetic linear arrangements. The written modality should not and can not be ignored in linguistic research, as it holds the potential to present a whole new perspective on the nature of language. The current study starts with an examination of the Chinese writing system.

3.4 Applicability across Sensory Systems

Discussions about commonalities and differences between auditory perception (as employed in listening) and visual perception (as employed in reading) are necessary for a cross-sensory-system mapping of the attentional factors from spoken English to written Chinese. Apart from the obvious fact that when two people are engaged in spoken conversations, hearing is more important than vision and that it is exactly the opposite while reading, another reason why distinctions between auditory and visual perception must be analyzed in detail in this study is that the connections among the written and spoken modality, the writing system and orthography in the two languages are different.

While it is taken for granted, though mistakenly, that the alphabetic phonemic written English serves as a passive recorder of Spoken English, the status of the logographic morphosyllabic Written Chinese in relation to Spoken Chinese has been much debated. Chinese characters, especially the hieroglyphic characters (for example, the character 火 meaning "fire" resembles the shape of burning wood), carry meaning through their visual appearances in themselves. English speakers, however, when seeing the word "f-i-r-e", firstly think of the pronunciation and then go on to recognize the meaning. A further perceptual difference in the writing systems of English and Chinese is that, while English words are unanimously in left-right linear order, Chinese characters mainly fall into six kinds of structures, i.e., 1) left-right (as in 泪，刘)；2) left-middle-right (as in 辩，鞭) 3）up-down ( as in 草，吴); 4）up-middle-down (as in 蕙，翼); 5) inside-outside in full enclosure (as in 国，圆)；and 6) inside-outside in half enclosure (as in 同，间). These structures of characters may also influence visual perception.

In English, the memorizing difficulty of the word is usually related to its length, i.e., the number of alphabets contained. In Chinese, however, each character takes up exactly the same size of square â–¡ and the difficulty lies in the intensity of strokes within this square. It is assumed that processing English and Chinese written forms requires different spatial recognition mechanisms.

3.5 Summary

4. Research Methodology

4.1 Introduction

4.2 Introspection

The current study is a theoretical analytic one in which the major method used is introspection.

"Introspective observation is what we have to rely on first and foremost and always" (p. 185), as William James, the Father of American Psychology, states in The Principles of Psychology (1890). The concept of introspection is closely related to "consciousness", "attention", "contemplation", or "thinking" (Carruthers, 1996; Dewey, 1958; Howe, 1990, 1991; Pronin & Kugler, 2007; Ryle, 1949; Stanley, 2012; Titchener, 1909). Introspection "refers to the direct attending to one’s own conscious state (say, how a cup is experienced)", while non-introspection "refers to attending to a stimulus" (Overgaard 2006: 662). To introspect is to observe the mental events within oneself (see Danziger 2001: 7888). It is "both the ability and activity of examining and describing one’s own internal psychological states and processes" (Willems 2012: 672). The term "introspection" can be defined as "intermittent re-representation of the contents of consciousness and is separated from a non-introspective form of consciousness and is separated from a non-introspective form of consciousness responsible for the primary processing of events" (Guggisberg et al. 2011: 1900). Simply put, introspection is "one thought ‘observing’ the preceding one" (Howe 1991: 28).

Introspection is frequently used in linguistic studies. Leonard Talmy, the Father of Cognitive Semantics, comments, "like any method in a scientific endeavour, introspection must be employed with rigor" and that "the method of introspection can be justified in much the same way as the methods settled on by any science" (Talmy 2000: 5). He further says that "the use of introspection must be recognized as an appropriate and arguably necessary methodology in cognitive science, together with the other generally accepted methodologies" (Talmy 2000: 6). In "the attention system of language—a report on work in progress" (Talmy, 2010a), Len Talmy states that in the investigation of attention in language, a combination of introspection and semantic and syntactic analysis in conjunction with analytical thought, which is introspective itself, is necessary, and that "systematic manipulation of ideas", "abstraction", "comparison", and "reasoning" need to be included (Talmy, 2010a).

The two basic structural characteristics of linguistic introspection are: 1) "two levels of consciousness"; and 2) "accessibility to introspection" (Talmy 2010a: 3-4).

Firstly, Talmy defines linguistic introspection as "conscious attention directed by a language user to particular aspects of language as manifest in his own cognition" (Talmy 2010a: 4). Certain aspects of language can appear in a language user’s consciousness "through perception of speech, by internal evocation, or spontaneously" and the "consciousness" here is termed as "level-1 consciousness", in distinction from "level-2 consciousness", which "can also occur in the same individual at the same time that has as its object (part of) the contents of the first level of consciousness" and "can be volitionally evoked and directed at a selected linguistic target on the first level" (Talmy 2010a: 4). In a word, introspection is the cognitive pattern when all the above components are engaged (Talmy 2010a: 4).

Secondly, Talmy explaines that accessibility to introspection includes both readiness and amenability. Different aspects of language vary in "their readiness to appear in first-level consciousness" and if such aspects of language are present in level-1 consciousness, they "differ in their amenability to attention directed at them from level-2 consciousness" (Talmy 2010a: 4). Further, if an aspect of language "has greater strength and clarity in level-1 consciousness and can remain more stably present there while attention is directed at it", it is more amenable (Talmy 2010a: 4); on the contrary, "it is less amenable if it is fainter, vaguer, or more elusive under such attempted scrutiny" (Talmy 2010a: 4).

Martina Lampert, in her Attention and recombinance: a cognitive-semantic investigation into morphological compositionality in English (2009) affirms the role of introspection in linguistic research. She says that there are at least two basic objections that can relativize "any rejection of introspection and theoretical analysis on principled grounds (most typically advanced by experimental scientists from ‘beyond’ the humanities)" (Lampert 2009: 150):

First, any bland refusal of introspection "as such" has to be scrutinized against an equally unfounded trust in an a priori acceptance of the empiricist methods "as such". Any such apodictic claims are subservient to perpetuating specific (though hegemonic) ideologies of scientific research; and, second, introspection (or analytic thought) is an unalienable step and a general prerequisite that cannot be circumvented even by the "hard" sciences. It allows for a systematic probing into all kinds of metacognitive queries that emerge in any scientific activity. Hence, research in both the corpus-based and –driven as well as the experimental paradigms, for instance, will have to employ introspective (that is, metacognitive) methods of analytic thoughts such as reasoning and deduction in both formulating and interpreting their respective hypotheses. (Lampert 2009: 150-151)

In the present study, introspection is needed for a reviewing past research, evaluating their relevance to the topic, identifying and formulating concrete research questions as well as an in-depth analysis of the results and implications.

4.3 Survey

Along with introspection, an empirical survey in the form of a questionnaire is the other research method employed in this study. The empirical part is conducted in order to provide specific data from written Chinese for Talmy’s Attention System of Language.

Due to limited time and space, the current study limits its disccussions within the 11 linguistic attentional factors in Category A (Factors involving properties of the morpheme) in the attentional framework. 11 multiple-choice questions are designed respectively in correspondence to the 11 factors, with the characters under examination all taken from the WATER Network (for the complete WATER Network, please refer to A3 in the Appendices attached). After the questionnaire is designed, a pilot study with 25 people (of which 10 are male and 15 are female; and whose average age is 25.72 years) is conducted in order to determine the validity of the questions, e.g., whether the answers given show a pattern that either confirms or contradicts the assumptions made by the linguistic attentional factors. Following the pilot study is a survey with 100 participants.

As the object of study is written Chinese, all the participants are required to be native Chinese speakers who hold at least a Bachelor’s Degree in either Arts or Science so as to ensure a relatively high level of literacy. As gender is assumed to be not a crucially relevant factor for the dicussion, either male or female participants are equally welcome. The age group is 18-30, with a range of 12 years. The questionnaire (which can be found in A4 of the Appendices) is firstly posted in the form of blogs on Renren (the biggest online social network in China and is the Chinese version of Facebook), on which I have 480 friends and to each friend a link to the blog is sent individually through the message box. After this, the questionnaire is then delivered by emails to nearly 1,000 contacts in my email list.

In total, this empirical survey ranges over one month from February 1st to February 28th, 2013, during which 1,480 people are invited to answer the questionnaire, of which 1/3 have replied. The answers from the first 100 of these cooperative participants who match the requirements are recorded in EXCEL work sheet (please refer to A5 in the Appendices for the detailed record). The data are analyzed with the statistical software SPSS and bar charts are drawn regarding the frequency distribution for the answers to each single question. The implications for these answer distributions will be discussed in more detail in the following chapter. It should be pointed out that 1) of the 100 participants, 45 are male and 55 are female, thus the gender ratio is roughly 1:1 (see Figure 02 below); 2) the average age of the participants is 25.02 years old, ensuring the average literacy level of the participants generally corresponds to the level of a fresh graduate from university with a Master’s Degree according to the Chinese education system (see Figure 03 below).

Figure 02: Gender distribution of the participants in the survey

Figure 03: Age distribution of the participants in the survey

To sum up, this statistical method is complementary to the introspective theoretical method and it provides direct empirical data for the performance of the 11 linguistic attentional factors of Category A (Factors involving properties of the morpheme) in Talmy’s Attention System of Language in written Chinese.

4.4 Summary