Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Text & Talk

An Interdisciplinary Journal of Language, Discourse & Communication Studies

Ed. by Sarangi, Srikant

6 Issues per year

IMPACT FACTOR 2016: 0.448
5-year IMPACT FACTOR: 0.686

CiteScore 2017: 0.63

SCImago Journal Rank (SJR) 2017: 0.326
Source Normalized Impact per Paper (SNIP) 2017: 0.625

See all formats and pricing
More options …
Volume 38, Issue 1


Evaluating news actors in Chinese hard news reporting: language patterns and social values

Changpeng Huan
  • Corresponding author
  • School of Foreign Languages, Shanghai Jiao Tong University, 800 Dongchuan Road, Minhang District, Shanghai, 200240 PR China
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
Published Online: 2017-12-08 | DOI: https://doi.org/10.1515/text-2017-0029


This paper interrogates the ways Chinese journalists judge behaviors of different news actors (individuals or institutions written about in a news report) in Chinese hard news reporting on risk events. Drawing on a newly constructed corpus of hard news reporting and following the appraisal framework, the study examines language patterns of judgement and suggests how lexicogrammatical patterns of judgement, semantic features of judgement and social orders are intimately connected. Corpus analytic findings show that news actors of different social statuses are judged in different manners. Social elites are primarily judged positively in relation to their “capacity” and “tenacity” in dealing with risk situations, whereas ordinary risk makers are mainly judged negatively in relation to their “propriety” in producing risk situations. Additionally, the same judgement marker in the same phraseological pattern may realize different valences of judgement in relation to news actors of different social statuses, and judgement of different news actors tends to select distinctive local patterns. It is argued that such different discursive patterns of judging news actors point to the distinction in the way social values are (re)produced and journalistic professional practices undertaken in the production of Chinese hard news.

Keywords: appraisal; Chinese hard news; corpus-based studies; judgement patterns; local grammar

1 Introduction

This paper sets out to interrogate the ways behaviors of news actors (i.e. individuals or institutions written about in a news report) are evaluated in Chinese hard news reporting (i.e. news items that cover political, economic and social issues) on risk events (i.e. earthquakes and food safety). These are resources delimited as a specific type of attitudejudgement – under the appraisal framework (Martin and White 2005). 1

Research interest in attitudinal language among Chinese linguists is of recent origin (Iwasaki and Yap 2015: 2). Most studies on attitude in general (not following the appraisal framework) in Chinese have largely laid emphasis on the exposition of syntactic and semantic idiosyncrasies or taken a grammaticalization perspective (e.g. Tao 2003; Su 2004; Wu 2004; Tao 2007; Lim 2011; Fuxiang 2011; Endo 2013; Wu 2015; Yang and Yap 2015). Research into attitude in news discourse in line with the appraisal framework has tended to focus on the analysis of emotions (e.g. emotional responses) (see Section 2 for details). The ways news actors are judged in news (e.g. how behaviors or characters of news actors are ethically evaluated in accord with social norms) have as yet tended to be neglected (see Feez et al. 2008 for an exception). This is particularly the case in Chinese news discourse, although studies of appraisal have sparked increasing scholarly interest in China since 2001 (Wang 2001). Hence, we have so far obtained insufficient understanding of how journalistic voices (e.g. reporter voice) are more or less compatible with journalists’ expression of judgement. The social factors that give rise to the realization of judgement in Chinese news discourse are underexplored as well.

To make a contribution to the literature on the link between specific lexicogrammatical expressions of judgement, semantic features of judgement, and the underpinning social values in Chinese hard news reporting, the paper draws on the judgement subsystem of the appraisal framework (Martin and White 2005), a framework that is concerned with dynamic processes of personae construction and ongoing negotiation of dialogic positions in discourse. According to Martin and White (2005: 35), judgement is a subsystem of attitude, together with affect (e.g. emotional reactions) and appreciation (e.g. evaluation of things). judgement itself is “the region of meaning construing our attitudes to people and the way they behave – their character” (Martin and White 2005: 52). It has five features grouped under “social esteem” and “social sanction.” According to Martin (2000: 156) and Martin and White (2005: 52), judgement of esteem deals with “normality” (how special someone is), “capacity” (how capable someone is) and “tenacity” (how resolute someone is); judgement of sanction has to do with “veracity” (how truthful someone is) and “propriety” (how ethical someone is). As Feez et al. (2008: 209–210) and Martin and White (2005: 52) observe, there is a significant difference separating “social esteem” and “social sanction.” “Social esteem” has to do with social admiration or contempt, whereas “social sanction” is a domain of moral right or wrong, a system of moral regulation. For instance, negative values of “social sanction” are usually considered to be an issue of ethical wrongdoing that are in breach of moral orders or punishable by law, whereas negative values of “social esteem” diminish or destroy someone’s esteem in the eyes of the public.

In the following, I will briefly review the literature on attitude studies in news discourse (Section 2). I then present the corpus data and research methods (Section 3), prior to a detailed presentation of corpus analytic findings (Section 4). These corpus analytic findings will be discussed in the context of social values that are currently operational in news production and in relation to journalistic voices (Section 5).

2 A brief review of the literature

In the field of journalism practices, journalistic attitude (e.g. journalists’ expression of emotions or evaluation of people’s behaviors) is often dismissed as a marker of unprincipled and flawed journalism (Chalaby 1998; Mindich 1998; Schudson and Anderson 2009; Peters 2011). Journalistic attitude – authorial (i.e. journalists’ own attitude) or observed (i.e. reporting third parties’ attitudinal responses) (Thomson et al. 2008; Stenvall 2014) – has largely been discussed in relation to journalists’ professional norms of objectivity or neutrality. For instance, Stenvall points out that “reporting on emotions inherently challenges the journalistic ideals of objectivity and factuality” (Stenvall 2008: 1569). Research has demonstrated that journalistic authorial attitude is hardly present in quality journalism in the Anglo-American tradition of news reporting, but that the reporting of observed attitude has become increasingly explicit (e.g. White 1997; Martin and White 2005; Bednarek 2006; Bednarek 2008; Stenvall 2008; Stenvall 2014; Pounds 2010; Peters 2011; Wahl-Jorgensen 2013a). The inclusion of observed attitude in news discourse is considered as a strategic ritual of objectivity in journalism practices (e.g. Tuchman 1972; White 1998; Wahl-Jorgensen 2013a; Wahl-Jorgensen 2013b). However, as Peters observes, “in recent years the dominance of objectivity as a fundamental paradigm in Anglo-American journalism has increasingly been called into question, not only by academics but within the industry itself (Rosen 1999; Frank 2003)” (Peters 2011: 298). Following a number of important recent publications, our understanding of attitude has shifted from a traditional view that considers attitude as linguistic markers in texts, to a practice-focused view that studies attitude as discursive acts of stance taking by writers/speakers vis-à-vis people, objects or propositions concerned (e.g. Martin and White 2005; Du Bois 2007; Jaffe 2009; Perrin 2012; Vertommen et al. 2012; Wahl-Jorgensen 2013b).

This paper builds on this view by considering expression of judgement (a subsystem of attitude) in news discourse as journalists’ discursive practices in enacting and concealing certain social and professional values. An important means of practicing journalism is realized by lexicogrammatical resources. Hence, the paper takes language resources of judgement as the analytical point of entry to understand how language patterns of judgement (re)produce social values and how social values regulate the actual expression of judgement in a specific discourse context, namely Chinese newspaper reports on risk events.

The study of Liu and Stevenson (2013) is one of few studies available in the literature examining judgement features in the context of Chinese news reporting. They examined judgement features in 15 news stories concerning the 2008 Sichuan earthquake in China (five from People’s Daily, five from Australian Chinese Daily, and five from Sydney Morning Herald). They found that the prominent judgement features in Peoples’ Daily and Australian Chinese Daily were related to the “tenacity” of rescue workers and that of the Chinese leadership, followed by evaluation of their “capacity” and “propriety.” However, in Sydney Morning Herald the primary judgement feature related to participants’ “capacity” to carry out rescue work. However, they did not study valences of judgement features or the lexicogrammatical patterns of judgement. To probe more fully into patterns of judgement, this paper defines it as phraseological patterns in which judgement markers are embedded, patterned distribution of semantic features of judgement, and patterned relations between semantic features of judgement, their valences, and social statuses of news actors.

3 Data and methods

This section first presents the corpus data on Chinese and Australian hard news reporting, and then explains the analytical procedures undertaken in annotating judgement in the corpus.

3.1 Corpus data

The corpus data comprises news reports on food safety (e.g. food poison) and earthquakes in Mainland China. These two risk events were selected due to their significance to people’s daily life in China. For instance, according to official figures, more than 69,197 people were confirmed dead, 374,176 injured and 18,222 missing in the event of the 2008 Sichuan earthquake (Xinhua Net, 20 July 2008). Additionally, food safety in China is a public concern in the press; for instance, the most infamous 2008 Chinese milk scandal raised wide public concerns about food safety domestically and internationally. 2

I selected news reports on earthquakes from the Zhōngguó Zhòngyào Bàozhǐ Shùjùkù (‘Chinese Database for Key Newspapers’). The selection of news reports on food safety drew on a website that provided regular updates on all food safety news stories covered in Chinese newspapers since 2004. 3 The selection of corpus data was confined to a period of one decade from 2004 to 2013. All news reports in the designated period were selected and carefully examined in line with the definition of hard news (Reinemann et al. 2011). Eventually, I selected 485 news reports on food safety (215,433 words) and 129 news reports on earthquakes (57,072 words). 4

3.2 Analytical procedures

In annotating local phraseological patterns of judgement in Chinese, it is necessary to briefly introduce the unique nature of Chinese grammar, especially its part-of-speech (POS) and local phraseological patterns. The nature of the Chinese grammar would not permit the investigation of language patterns of evaluation of the kind carried out in English by Bednarek (2008; Bednarek 2009) and Hunston (2011). This is because there is a general one-to-one correspondence between the POS of a word and its syntactic role in English. For instance, nouns typically function as subject or object, verbs as predicate, and adjectives as attributive. By contrast, the relationship in Chinese is that of one-to-many (Lu 2003). For instance, adjectives can perform all syntactic roles – subject, object, predicate, attributive, and adverbial. Despite this, local patterns in Chinese are rather limited and stable. In particular, I have identified five major patterns of judgement markers in Chinese for the purposes of this study (see Table 1).

Table 1:

Local phraseological patterns of judgement markers in Chinese.

In annotating markers of judgement, I assigned each marker a corresponding feature of judgement, and the decision was made on the basis of its meaning in context. Similar to the tripartite distinction of affect – positive, negative and neutral – advanced by Bednarek (2008), I added neutral judgement to the dichotomy of judgement polarity. Neutral judgement in my corpus mainly has to do with irrealis triggers of judgement (e.g. actions not happened yet) in cases such as “consumers should be cautious in selecting their purchases”.

In corpus-assisted appraisal studies, a combination of automated and manual analysis is usually the norm, and I have conformed to this norm. First, I extracted a list of potential judgement markers in Chinese based on the HowNet Knowledge Databases and a list of Chinese sentiment lexis by Xu and Tao (2003). I then relied on the corpus software SysConc (Wu 2009) to automatically generate all concordance lines containing the selected judgement markers. All concordance lines were subsequently transferred to a self-designed Filemaker-based coding scheme so as to facilitate coding of judgement. The coding scheme was also used to extract frequencies (e.g. count of each judgement marker and each judgement feature in question). The next step was to manually determine the relevance of the senses of these items to judgement. This issue concerns the problem of polysemy and homonymy, which is a central issue to the annotation of a sense-sensitive corpus.

The coding was conducted on the basis of the semantics of each judgement marker in its context. In cases of clashes, I consulted other native speakers of Chinese to elicit their understanding of the ambiguous meanings of a judgement item. For general judgement items (e.g. good, bad, great), their meanings were decided on the premise of their actual meaning in context. For instance, “good” in “a good man” is a judgement marker but it is a marker of appreciation in “that’s a good way of knowing the truth”. In some cases, one appraisal item actually realizes another category. For instance, 假冒 (‘faked’) in假冒酱油 (‘faked soy source’) evaluates the quality of the soy source, but it actually realizes judgement of negative “propriety” in its embedded phrase 某人生产假冒酱油 (‘Someone produces faked soy source’). Cases of such target-value mismatches are called “Russian dolls” (Thompson 2014: 47), meaning that “an expression of one category of attitude may function as a token (an indirect expression) of a different category; and that token may itself function as an indirect expression of yet another category, and so on” (Thompson 2014: 47). In coding cases of this kind, I assigned judgement feature to their “value” (actual attitudinal meaning).

On the basis of the semantic elements identified in studies of evaluative language by Bednarek (2008), Galasinski (2004), Hunston (2003), and Martin and White (2005), I distinguished between four main elements for describing semantic elements of judgement: the appraiser (i.e. the one who judges other persons’ behaviors or characters); the judgement itself (i.e. the particular feature of judgement involved); the target of judgement (i.e. the ones whose behaviors or characters are evaluated); and trigger (i.e. what causes the judgement).

In an attempt to categorize news actors, Bell (1991: 194) argues that most news actors fall into one of the following categories: political figure, official, celebrity (e.g. film or music star), sportsperson, professional or other public figure (e.g. lawyer), criminal or accused, human interest figure, and participant (e.g. victim or witness). Premised on this categorization and also on the news actors actually identified in my corpus (e.g. celebrity and sportsperson are irrelevant to our study), I broadly divided news actors into two groups: the powerful (i.e. social elites, including political and professional figures) and the powerless (i.e. citizens or consumers, and risk makers). Risk makers are those individuals or institutions (e.g. food producers) who produce food that threatens others’ health or life. In the coding process, I also noticed the presence of journalists’ explicit reference to themselves (e.g. 记者仔细察看 – ‘reporters carefully examine …’), although the occurrence was rare (five occurrences in total). For those news actors that are unknown or implicit, I coded them as unspecific.

4 Corpus analytic findings: patterns of judgement

This section first outlines corpus analytic findings in relation to semantic features of judgement, their valences, and social statuses of news actors (Section 4.1). It then examines phraseological patterns of judgement (Section 4.2).

4.1 An overview of corpus analytic findings

Table 2 shows the normalized percentage of the distribution of judgement features identified in the corpus. 5

Table 2:

Overall distribution of judgement features in relation to polarity.

We can observe from Table 2 that overall the number of positive “social esteem” (comprising “normality,” “capacity” and “tenacity”) markers (46 %) overtakes that of negative markers (10 %) but the reverse is true in the realization of “social sanction.” Neutral judgement markers are minimally relevant in the realization of judgement. If we continue examining the valences of “social esteem” and “social sanction,” we will find that all “social esteem” features (i.e. “normality,” “capacity” and “tenacity”) prefer positivity over negativity. This is also the case in realizing “veracity,” a feature of “social sanction.” However, in realizing “propriety,” negative markers (34 %) by far outnumber positive ones (4 %).

This finding largely corroborates the one presented in the study of Liu and Stevenson (2013). They have found that Chinese news coverage of risk events (i.e. the 2008 Sichuan earthquake) tended to foreground “tenacity,” “capacity” and “propriety” of Chinese power elites, and also presented cross-cultural differences in the distribution of judgement features. I have elsewhere found similar cultural differences in presenting journalistic judgement (Huan 2015). The combined findings point to the cultural distinction of judgement in news discourse, suggesting that it is necessary to probe more fully into the underpinning sociocultural values to understand the expression of judgement in news discourse (see Section 4).

Figure 1 displays corpus distribution of news actors in relation to the valences of judgement features. The most obvious difference lies in the fact that risk makers are predominantly judged negatively, but social elites are overwhelmingly judged in a positive manner (see Figure 1 below). Such a contrast may not be particularly striking, but the details are of interest (see Section 4.2 below). Neutral judgement markers chiefly relate to judgement of citizens and they are hardly enacted in judging other news actors. Hence, my discussion on judgement markers will focus on positive and negative markers. Additionally, the analysis will foreground three groups of news actors, namely risk makers, social elites and citizens, in that judgement of other news actors is minimally relevant in the corpus. More specifically, the number of authorial judgement is negligible with five occurrences, and four of them are realized by the marker 仔细 (‘careful’) to show journalists’ careful reading of the information of a risky product (e.g. food with potential health risk). Unspecific news actors will be ignored because we are unable to know who are judged in news reporting.

Distribution of judgement valences in relation to news actors.
Figure 1:

Distribution of judgement valences in relation to news actors.

Let us now turn to examine the relationship between judgement features and news actors. Table 3 shows the distribution of judgement features vis-à-vis their polarity orientations and the social statuses of news actors.

Table 3:

Distribution of judgement features in relation to news actors and valences.

Table 3 shows that negative “propriety” markers are predominantly adopted to judge risk makers (279 occurrences) rather than to judge social elites (one occurrence) or citizens (zero occurrences). Positive markers of “capacity” and “tenacity” are mainly deployed to judge social elites. The predominant negative markers in relation to judgement of social elites relate to their negative “capacity” (eight occurrences). Citizens are hardly judged negatively by reference to “social sanction” (i.e. “veracity” and “propriety”). Positive judgement of citizens mostly relates to their “tenacity,” whereas negative judgement mostly has to do with their “capacity.”

Bringing together the corpus analytic findings in this section, it appears that “social esteem” is oriented towards positivity, while “social sanction” is oriented towards negativity. Among all news actors concerned, risk makers are predominantly judged negatively, but social elites are overwhelmingly judged positively. Negative judgement of risk makers primarily relates to that of “propriety,” and positive judgement of social elites chiefly concerns their “capacity” and “tenacity.” In the next section I examine the local patterns of judgement markers.

4.2 Local patterns of judgement

For reasons of scope and due to their infrequent occurrences (see Section 4.1 above), markers of “normality” and “veracity” will be excluded from pattern analysis. In other words, I will focus on the local patterns of “capacity,” “tenacity” and “veracity” in relation to risk makers, citizens and social elites (see Table 4). Those markers with less than five occurrences will be excluded from analysis in that they occur too infrequently to establish a solid ground for pattern analysis.

Table 4:

Distribution of “capacity,” “tenacity” and “propriety,” and their valences across POS.

In order to avoid generating a pattern analysis that is too dense to be comprehensible, I decided to further confine pattern analysis to verbal and adjectival patterns of “capacity,” “tenacity” and “propriety” in relation to citizens, risk makers and social elites. Nominal and adverbial patterns will be briefly discussed here due to their infrequent occurrences. Specifically, Table 4 shows that nouns mainly realize positive “capacity” and positive “propriety,” and that adverbs chiefly realize positive “capacity” and negative “propriety.”

All the six nominal markers of “capacity” serve to positively judge social elites (困难 – five occurrences; 艰难险阻 – one occurrence; both meaning ‘difficulties’) in the pattern of Shubin (see example [1]).




‘The rescue traffic is overcoming various difficulties to arrive at the disastrous areas faster.’

The stance noun itself (困难, ‘difficulty’) bears a negative meaning. However, when it is adopted in the pattern of Shubin and especially when collocating with the verb ‘overcome’ (克服), it enacts positive evaluation of the “capacity” of the rescue team.

Likewise, all nominal markers of “propriety” (i.e. 爱心 – ‘love’, six occurrences) construe positive judgement of social elites. The appraisal item 爱心 (‘love’) often collocates with the verb (‘donate’) in the pattern of Shubin (see example [2] below). The beneficiaries are typically earthquake victims or those persons influenced by the disaster. As such, news actors (social elites in this case) who perform the donating actions are positively judged.




‘Staff from all sectors of the CPC and national bureaus donate money enthusiastically to show their love to those in the disaster areas.’

In addition to the Shubin pattern, the corpus has documented the pattern of Dingzhong (attributive + head noun) (see example [3] below), in which the head explicates the persons under evaluation while the attributive assigns a positive judgement value.




‘X has received the donation of 241.71 million CNY from caring people from all walks of life and those.’

Bringing together nominal patterns of positive “capacity” and positive “propriety,” it was found that they both prefer the pattern Shubin, and that they both serve to positively judge social elites.

As to adverbial markers, although no single adverbial marker of positive “capacity” occurs more than five times in the corpus to be eligible for pattern analysis, it is interesting to notice that all adverbial markers of positive “capacity” occur in the same pattern (i.e. Zhuangzhong) to positively judge social elites (see example [4]).




‘General Secretary Hu Jintao, and Premier Wen Jiabao requested (government officials) to spare no efforts to rescue those people in the disaster areas.’

In like manner, the primary pattern of adverbial markers in realizing negative “propriety” is also that of Zhuangzhong (see example [5]). However, all these markers serve to negatively judge risk makers, who illegally produce risky food.




‘The investigation shows that the criminals Gao and Song illegally obtained profits.’

Collating adverbial patterns of positive “capacity” and negative “propriety” in the corpus, it was found that they both favored the pattern Zhuangzhong. However, the target of judgement in adverbial patterns of positive “capacity” was social elites, whereas the target in adverbial patterns of “propriety” was risk makers.

4.2.1 Patterns of “capacity”

Table 5 shows that verbal markers of “capacity” mainly serve to negatively judge risk makers (six occurrences) but to positively judge social elites (22 occurrences). This is also the case for adjectival markers of “capacity.”

Table 5:

Distribution of “capacity” of selected news actors across POS and valences.

As to verbal markers, both negative “capacity” of risk makers (see example [6]) and positive “capacity” of social elites (see example [7]) tend to occur in the pattern Zhuwei.




‘Xi Jinping … discerned in detail the construction situation.’



王老板称, 他非常了解一些地下加工厂制作猪红的制作过程。

‘Mr. Wang said, he was quite familiar with the processes of lards production in illegal workshops.’

It is worth noting that the same judgement item in the same pattern may realize rather different valences of judgement. For instance, the judgement marker ‘discern’ (了解 in examples [6] and [7]) is employed to inscribe positive judgement when news actors under evaluation are the powerful voice (social elites in example [6]). In contrast, when news actors under evaluation are powerless (risk makers in example [7]), it realizes negative judgement. The finding thus points to an important fact that valences of judgement are closely linked to social status of news actors under evaluation.

Adjectival markers of “capacity” of risk makers, citizens and social elites all primarily take the pattern Zhuangzhong (adverbial + head verb), in which the adjectives modify a verbal group (see example [8]). In this pattern nominal group, preceding adjectival markers realize the target of judgement, and the verbal group explicates their behaviors.




‘Several women swiftly picked them up and threw to the plastic bags.’

Despite the similarity, local adjectival patterns in realizing “capacity” of risk makers, citizens and social elites also exhibit disparity. For instance, another salient adjectival pattern in realizing “capacity” of risk makers is that of Zhongbu (head + complement), with the head realized by a verbal group (example [9]).




‘The factory said this was because dealers improperly store (the goods).’

By contrast, other prominent adjectival patterns in realizing “capacity” of social elites are Dingzhong (attributive + head) and Zhuwei (subject + predicate). In the former pattern, social elites (political figures in most cases) are grammatically realized as the head and are directly modified by the evaluative adjectives (e.g. 高效、务实的党委、政府 – ‘Efficient and pragmatic Party Committee and government’). In the latter pattern, adjectival markers of “capacity” function as the predicate of the clause (e.g. 他认为中国政府很了不起。‘He thinks the Chinese government is doing a good job’).

Due to the infrequent occurrences, it is difficult to observe the preferred local adjectival patterns in realizing “capacity” of citizens other than Zhuangzhong.

However, bringing together the pattern differences in realizing “capacity” of risk makers, citizens and social elites, we find that the pattern Zhongbu (head + complement) (see example [10]) appears to be associated with a negative value of judgement of “capacity” of risk makers.




‘The mouldy smell of lantern festival dumplings may be caused by consumers improperly storing (them).’

In addition, the pattern Zhongbu is hardly adopted to evaluate “capacity” of social elites. This finding suggests that social statuses of news actors may select certain local patterns and largely determine valences of judgement.

4.2.2 Patterns of “tenacity”

Table 6 shows that the “tenacity” of risk makers, social elites and citizens is primarily realized in adjectival forms and that verbal markers of “tenacity” of these three news actors only sporadically occur. Thus, this section will focus on adjectival markers of “tenacity.”

Table 6:

Distribution of “tenacity” of selected news actors across POS and valences.

Adjectival markers of both positive and negative judgement of “tenacity” of risk makers are primarily embedded in the pattern Zhuangzhong (adverbial + head verb), in which the judgement adjectives serve to modify the following verb. It counters our expectation to find more adjectival markers of positive “tenacity” than those negative ones in relation to risk makers. However, a close examination of the targets of judgement reveals that positive “tenacity” of risk makers relates to institutions (e.g. large food factories) (see example [11]), whereas negative “tenacity” relates to individual food producers (see example [12]). In the event of food safety, institutions are represented as news actors who actively seek cooperation with government in dealing with the risk or in initiating food recall. By contrast, individuals are represented as those who recklessly place or produce food.




‘Mr. Liao, a staff responsible for marketing in the Beijing X Food Corporation, said that (they) will actively collaborate with relevant government departments to solve the issue.’




‘Officials of Food and Medicine Administration saw on the production site that some food was recklessly placed on ground.’

All adjectival markers of “tenacity” in relation to social elites are positive in terms of valence. The primary local pattern is that of Zhuangzhong as well (see example [13]), expressing judgement of political figures’ resolute, careful and serious dealing with risk situations. In addition, the pattern Dingzhong (attributive + head) is noticeable (see example [14]). This pattern often positively judges the resolute leadership of the central government. The adjectival marker (i.e. attributive) directly modifies and thus evaluates political figures and Chinese government (i.e. head).




‘Xi Jinping … earnestly watched the reconstruction plan.’




‘With the resolute leadership of the Central Party Committee and State Council, we will eventually win the battle against earthquake after-effects.’

Adjectival markers of “tenacity” in relation to citizens are typically collocated to their left with a modal verb, expressing obligation (see example [15]) primarily in the pattern Zhuangzhong.




‘The Administration for Industry and Commerce reminds that citizens should carefully check product labels before purchasing any dairy drink.’

In short, positive judgement of “tenacity” upon citizens is often associated with kind reminders offered by elites.

4.2.3 Patterns of “propriety”

Table 7 shows that both verbal and adjectival markers mainly realize negative judgement of risk makers and positive judgement of social elites, and that adjectival markers far outnumber verbal ones.

Table 7:

Distribution of “propriety” of selected news actors across POS and valences.

Verbal markers of positive “propriety” in relation to social elites mainly express their adherence to the common interest of the masses in the event of an earthquake, but no single marker occurs frequently enough to be eligible for pattern analysis. Hence, I turn to examine verbal markers of negative “propriety” in relation to risk makers, as in example (16).




‘Some dealers fake the production date of food.’

Negative “propriety” mainly concerns the process of faking food (假冒, 伪造). The verbal markers are mainly realized in the pattern of Zhuwei as above, where the first nominal group refers to individuals or institutions under evaluation.

As to adjectival markers of “propriety,” markers of positive “propriety” are primarily adopted to evaluate social elites in the local pattern of Zhuangzhong (adverbial + head verb). These markers mainly refer to how selfless social elites are in donating money to those persons influenced by the disaster at issue. The only negative judgement of “propriety” of social elites refers to an anonymous disgraced government official, as in the local phrase of 腐败官员 (‘a corrupted official’). I will not present those patterns here due to their infrequent occurrence (less than five occurrences). Instead, I now turn to adjectival markers of negative “propriety” in relation to judgement of risk makers.

While positive judgement of “propriety” of social elites favors the pattern Zhuangzhong, negative judgement of “propriety” of risk makers prefers the pattern Dingzhong. In the former pattern, the judgement item is adopted to evaluate the action conducted by social elites, while in the latter pattern the judgement item directly judges the character of risk makers against morality and law-abidingness (e.g. 黑心商户 – ‘black-hearted businessmen’; 地下窝点 – ‘illegal food workshop’).

There are also cases where markers that are usually adopted to evaluate entities (i.e. appreciation) may realize judgement of “propriety.” For instance, the appraisal marker 假冒 (‘counterfeited’) in example (17) realizes appreciation of the soy source at issue in its immediate local pattern (i.e. Dingzhong).




‘This criminal group has been producing and selling counterfeited soybean source since early 2009.’

However, it is also embedded in a larger pattern of Shubin (verb + object) in which the entire local pattern of Dingzhong functions as the object. It is in the pattern of Shubin that the appraisal item (i.e. 假冒, ‘counterfeited’) realizes and triggers the negative judgement of “propriety.” Examples of this kind are typical cases of what Thompson (2014: 47) called “Russian dolls.” attitudinal markers of this kind have posed substantial challenges for their annotation and subsequent analysis. On the other hand, these markers are significant for understanding the interwoven relation between evaluative language and local patterns (e.g. Hunston 2011).

5 Discussion and conclusion

Collating the corpus analytic findings, it is the case that positive markers of “normality,” “capacity,” “tenacity” and “veracity” outnumber the negative ones but that the reverse is true in the case of “propriety.” Negative judgement is chiefly concerned with risk makers, whereas positive judgement is mainly linked to social elites. In other words, social elites are more likely to be positively judged, but risk makers are less likely to be so. Among negative judgement of risk makers, it is their “propriety” that is often foregrounded, whereas positive judgement of social elites highlights “tenacity” and “capacity.”

The choice of judgement markers in hard news appears to be incompatible with journalistic professional norms of objectivity or neutrality of news reporting. Instead, it signals Chinese journalists’ attitudinal stance vis-à-vis news actors and events in news discourse. By positively judging behaviors of social elites, Chinese journalists position themselves in alignment with the powerful. This is immediately relevant to the social values that are currently operational in news production in the Chinese context where journalists and their affiliated newspapers need to toe the government line (Zhao 1998; Zhao 2008; Lee et al. 2007). The primary and guiding working principle of all Chinese news agencies is that of dang guan meiti (‘The government should supervise the work of Chinese news agencies’), as prescribed in media policies. 6 Under this principle, what is considered to be of social import by the government will and should be augmented accordingly by Chinese news agencies in news production processes and in news products (e.g. news stories) as well. In this sense, newspapers are supposed to disseminate the sociopolitical values desired by the government. To do so, it is necessary that power elites be discursively construed in a positive way in news reporting, for instance how resolute and capable they are in dealing with risk situations. This explains in part our corpus analytic findings concerning the predominant positive judgement of “capacity” and “tenacity” of social elites.

While it is not the case that only Chinese journalism practitioners share the sociopolitical values of the government, as shown in studies on racist ideology in what is often labelled as Western democratic countries (e.g. Van Dijk 1998; Wodak and Van Dijk 2000), the point is that, as has been reported elsewhere (Huan 2016), Chinese journalists are more likely to defer to news values criteria and professional values in news production in harmony with the dominant social values (Huan 2016). More specifically, in the context of increasing media competition and decreasing government subsidies, Chinese press tends to augment the news values of Eliteness (i.e. the high status of individuals, institutions or nations involved in an event) and Positivity (i.e. the positive aspects of an event) to cater to the political power of the government, and at the same time to foreground the news values of Personalization (i.e. the personal or “human” face of an event) and Negativity (i.e. the negative aspect of an event) to accrue economic profits. Such a distinctive way of enacting news values in the Chinese press is relevant to my corpus analytic finding that the powerful (i.e. social elites) are predominantly evaluated in a positive manner, whereas the powerless (risk makers) are negatively evaluated. The media pursuit of economic profits may also explain why institutional risk makers are more likely to be represented positively, in marked contrast with the way individual risk makers are represented.

Additionally, the involvement of judgement markers in Chinese hard news requires revisiting the often-assumed natural relationship between “reporter voice” and the genre of hard news, especially in the Western context of hard news reporting. According to Feez et al. (2008), “reporter voice” is a feature of “authorial voice” which is defined as “certain language features in texts that contribute to a sense of the presence or personality of a writer” (Feez et al. 2008: 201). The system of “authorial voice” is divided into those language resources that are reserved for expressing objectivity (i.e. “reporter voice”) and those that are adopted to be subjective (i.e. “writer voice”). “Writer voice” is further divided into “correspondent voice” (where judgement is confined to values of “social esteem”) and “commentator voice” (which has access to the full array of judgement values of “social esteem” and “social sanction,” with a tendency to give prominence to “social sanction”) (Feez et al. 2008: 211). As Feez et al. (2008: 198) point out, the objective of hard news reporting is to construct a “reporter voice” featuring an impersonal, anonymous or even absent authorial voice. In other words, as White observes, “the meanings typically avoided [in journalistic reporter voice] include explicit value judgements by the reporter about the morality, competence, normality, etc. of participants” (White 1997: 107). According to Thomson et al., “within English-language journalism, authorial ‘neutrality’ and use of the ‘inverted pyramid’ structure are frequently seen to be distinctive features of the modern hard news report and one of the grounds by which journalists assert the ‘objectivity’ of their writing” (Thomson et al. 2008: 212). However, my corpus analytic finding has shown the involvement of both judgement of esteem and sanction in Chinese hard news. The finding indicates that, on the one hand, there is no natural relation between reporter voice and hard news; at least this is not necessarily true in Chinese hard news. On the other hand, the shift from a “reporter voice” to a “commentator voice” in Chinese hard news points to an important pragmatic function of hard news, that is to (re)produce the dominant power relations in news discourse. This is because adopting a “commentator voice” can substantially facilitate undertaking certain stances by Chinese journalists in news discourse.

The present study has also demonstrated the interlocked relationship between local patterns of judgement markers and their corresponding judgement features. For instance, the judgement marker of kunnan can only express judgement of “capacity” in the pattern Shubin. By examining language patterns of judgement, I found that the same judgement marker (e.g. liaojie) in the same grammatical pattern may realize different valences of judgement in relation to news actors of different social statuses. In addition, certain grammatical patterns (e.g. Zhongbu) are less likely to be adopted to judge the “capacity” of social elites. Furthermore, there seems to be a close link between judgement meanings and word classes. For instance, “tenacity” is chiefly realized in adjectival forms, but seldom in other forms.

To conclude, the present study of the ways Chinese journalists judge behaviors or characters of different news actors in newspaper hard news reporting has documented an interwoven relation between semantic features of judgement, their valences, social statuses of news actors and local grammatical patterns of judgement markers. For reasons of scope, the study could only provide a partial analysis of Chinese journalists’ enactment of attitude in hard news reporting with a focus on judgement. The study is further limited by its focus on print media coverage of earthquakes and food safety in China, and thus it is not possible to generalize the corpus analytic findings to other news genres (e.g. soft news), media outlets (e.g. broadcast news) or other societies. The study has tended to focus on some salient risks in China, and thus has not differentiated man-made risk (i.e. food safety) from natural risk (i.e. earthquake). Future studies are required to probe more fully into their differences in that attribution of responsibility (reference to people responsible for the risks) may vary depending on different kinds of risk.


I express my sincere gratitude to the late Emeritus Professor Chris Candlin for his excellent supervision of my doctoral project, of which this article is one part. I would also like to thank the editor-in-chief Professor Srikant Sarangi and anonymous reviewers for their comments and suggestions to improve this paper. I assume responsibility for any remaining errors.


  • Bednarek, Monika. 2006. Evaluation in media discourse: Analysis of a newspaper corpus. London/New York: Continuum. Google Scholar

  • Bednarek, Monika. 2008. Emotion talk across corpora. Basingstoke: Palgrave Macmillan. Google Scholar

  • Bednarek, Monika. 2009. Language patterns and attitude. Functions of Language 16(2). 165–192. CrossrefGoogle Scholar

  • Bell, Allan. 1991. The language of news media. Oxford: Blackwell. Google Scholar

  • Chalaby, Jean K. 1998. The invention of journalism. Houndmills: Macmillan Press. Google Scholar

  • Du Bois, John. 2007. The stance triangle. in Robert Englebretson (ed.), Stancetaking in discourse: Subjectivity, evaluation, interaction, 130–182. Amsterdam: Benjamins. Google Scholar

  • Endo, Tomoko. 2013. Epistemic stance in Mandarin conversation: The positions and functions of wo juede “I think”. in Yuling Pan & Daniel Kádár (eds.), Chinese discourse and interaction: Theory and practice, 12–34. London: Equinox. Google Scholar

  • Feez, Susan, Risk Iedema & Peter R.R. White. 2008. Media literacy. Surry Hills, N.S.W.: NSW Adult Migrant Education Service. Google Scholar

  • Frank, Russell. 2003. These ‘crowded circumstances’: When pack journalists bash pack journalism. Journalism 4(4). 441–458. CrossrefGoogle Scholar

  • Fuxiang, Wu (ed.). 2011. Hanyu zhuguanxing yu zhuguanhua yanjiu (Studies in subjectivity and subjectification in Chinese). Beijing: Commercial Press. Google Scholar

  • Galasinski, Dariusz. 2004. Men and the language of emotions. Basingstoke: Palgrave Macmillan. Google Scholar

  • Huan, Changeng. 2015. Journalistic stance in Chinese and Australian hard news. Sydney: Macquarie University doctoral thesis. Google Scholar

  • Huan, Changpeng. 2016. Leaders or readers, whom to please? News values in the transition of the Chinese press. Discourse, Context & Media 13. 114–121. CrossrefGoogle Scholar

  • Hunston, Susan. 2003. Frame, phrase or function: A comparison of frame semantics and local grammars. in Dawn Archer, Paul Rayson, Andrew Wilson & Tony McEnery (eds.), Proceedings of the corpus linguistics 2003 conference, 342–358. Lancaster: UCREL. Google Scholar

  • Hunston, Susan. 2011. Corpus approaches to evaluation: Phraseology and evaluative language. New York: Routledge. Google Scholar

  • Iwasaki, Shoichi & Foong Ha Yap. 2015. Editorial: Stance-marking and stance-taking in Asian languages. Journal of Pragmatics 83. 1–9. CrossrefGoogle Scholar

  • Jaffe, Alexzandra. 2009. Stance: Sociolinguistic perspectives. Oxford: Oxford University Press. Google Scholar

  • Lee, Chin-Chuan, He Zhou & Yu Huang. 2007. Party-market corporatism, clientelism, and media in Shanghai. The Harvard International Journal of Press/Politics 12(3). 21–42. CrossrefGoogle Scholar

  • Lim, Ni-Eng. 2011. From subjectivity to intersubjectivity: Epistemic marker wo juede in Chinese. in Yun Xiao, Liang Tao & Hooi Ling Soh (eds.), Studies in Chinese linguistics in the new era, 129–166. Cambridge: Cambridge University Press. Google Scholar

  • Liu, Lian & Marie Stevenson. 2013. A cross-cultural analysis of stance in disaster news reports. Australian Review of Applied Linguistics 36(2). 197–220. CrossrefGoogle Scholar

  • Lu, Jinming. 2003. 现代汉语语法研究 [Studies on contemporary Chinese grammar]. Beijing: Pekin University Press. Google Scholar

  • Martin, James R. 2000. Beyond exchange: Appraisal systems in English. in Susan Hunston & Geoff Thompson (eds.), Evaluation in text: Authorial stance and the construction of discourse, 142–175. Oxford: Oxford University Press. Google Scholar

  • Martin, James R. & Peter R.R. White. 2005. The language of evaluation. Basingstoke: Palgrave Macmillan. Google Scholar

  • Mindich, David. 1998. Just the facts: How objectivity came to define American journalism. New York: New York University Press. Google Scholar

  • Perrin, Daniel. 2012. Stancing: Strategies of entextualizing stance in newswriting. Discourse, Context & Media 1(2-3). 135–147. CrossrefGoogle Scholar

  • Peters, Chris. 2011. Emotion aside or emotional side? Crafting an ‘experience of invovlement’ in the news. Journalism 12(3). 297–316. CrossrefGoogle Scholar

  • Pounds, Gabrina. 2010. Attitude and subjectivity in Italian and British hard-news reporting: The construction of a culture-specific ‘reporter’ voice. Discourse Studies 12(1). 106–137. CrossrefGoogle Scholar

  • Reinemann, Carsten, James Stanyer, Sebastian Scherr & Guido Legnante. 2011. Hard and soft news: A review of concepts, operationalizations and key findings. Journalism 13(2). 221–239. Google Scholar

  • Rosen, Jay. 1999. What are journalists for?. New Haven, CT: Yale University Press. Google Scholar

  • Schudson, Michael & Chris Anderson. 2009. Objectivity, professionalism, and truth seeking in journalism. in Karin Wahl-Jorgensen & Thomas Hanitzsch (eds.), The handbook of journalism studies, 88–101. New York & London: Routledge. Google Scholar

  • Stenvall, Maija. 2008. On emotions and the journalistic ideals of factuality and objectivity - Tools for analysis. Journal of Pragmatics 40. 1569–1586. CrossrefGoogle Scholar

  • Stenvall, Maija. 2014. Presenting and representing emotions in news agency reports. Critical Discourse Studies 11(4). 461–481. CrossrefGoogle Scholar

  • Su, Lily. 2004. Subjectification and the use of the complementizer SHUO. Conccentric: Studies in Linguistics 30(1). 19–40. Google Scholar

  • Tao, Hongyin. 2003. Phonological, grammatical, and discourse evidence for the emergence of zhidao constructions in mandarin conversation. Zhongguo Yuwen [中国语文] 4(295). 291–302. Google Scholar

  • Tao, Hongyin. 2007. Subjectification and the development of special-verb existential/presentative constructions. Language and Linguistics 8(2). 575–602. Google Scholar

  • Thompson, Geoff. 2014. Affect and emotion, target-value mismatches, and Russian dolls. in Geoff Thompson & Laura Alba-Juez (eds.), Evaluation in context, 47–66. Amsterdam: John Benjamins. Google Scholar

  • Thomson, Elizabeth, R.R. White Peter & Philip Kitley. 2008. “Objectivity” and “hard news” reporting across cultures: Comparing the news report in English, French, Japanese and Indonesian journalism. Journalism Studies 9(2). 212–228. CrossrefGoogle Scholar

  • Tuchman, Gaye. 1972. Objectivity as strategic ritual: An examination of newsmen’s notions of objectivity. American Journal of Sociology 77. 660–679. CrossrefGoogle Scholar

  • Van Dijk, Teun A. 1998. Ideology: A multidisciplinary Study. London: Sage. Google Scholar

  • Vertommen, Bram, Astrid Vandendaele & Van Praet. Ellen. 2012. Towards a multidimensional approach to journalistic stance. Analyzing foreign media coverage of Belgium. Discourse, Context & Media 1(2). 123–134. Google Scholar

  • Wahl-Jorgensen, K. 2013b. Subjectivity and story-telling in journalism: Examining expressions of affect, judgement and appreciation in Pulitzer prize-winning stories. Journalism Studies 14(3). 305–320. CrossrefGoogle Scholar

  • Wahl-Jorgensen, Karin. 2013a. The strategic ritual of emotionality: A case study of Pulitzer prize-winning articles. Journalism 14(1). 129–145. CrossrefGoogle Scholar

  • Wang, Zhenhua. 2001. 评价系统及其运作 [Appraisal systems and their operations: A new development in the systemic functional linguisitcs]. Journal of Foreign Languages 6. 13–20. Google Scholar

  • White, Peter R. R. 1997. Death, disruption and the moral order: The narrative impulse in mass-media hard news reporting. in Francis Christie & James R. Martin (eds.), Genres and institutions: Social processes in the workplace and school, 101–133. London: Cassell. Google Scholar

  • White, Peter R. R. 1998. Telling media tales: The news story as rhetoric. Sydney: The University of Sydney doctoral thesis. Google Scholar

  • Wodak, Ruth & Teun A. Van Dijk (eds.). 2000. Racism at the top. Parliamentary discourses on ethnic issues in six European states. Klagenfurt: Drava. Google Scholar

  • Wu, Canzhong. 2009. Corpus-based research. in Michael A.K. Halliday & Jonathan Webster (eds.), Continuum companioon to systemic functional linguistcs, 128–142. London: Continuum. Google Scholar

  • Wu, Haiping. 2015. Encoding subjectivity with totality: A corpus-based study of [zhengge yi (CL) + X] in Mandarin. Journal of Pragmatics 83. 27–40. CrossrefGoogle Scholar

  • Wu, Ruey-Jiuan Regina. 2004. Stance in talk: A conversation analysis of mandarin final particles. Amsterdam/Philadelphia: John Benjamins. Google Scholar

  • Xu, Xiaoying & Jianhua Tao. 2003. A study of affect in Chinese. Proceedings of the 1th Chinese conference on affective computing and intelligent interaction, 199–205. Google Scholar

  • Yang, Ying & Foong Ha Yap. 2015. “I am sure but I hedge”: Fear expression kongpa as an interactive rhetorical strategy in Mandarin broadcast talk. Journal of Pragmatics 83. 41–56. CrossrefGoogle Scholar

  • Zhao, Yuezhi. 1998. Media, market, and democracy in China: Between the Party line and the bottom line. Urbana: University of Illinois Press. Google Scholar

  • Zhao, Yuezhi. 2008. Communication in China: Political economy, power, and conflict. Lanham, Maryland: Rowman & Littlefield Publishers. Google Scholar


  • 1

    Following the convention, this paper uses small caps to indicate subsystems of the appraisal framework (e.g. judgement), whereas scare quotes will be adopted to indicate appraisal features (e.g. “capacity”). 

  • 2

    The scandal was a food safety incident in China. It involved tens of thousands of infants being hospitalized after being fed infant formula adulterated with melamine. 

  • 3

    www.zccw.info, accessed 31 January 2014. 

  • 4

    Word count was produced based on the MyZiCiFreq. Tool (http://www.cncorpus.org/Resources.aspx, accessed 1 March 2014). 

  • 5

    The percentage was calculated by dividing the number of the occurrence of each judgement feature by the total occurrence (i.e. 825). This rule also applies to the following tables and figures. 

  • 6

    http://www.gapp.gov.cn, accessed 1 September 2014. 

About the article

Changpeng Huan

Changpeng Huan completed his PhD in linguistics from Macquarie University, Sydney, under the supervision of the late Emeritus Professor Chris Candlin, and is currently Lecturer at Shanghai Jiao Tong University, Shanghai, People’s Republic of China. His research interests include discourse studies, evaluative language, corpus linguistics, and linguistic ethnography. His most recent publications appear in Applied Linguistics, Critical Discourse Studies, Discourse & Communication, Discourse, Context & Media, Discourse & Society, and Journal of Pragmatics, as well as in edited volumes (Palgrave Macmillan).

Published Online: 2017-12-08

Published in Print: 2017-12-20

The publication of the article was sponsored by the Humanities and Social Sciences Research Grant (Culture research Project [grant number 16JCWH13]) at Shanghai Jiao Tong University, and the Shanghai Pujiang Program (grant number 17PJC058).

Citation Information: Text & Talk, Volume 38, Issue 1, Pages 23–45, ISSN (Online) 1860-7349, ISSN (Print) 1860-7330, DOI: https://doi.org/10.1515/text-2017-0029.

Export Citation

© 2018 Walter de Gruyter GmbH, Berlin/Boston.Get Permission

Comments (0)

Please log in or register to comment.
Log in