RSS feed

Epistemic modality in translated and non-translated English court judgments of Hong Kong: A corpus-based study

Yanfang Su, The Hong Kong Polytechnic University
Kanglong Liu, The Hong Kong Polytechnic University
Andrew K. F. Cheung, The Hong Kong Polytechnic University

ABSTRACT

Court judgments serve as important precedents for future judicial decision-making in common law systems. The legal meanings of judgments are conveyed by specific linguistic devices, among which epistemic modality plays an important role in indicating the probability of propositions to construct convincing arguments and recognise potential differing opinions (Abbuhl 2006). This study adopts a corpus-based approach to compare different categories of epistemic modality in translated and non-translated English court judgments in Hong Kong. Based on the framework put forward in Halliday and Matthiessen (2004), epistemic modality is categorised by orientation, value, and linguistic realisation. The findings reveal that the non-translated court judgments are characterised by a higher proportion of epistemic modality. In terms of orientation, the two corpora also exhibit some different distribution patterns. As for value, the overwhelming proportion is represented by median-value epistemic modality markers, whereas high-level modality markers are least represented in both corpora. Some variations related to the linguistic realisation of epistemic modality were also observed between the two corpora, which could mainly be attributed to the influence of the Chinese source texts, the translation process and different judicial thinking.

KEYWORDS

Epistemic modality, court judgments, translated judgments, corpus-based, value, orientation.

1. Introduction

A court judgment is a written decision made by the court to resolve disputes which regulates the rights and obligations of the parties in a lawsuit (Atkinson 2002:1). Judgments are an integral part of judicial procedure and serve as precedents for future judicial decision-making in common law systems (Garner 2001; Hinkle et al. 2012). As a distinct type of legal discourse, the way different linguistic devices are employed to convey specific legal meanings in judgments has always been a central concern of cross-disciplinary studies of law and language. In particular, Hong Kong, as the only common-law jurisdiction where both Chinese and English are official court languages, is especially worth studying from both a linguistic and legal perspective (Cheng and He 2016)1. Prior studies have provided valuable insight into the history of Hong Kong’s bilingual judiciary and problems of operating in two official court languages from the perspective of translation (Poon 2006; Cheng and He 2016). Several attempts have also been made to compare the legal meanings and judicial powers conveyed by judgments in Hong Kong and other jurisdictions, such as Scotland, the Chinese mainland, and Taiwan (Wong 2006; Cheng 2010; Cheng and Cheng 2014). There is a paucity of research comparing the different linguistic devices used in original and translated court judgments in Hong Kong. One device that deserves particular attention is the use of epistemic modality, i.e., the speaker or writer’s view of the probability of a proposition (Palmer 1986; Halliday and Matthiessen 2004). In legal discourse, epistemic modality is critical for the logic of reasoning and the prevention of challenges from people holding different opinions (Abbuhl 2006; Hinkle et al. 2012). While previous studies have addressed the variation of epistemic modality in non-translated civil judgments (Cheng and Cheng 2014) and have compared translated and non-translated judgments within the European context (Koźbiał 2020; Szczyrbak 2017), there remains a gap in understanding the specific nuances of epistemic modality in the context of Hong Kong courts, which operate within a common law system, and the challenges involved in translating court judgments within this legal framework. To address this research gap, our study aims to conduct a systematic comparison of epistemic modality in two genres, namely translated and non-translated judgments in Hong Kong courts. By doing so, we seek to enhance our understanding of the underlying judicial reasoning reflected in the use of epistemic modality in Hong Kong's common law system, while also shedding light on the potential issues and constraints associated with translating court judgments within this specific legal context.

2. Literature review
2.1 Judgments and translation

A court decision is a statement of the court's decision and reasons for sentencing in a particular case, which consist of the revelation of facts, references to the relevant law, the application of the law to the facts, and the conclusions made (Atkinson 2002). As part of the judicial process, court judgments are a fundamental element of the trial procedure (Hinkle et al. 2012). The primary function of court judgments is to resolve disputes between litigants by establishing a dialogue between judges, litigants, and witnesses (Aldisert 2009). Moreover, court judgments also perform the function of clearly expressing the thinking of the judges, justifying the decisions to the parties and the lawyers, interpreting the law, informing the public, and providing facts and reasoning as references for the court (Atkinson 2002; Biel et al. 2019). In the common law system, court judgments also serve as precedential references for judicial decision-making in future cases (Garner 2001), and thus are the cornerstone of the legal doctrine stare decisis (Leflar 1961; Aldisert 2009).

All the functions of judgments are achieved through flexible but patterned linguistic features (Maley 1994; Cheng and He 2016). Compared with other types of legal discourse such as legislation which represents a more strictly institutionalised type of language (Bhatia et al. 2004), court judgments enjoy a higher degree of diversity. The diction of judgments determines how effectively the legal meanings can be conveyed in court judgments (Leflar 1961). In view of the uniqueness and importance of court judgments in common law jurisdictions, researchers have long taken an interest in studying the grammatical and lexical devices embodied in them in order to find out how judicial power and thinking are exercised in this type of legal discourse (Leflar 1961; Cheng 2010). A plethora of empirical studies have been conducted to identify and explain the distinctive linguistic features of court judgments. For example, Bhatia et al. (2004) analysed the distribution of the top-4 high frequency verbs in 105 civil and criminal law cases in the common law system and found that different moves in the genre showed a preference for different verbs. Hobbs (2007) analysed judges’ use of the language of humour in three legal cases. He argued that the use of humour of judgments can serve the pragmatic function of punishing the ‘wrongdoers’ and alerting the lawyers to behave in a more prudent manner. While many of the previous studies on judgments are descriptive in nature involving the analysis of typical cases, Cheng (2010) adopted a corpus-based genre analysis of court judgments in Hong Kong, Taiwan, and the Chinese mainland, and found that the Hong Kong judgments displayed more diversity compared with the judgments of the other two jurisdictions. Cheng’s study indicates that the linguistic features of the Hong Kong judgments, both translated and non-translated, deserve more in-depth exploration to investigate how judicial thinking in court judgments is realised and transferred from one language to another.

Although it represents an important genre in bilingual jurisdiction systems, the translations of judgments have not been extensively studied (Cheng 2010). Considering that mistakes in translation of judgments can jeopardise the legal meanings conveyed in the judgments and potentially have a knock-on effect for the jurisdiction (Wugalter 2016), the importance of the accurate and appropriate translation of judgments should never be neglected. This is especially the case for jurisdictions that practise legal bilingualism. In this regard, Hong Kong as the only jurisdiction administering a common law system in both English and Chinese is worth our attention (Leung 2004; Cheng and He 2016). The use of Chinese in Hong Kong as an official language has challenged the traditionally predominant role of English in the common law system (Poon 2006). Although translated court judgments in Hong Kong do not have the same legal status as the original non-translated judgments, they serve multiple purposes in improving the bilingual legal system, including assisting foreign judges in preparing judgments in both languages, enabling judges in other jurisdictions to understand judgments written in Chinese, specifying the obligations and rights for parties of disputes who cannot understand the original language of judgments, and informing the public or media of their interests (Judiciary Administration of Hong Kong 2003). However, the translation quality of judgments between English and Chinese has been a central concern for the practice of legal bilingualism in Hong Kong (Poon 2006). While English has developed a more sophisticated and technical legal language than Chinese (Chan 2007), it is cautioned that these rich legal meanings cannot always be effectively conveyed when translated from English into Chinese (Yeung 2003). Poon (2006) probed into the difficulties in translations of judgments in Hong Kong and proposed that translators should use plain language in translating certain terms from English into Chinese. From a sociosemiotic perspective, Cheng and He (2016) expressed greater confidence in achieving semantic equivalence in the translation of court judgments and suggested that subsequent researchers could explore how legal meanings are transferred through translation.

To recap, previous studies on translated judgments in Hong Kong were largely not based on systematic investigation of empirical materials and relied strongly on the researchers’ own subjective evaluations. To bridge the gap, more corpus-based comparisons of original and translated and judgments are needed to understand how legal meanings are expressed or influenced by the specific linguistic features (cf. Biel 2010).

2.2 Epistemic modality: concepts and research in legal discourse

Epistemic modality expresses the degree of probability the speaker assigns to certain statements through the use of certain linguistic features (Palmer 1986). Researchers have proposed various categorisation methods in order to understand the meaning of different epistemic modalities (Palmer 1986; Hyland 1998). In systemic functional grammar (SFG), epistemic modality can be classified based on the dimensions of orientation and value (Halliday and Matthiessen 2004: 128). Specifically, orientation concerns “how far the speaker overtly accepts responsibility for the attitude being expressed” and “the extent to which the speaker openly accepts responsibility for the subjective assessments being expressed” (Thompson 2014: 73-75). Value, on the other hand, refers to “the degree of belief a speaker has in the validity of a proposition” (Thompson 2014: 73). Epistemic modality can be categorised into low, median, and high values, which represent the meaning of ‘possibly’, ‘probably’, and ‘certainly’, respectively. Compared with other methods, the categorisation of SFG is both systematic and operational, and thus has been widely applied in a range of studies (e.g., Cheng and Cheng 2014; Yang et al. 2015; Zhang and Cheung 2022).

Epistemic modality plays an important role in legal discourse, such as helping the writer to construct cogent flow of reasoning and acknowledging potentially deviating opinions on the part of readers (Abbuhl 2006). More importantly, the use of different categories of epistemic modality is closely tied to the standard of proof in a jurisdiction, which refers to the level of evidence required to establish the validity of legal claims or assertions (Cheng and Cheng 2014). This connection is significant because it can affect the precedent value of decisions for future judicial decision-making (Hinkle et al. 2012). At the same time, understanding epistemic modality and making use of it appropriately can be challenging, as it exhibits diverse forms and usages (Lewin 2005) and is subject to the influence of genre and socio-cultural factors (Tessuto 2011). Considering the importance of epistemic modality and the difficulties of its appropriate use in legal discourse, researchers have employed a corpus approach to tackle this issue. For example, Greco (2018) investigated the evidential and epistemic modality in the witness testimony of Italian criminal trials and identified the influence of various social and cultural factors in its use. Cheng and Cheng (2014) adopted an SFG categorisation to examine how epistemic modality was employed in court judgments in Hong Kong and Scotland. They found that the two jurisdictions showed similarities in terms of the criteria in employing epistemic modality in judgments. While Cheng and Cheng (2014) investigated judgments originally written in English, Koźbiał (2020) expanded the scope of the analysis and found that the use of epistemic modality was salient both in EU judgments translated into English and Polish and in non-translated Polish judgments passed by the Supreme Court of Poland, confirming epistemic modality as a generic feature of court judgments. However, the EU judgments and Polish judgments show different distribution patterns in terms of the types of epistemic markers. So far, the use of epistemic modality in translated and non-translated judgments has remained largely underexplored, especially between language pairs that are genetically more distinct from each other, such as Chinese and English.

Based on the foregoing literature review, we can see there remains a gap in the investigation of epistemic modality in translated judgments, particularly in the Hong Kong context. Therefore, this study aims to compare how epistemic modality is represented in translated as opposed to non-translated English court judgments in Hong Kong. The three research questions are as follows:

RQ1. What are the distribution patterns of linguistic devices that express epistemic modality of different values and orientations in translated and non-translated English court judgments in Hong Kong?

RQ2. What are the similarities and differences, if any, in the distribution patterns of epistemic modality between the two types of court judgments?

RQ3. If the distribution patterns vary or are similar between the two types of judgments, what are the possible factors accounting for such variations or similarities?

3. Methods
3.1 Corpora and court judgments

The current study analyses data from two corpora, i.e., Corpus of Non-translated Hong Kong Court Judgments (hereinafter HKN), Corpus of Translated Hong Kong Court Judgments (hereinafter HKT). Each corpus consists of 200 judgments from the High Court of Hong Kong published between 2000 to 2020 with a length between 2000 words to 6000 words. All the judgments were randomly extracted from the Legal Reference System of the Judiciary of Hong Kong (https://legalref.judiciary.hk/lrs/common/ju/judgment.jsp). More specifically, HKN consists of judgments originally written in English and HKT consists of judgments translated from Chinese into English for “publication and reporting” purposes (Judiciary Administration of Hong Kong 2003: 3). Details of the two corpora are presented in Table 1.

Corpus

Texts

Time Span

Total Token

Mean Token

Standard Deviationa

HKN

200

2000-2020

755,633

3785

1377

HKT

200

2000-2020

568,389

2842

566

a The legal judgments in the corpora may vary significantly in terms of sophistication, with some cases being relatively simple, while others involve multiple issues or intricate legal arguments. This variation in case sophistication can lead to a wide range of text lengths, thereby contributing to the high standard deviation.

Table 1. Corpus Design

3.2 Data analysis

The corpus linguistic approach was adopted for data analysis. Wordsmith 8.0 was used for retrieval of word frequencies and distribution of n-grams, as well as for concordance analysis. As a first step, a single word frequency list (18,013 words) and a list of 3- and 4-grams that appeared at least three times and were found in a minimum of two judgments (57,404 n-grams) were retrieved for both corpora. The researchers then identified a list of 112 epistemic-related items in the two lists by making reference to the epistemic marker lists proposed by Cheng and Cheng (2014), Vass (2017), and Koźbiał (2020), which contained detailed lists of epistemic markers in legal contexts. Further manual analysis was carried out to distinguish epistemic markers from non-epistemic markers. For example, instances where may was used as a deontic marker of obligation were excluded from the analysis. After identifying all epistemic modality markers, concordance analysis was conducted to categorise them in terms of orientation and value levels based on the SFG framework (Halliday and Matthiessen 2004). The linguistic realisations of epistemic modality in relation to the dimension of orientation were also annotated based on the categorisation of Yang et al. (2015: 3), as shown in Table 2.

After data annotation, the normalised frequencies per million words for each epistemic modality category in the two corpora were calculated to answer RQ1. To address RQ2, the frequency distributions of different categories of epistemic modality in HKN and HKT were compared. Log-likelihood tests were performed to determine the significance of the differences between the frequency distributions of epistemic modality in the two corpora. The log-likelihood values were compared to critical values to assess significance, with values above certain thresholds indicating significance. Specifically, a log-likelihood value of 3.84 or above at p < 0.05, 6.63 or above at p < 0.01, 10.83 or above at p < 0.001, and 15.13 or higher at p < 0.0001 were considered significant. Building upon the findings, we address RQ3 in the discussion.

Orientation

Linguistic Property

Example

Explicit subjective (ES)

Predicate with a first-person subject (VERB1)

We believe that there was unresolved psychological problem... (HKN)

Prepositional phrases with a first-person pronoun (PHRASE)

In my opinion, an ordinary reader of the e-mails would understand the statement in question. (HKN)

Implicit subjective (IS)

Modal verbs (MODAL)

The Applicant might not have any premeditated plan to deliberately get the Complainant to drink... (HKT)

Explicit objective (EO)

Nouns of modality (NOUN)

From its terms, the probability is that he did not do so... (HKN)

Lexical verbs with an inanimate subject (VERB2)

The content of the record of interview (exhibit P33) really suggested that the applicant had admitted her guilt... (HKT)

projecting clauses of relational or attributive process (PROJ)

It is likely that real estate development is not included as one of the permitted business activities for the plaintiff. (HKN)

Implicit objective (IO)

Modal adjuncts (ADJUN)

… and perhaps PW1 had misunderstood his actions. (HKT)

Adjectives (ADJEC)

In my judgment the risk is not such as makes his account unlikely or improbable (HKN).

Expansion of predicators or modal verbal group complex (PRED)

Direct proof that the neglect, ill-treatment, etc, did in fact, or was likely to, cause unnecessary suffering or injury to health is not required. (HKT)

Table 2. Orientation and linguistic property of epistemic modality
(adapted from Yang et al. 2015: 3)

4. Results
4.1 Orientation of epistemic modality

Both HKN and HKT were found to contain all four epistemic modality orientations. Some similarities, as well as differences, were observed in the distribution patterns of the two corpora (see Table 3). As indicated by the normalised frequency, HKN contains significantly more instances of epistemic modality than HKT. Both corpora, particularly HKT, are dominated by the use of subjective epistemic modality, and the use of explicit objective epistemic modality is underrepresented in both corpora.

4.1.1 Subjective epistemic modality

Explicit subjective epistemic modality accounted for the largest proportion among the four orientations in both corpora. The normalised frequency of explicit subjective epistemic modality in HKT is lower than in HKN. We conducted detailed concordance analysis to examine how the explicit subjective epistemic modality is expressed differently in the two corpora.

Orientation

Linguistic
Property

HKN

HKT

LLl

F (%)a

NFb

F (%)

NF

Explicit Subjective

VERB1c

759 (20.70%)

1004

346 (15.76%)

609

62.89****

PHRASEd

324 (8.84%)

429

395 (18.00%)

695

41.77****

Implicit Subjective

MODALe

915 (24.95%)

1211

653 (29.75%)

1149

1.06

Explicit Objective

NOUNf

177 (4.83%)

234

22 (1.00%)

40

97.38****

VERB2g

70 (1.91%)

93

57 (2.60%)

100

0.20

PROJh

349 (9.52%)

462

129 (5.88%)

227

52.18****

Implicit Objective

ADJUNi

582 (15.87%)

770

436 (19.86%)

767

0.00

ADJECj

98 (2.67%)

130

56 (2.55%)

99

2.75

PREDk

393 (10.72%)

520

101 (4.60%)

178

111.23****

Total

3667 (100%)

4853

2195 (100%)

3862

72.86****

aF(%) represents raw frequency of the epistemic modality and the proportion of the category in relation to total epistemic modality
bNF represents normalised frequency of epistemic modality per million words
cVERB1represents predicate with a first-person subject
dPHRASE represents prepositional phrases with a first-person pronoun
eMODAL represents modal verbs
fNOUN represents nouns of modality
gVERB2 represents lexical verbs with an inanimate subject
hPROJ represents projecting clauses of relational or attributive process
iADJUN represents modal adjuncts realised by adverbial groups or prepositional phrases
jADJEC represents adjectives
kPRED represents expansion of predicators or modal verbal group complex
lThe log-likelihood value (LL) is used to measure significance. The symbols “*”, “**”, “***”, and “****” are used to represent significance levels at p < 0.05, p < 0.01, p < 0.001, and p < 0.0001, respectively.

Table 3. Distribution of orientations of epistemic modality in HKN and HKT

The prepositional phrases with first person-pronouns, such as in my/our view/judgment/opinion, mark the predominant linguistic realisation of explicit subjective epistemic modality in translated judgments. According to the log-likelihood tests, these phrases are found to be significantly more frequent in translated judgments compared to non-translated ones. In comparison, predicates with first-person pronouns as subject (e.g., I think, I believe), which is known as a metaphorical expansion of modality (Halliday and Matthiessen 2004: 613), dominated the non-translated judgments. Moreover, HKN and HKT differed from each other in the use of the singular and plural first-person pronouns, with the former exhibiting a clear preference for singular first-person pronoun and the latter showing a balanced use of both singular and plural pronouns. For example, 108 instances (raw frequency) of in my view and 126 instances of in our view were identified in HKT, whereas 192 instances of in my view and none of in our view were found in HKN. Examples (1) and (2) are two typical examples extracted from HKN and HKT. In the non-translated judgment, the judge used a predicate with the singular first-person pronoun I as subject to represent possibility, which also showed his own process of judgment. On the other hand, the prepositional phrase with a plural pronoun our was used in the translated judgment, indicating the decision was made by the court instead of the judge alone. The High Court in Hong Kong consists of the Court of Appeal and the Court of First Instance (https://www.judiciary.hk/en/court_services_facilities/hc.html#1). Typically, judgments of the Court of Appeal are composed of three judges, while judgments of the Court of First Instance are issued by a single judge. The difference in the use of first-person pronouns in translated and non-translated English judgments may be determined by the nature of the two corpora. Since only judgments with jurisprudential value are translated, the translated judgments may contain a larger proportion of judgments from the appellate courts than the non-translated judgments.

(1) I think that this is a useful test and when applied to the facts before me, I have no doubt that the party’s dominant intention was to settle disputes. (HKN)
(2) In our view, although the charges Wong Cheung and Chau Kei faced are slightly different, there is no significant difference between their culpability. (HKT)

Expression

HKN

HKT

LLc

Fa

NFb

F

NF

may

431

570

283

497

3.18

might

245

324

243

428

9.30**

must

150

199

77

135

7.70**

could

56

74

36

63

0.55

can

33

44

14

25

3.44

aF(%) represents raw frequency of the epistemic modality and the proportion of the category in relation to total epistemic modality
bNF represents normalised frequency of epistemic modality per million words
cThe log-likelihood (LL) value indicates the significance level, with ** denoting significance at p < 0.01.

Table 4. Raw frequencies and normalised frequencies of implicit subjective epistemic modality

While the use of explicit subjective epistemic modality overtly shows personal involvement in the judgment, the use of implicit subjective epistemic modality, often realised by modal verbs, conceals the use of first-person pronouns to express subjectivity.

Table 4 presents the raw frequencies and normalised frequencies of the modal verbs that were identified to convey epistemic modality in the corpora. The results showed that may was most frequently used to indicate epistemic possibility in court judgments in both corpora, followed by might, must, could, and can, which all show different proportions and frequencies between the two corpora. It is important to note that instances of may used as deontic markers of obligation were excluded from the analysis. Specifically, while many instances of may and must were found in both corpora, not all of them were used to convey epistemic possibility. On the other hand, all the instances of might identified in both corpora were used to express epistemic modality. As for the high-value modal verb must, less than 20% of the instances were found to express certainty in both corpora (see Example 3), and more instances of must are used to express deontic modality indicating requirement or obligation (see Example 4). Generally speaking, compared with HKN, the use of the modal verb must that expresses certainty is underrepresented and might that expresses uncertainty is overrepresented in HKT. In addition, very few instances of could and can were identified to be used as epistemic modality markers in either corpus (see Example 5).

(3) Although the fruit knife was wrapped, it must have caused people at the scene to fear for their personal safety. (HKT)
(4) The first is that the prosecution must prove that the child defendant did the act charged and that when doing that act he knew that it was a wrong act as distinct from an act of mere naughtiness or childish mischief. (HKT)
(5) Such situations at best can lead to divided loyalty and at worst can result in corruption or other questionable practices. (HKT)

4.1.2 Objective epistemic modality

Compared with subjective epistemic modality, the use of objective epistemic modality signals a certain degree of objectivity and impersonality, which also contributes to the authoritativeness of judgments. As previously discussed, the use of explicit objective epistemic modality is underrepresented in both corpora and the implicit means is preferred. Based on the statistics, explicit objective markers are less used in HKT than in HKN (see Table 3). Overall, explicit objective epistemic modality represented by projecting clauses and nouns of modality is seldom found in HKT. For example, concordance analysis shows that It appears that is regularly used by judges to make inferences from the evidence in HKN (see Example 6), but seldom used in HKT. Moreover, while nominalisations of modal adjectives such as possibility and likelihood are frequently used in HKN to evaluate a proposition, such words are not found in HKT either. As for the similarities, statistical results showed that explicit objective epistemic modality is scarcely represented by lexical verbs with inanimate subjects in both corpora.

(6) It appears that the parties are determined to occupy the time of the court with satellite litigation rather than direct their attention to the real issues between them. (HKN)

As for implicit objective epistemic modality as measured by the normalised frequencies, the frequency in HKN exceeded that of HKT (see Table 3). Results of concordance analysis showed that the proportion of different forms of linguistic realisation in the two corpora also vary. In this regard, both corpora used more modal adjuncts and less adjectives. HKT uses a higher proportion of modal adjuncts than HKN. The log-likelihood tests indicate that the expansion of predictors (e.g., be likely to) and verb group complexes (e.g., appear to, seem to) to represent implicit objective epistemic modality are salient in HKN but relatively scarce in HKT.

4.2 Value of epistemic modality

In the SFG framework, value represents another dimension of epistemic modality in addition to orientation. As Table 5 shows, median value epistemic modality is predominant in both corpora, followed by low-value and high-value epistemic modality.

The log-likelihood tests reveal a significantly higher frequency of median and high-value epistemic modality expressions in HKN compared to HKT. Furthermore, HKN exhibits a greater proportion of epistemic modality expressions with a median value compared to HKT.

Orientation

HKN

HKT

LLc

F(%)a

NFb

F(%)

NF

Low

1011 (27.57%)

1338

695 (31.66%)

1222

3.36

Median

1878 (51.21%)

2486

950 (43.28%)

1671

103.12****

High

778 (21.22%)

1029

550 (25.06%)

968

24.84****

aF(%) represents raw frequency of the epistemic modality and the proportion of the category in relation to total epistemic modality
bNF represents normalised frequency of epistemic modality per million words
cLL represents the log-likelihood value; **** indicates significance at the p < 0.0001 level.

Table 5. Distribution of different values of epistemic modality

4.2.1 Low-value epistemic modality

Linguistic expressions of low-value epistemic modality with a minimum frequency of five occurrences are shown in Table 6. Concordance analysis showed that both translated and non-translated judgments make use of similar expressions in this case, but their patterns of distribution are distinctive. The linguistic realisations of low-value epistemic modality are relatively less diverse compared to other median- and high-value realisations, with low-value epistemic modality mostly realised by the use of modal verbs.

Expression

HKN

HKT

LLa

F

NF

F

NF

may

431

570

283

497

3.18

might

245

324

243

428

9.30**

possible

108

143

35

62

21.18****

possibility/possibilities

56

74

0

0

62.82****

perhaps

52

69

17

30

10.03**

could

47

62

29

51

0.71

possibly

29

38

53

93

15.62****

…not sure…

9

12

25

44

13.08***

…not convinced

8

11

0

0

8.97**

I have doubt…

8

11

1

2

4.39*

I doubt…

7

9

1

2

3.52

… doubtful

5

7

2

2

0.62

aLL represents the log-likelihood value; * indicates significance at the p < 0.05 level, ** indicates significance at the p < 0.01 level, *** indicates significance at the p < 0.001 level, and **** indicates significance at the p <0.0001 level.
bDifferent tenses of the verbs are included.

Table 6. Raw frequencies and normalised frequencies of low-value epistemic modality expressionsb

4.2.2 Median-value epistemic modality

Table 7 illustrates that the linguistic realisations of median-value epistemic modality exhibit a greater degree of diversity. Prepositional phrases such as in my view and predicates with first-person pronouns as subjects, for example I think, are the most commonly used expressions. In addition, the nominalised form of modal adjective, probability, also appeared with a high frequency in both corpora. In particular, many instances of probability occurred in the collocation balance of probabilities (14 instances in HKN and 15 instances in HKT), which specifies the standard of proof when making judicial decisions in civil cases. 

Some differences were also identified between the two corpora. As has been pointed out in the findings regarding explicit subjective epistemic modality, some expressions of median-value modality markers with plural first-person pronouns are found frequently in HKT but absent from HKN, such as in our view, in our judgment, we are of the viewwe do not think, we thinkwe do not considerwe take the view that. Likewise, there are certain expressions that can only be found in HKN but not in HKT, such as likelihoodpresumablyI infer, and I assume that.

4.2.3 High-value epistemic modality

The occurrence of instances of high-value epistemic modality accounts for the smallest proportion of all three types in both corpora. As for the linguistic realisations of high-value epistemic modality, the expressions used are quite similar between HKN and HKT. Overall, modal adjuncts occupy a central place in both corpora (see Table 8). Our analysis also showed that certainty is also regularly expressed by projecting clauses in HKN, but to a lesser extent in HKT.


Expression

HKN

HKT

LL

F

NF

F

NF

in my view

192

254

108

190

5.98*

it seems that

147

194

14

25

93.44****

likely to

134

177

50

88

19.60****

I think

117

155

19

33

53.37****

in my judgment

100

132

80

141

0.17

in our view

0

0

126

222

213.10****

It appears that

91

120

24

42

24.86****

I do not think

86

114

21

37

26.02****

I consider…

79

105

33

58

8.63**

… appear to

78

103

33

58

8.21****

I believe that…

77

102

23

40

17.42****

…seem to…

59

78

6

11

36.31****

probability/probabilities

53

70

22

39

5.89*

probably

53

70

14

25

14.45***

I am of the view that

53

70

59

104

4.29*

... indicate that…

32

42

44

77

6.86**

I do not consider…

43

57

14

25

8.36**

in our judgment

0

0

56

99

94.71****

we are of the view…

0

0

50

88

84.56****

…suggest that…

38

50

8

14

13.65***

likelihood

36

48

0

0

40.38****

can

33

44

13

23

4.23*

we consider…

0

0

34

60

57.50****

It is likely that…

20

26

14

25

0.04

in my opinion

29

38

3

5

17.69****

we do not think…

0

0

27

48

45.66****

presumably

19

25

0

0

21.31****

I indicate…

15

20

3

5

5.68*

I take the view that

13

17

4

7

2.80

we think…

0

0

15

26

25.37****

I infer…

10

13

0

0

11.22***

we do not consider…

0

0

10

18

16.91****

the inference is that

9

12

2

5

3.05

we take the view that...

0

0

9

16

15.22****

I assume that…

8

11

0

0

8.97***

we believe that…

6

8

5

9

0.03

draw the inference

5

7

2

4

0.62

aLL represents the log-likelihood value; * indicates significance at the p < 0.05 level, ** indicates significance at the p < 0.01 level, *** indicates significance at the p < 0.001 level, and **** indicates significance at the p <0.0001 level.
bDifferent tenses of the verbs are included.

Table 7. Raw frequencies and normalised frequencies of median-value epistemic modality expressionsb

Expression

HKN

HKT

LL

F

NF

F

NF

must

150

199

77

135

7.70**

of course

107

142

66

116

1.63

apparently

85

112

23

40

22.39****

certainly

69

91

62

109

1.03

obviously

67

89

107

188

24.18****

there is no doubt that

44

58

3

5

32.12****

impossible

38

50

21

37

1.32

no doubt

32

42

12

21

4.63*

I have no doubt that

30

40

2

4

22.07****

absolutely

19

25

30

53

6.61*

I conclude that

13

17

7

12

0.52

sure

2

3

29

51

36.46****

definitely

12

16

23

40

7.36**

improbable

10

13

15

26

2.94

could not

9

12

8

14

0.12

surely

8

11

6

11

0.00

certainty

7

9

0

0

7.85**

unlikely

7

9

14

25

4.80*

evidently

6

8

3

5

0.35

we conclude that

1

1

5

9

4.17*

beyond doubt

0

0

5

9

8.46**

aLL represents log-likelihood value; * indicates significance at the p < 0.05 level, ** indicates significance at the p < 0.01 level, *** indicates significance at the p < 0.001 level, and **** indicates significance at the p < 0.0001 level.
bDifferent tenses of the verbs are included.

Table 8. Raw frequencies and normalised frequencies of high-value epistemic modality expressionsb

4.3 Relationship between orientation and value

The purpose of this section is to analyse the distribution of different orientations of epistemic modality markers over a continuum of value levels. Exploring the relationship between orientation and value of epistemic modality can provide more insight into how judges employ epistemic modality strategically to communicate confidence in their propositions and defend themselves from possible challenge.

In Table 9, we present the normalised frequency of epistemic modality across different values and orientations. The log-likelihood results reveal both similarities and variations in the distributions of epistemic modality values with respect to different orientations between HKN and HKT. The most notable variation is the median-value modality of objective orientation. It is worth mentioning that a high proportion of subjective epistemic modality expressions denote possibility and probability rather than certainty, which highlights the judge’s active role in making the propositions. For instance, Example 7 shows how judges typically state their subjective evaluation of evidence in an uncertain way, suggesting that while judicial decision-making requires reasoning based on facts and evidence, judges tend to avoid stating their propositions in an absolute or definite way. Expressing personal evaluations using low- or median-value modality also makes the proposition less falsifiable since more facts may be found to verify it (Meyer 1997). Implicit subjective epistemic modality makes up the highest percentage of low-value epistemic modality in both corpora, indicating that the judges try to downplay personal commitment in uncertain statements while necessitating a certain degree of authoritativeness. An overwhelming proportion of median-value epistemic modality in both corpora belongs to the explicit subjective category, whereas relatively few are found in the implicit subjective category. This might be a result of the tendency that the modal verb can is rarely used in legal discourse to express epistemic modality.

For propositions of high certainty, it is natural that epistemic modality is represented through an objective rather than subjective orientation, which greatly downplays the involvement of the judge while enhancing the impression that the decisions are made based on facts. In addition, both translated and non-translated judgments favour the implicit method of conveying high certainty, that is, high-value modality is expressed “in the same clause as the main proposition” (Thompson 2014: 75) in an implicit way. Example 8 shows the reasoning process of decision-making, in which the judge objectivises the statement in order to maintain an impersonal style.

(7) Under such circumstances, I doubt if PW1 could really hear clearly all the words of the abusive language, or could hear from what was said the word “beat” or words to that effect. (HKT)
(8) No doubt, the defendant has since early 2005 been requested to account for the withdrawals made from the plaintiff's account and his dealings with the money withdrawn. (HKN)

Category

Low

Median

High

HKN

HKT

LL

HKN

HKT

LL

HKN

HKT

LL

Explicit Subjective

37

5

16.77****

1329

1267

0.96

67

33

7.49***

Implicit Subjective

957

976

0.13

44

23

4.23*

210

150

6.66***

Explicit Objective

112

62

9.67**

549

200

108.33****

127

104

1.51

Implicit Objective

232

179

4.28*

564

183

128.92****

625

681

1.57

aLL represents the log-likelihood value; * indicates significance at the p<0.05 level, ** indicates significance at the p<0.01 level, *** indicates significance at the p<0.001 level, and **** indicates significance at the p<0.0001 level.
bDifferent tenses of the verbs are included.

Table 9. Normalised frequencies of epistemic modality

5. Discussion
5.1 Subjectivity and objectivity

Based on the distribution patterns of epistemic modality of different orientations, we can observe both similarities and differences in the expression of epistemic modality between translated and non-translated English judgments of Hong Kong.

Three types of epistemic modality, i.e., explicit subjective, implicit subjective, and implicit objective, are found to be represented in comparable proportions in both corpora, whereas the explicit objective type is relatively underrepresented. This is incongruent with the findings presented by Cheng and Cheng (2014) according to which subjective epistemic modality forms an overwhelming proportion in Hong Kong judgments. This inconsistency is probably due to Cheng and Cheng’s (2014) selection of relatively fewer epistemic markers than the current study. It should also be noted that Cheng and Cheng analysed only civil court judgments whereas the current study examined both civil and criminal court judgments. Moreover, the use of explicit subjective epistemic modality is found to be the principal type in both HKN and HKT, possibly because the judges’ individual assessment of the propositions framed as viewpoints instead of definitive facts can be better conveyed through such an orientation (Yang et al. 2015; Vass 2017). This salience of subjective epistemic modality in Hong Kong court judgments, to a large extent, also reflects the profound influence of UK legal culture and traditions on Hong Kong’s legal system (Liu and Zhu 2021). As Biel et al. (2022: 14) found, UK Supreme Court judgments also utilise a strikingly high frequency of first-person pronouns to express the judges’ personal beliefs and statements, which is an indication of the judges’ “authorial presence through self-references”. Subjective epistemic modality in judgments also indicates that judges may exercise judicial discretion when the precedents are not applicable to their case (Bingham 2000: 36). Under the common law system, judicial decisions supplement and constantly replenish the codified law. Generally speaking, judges follow the precedents, but they also enjoy flexibility to overturn previous cases by pointing out the errors and setting a new precedent for reference (Dainow 1966). Also, since judgments also serve as a means for communication with the public, the frequent use of subjective epistemic modality might encourage readers to participate in the discourse and openly evaluate the current state of affairs (Hinkle et al. 2012).

In comparison, the use of objective epistemic modality disguises speculative possibilities and adds a great degree of impersonality (Hinkle et al. 2012), with objectivised statements shifting the focus away from the judge’s personal evaluation. In this case, the judge’s estimation of the state of affairs appears to derive from credible facts and logical reasoning, which might also function to reduce the public’s suspicion of the judges’ subjective evaluation or intuition (Posner 1994; Yang et al. 2015). In the current study, the relative underrepresentation of explicit objective epistemic modality in both corpora indicates that the judges are active participants in the judicial reasoning process. The use of more subjective epistemic modality, rather than obscuring their role as judges, in fact highlights their authoritativeness. This is especially the case with translated judgments, indicating that the Chinese source texts put even more emphasis on the role of judges in this regard.

5.2 Degree of certainty

The analysis of the values of epistemic modality reveals that median-level epistemic modality is prevalent in both translated and non-translated English judgments of Hong Kong, followed by low-value and high-value modality, which is consistent with the findings of Cheng and Cheng (2014). The overrepresentation of median-value epistemic modalities might be related to the principle of proof “on a balance of probabilities” in civil judgments (Judiciary Administration of Hong Kong 2017). When applying the balance of probabilities, an event is deemed by the court to have occurred based on evidence that “it was more probable than not that the event occurred” (Nesson 1985: 1362). When low and median values of epistemic modality are used, the judges’ thinking is formulated with a greater degree of reservation, suggesting that the propositions are not regarded as being of absolute certainty (Yang et al. 2015). This means that a judicial decision can be made without satisfying a high standard of proof. In addition, refraining from using absolute statements in the judgments might also help to avoid potential legal challenges (Channel 1994: 157). In terms of the use of high-value epistemic modality, more definitive language shows the judge’s confidence in the proposition, which can then be used as robust precedents that can be relied on in future rulings (Hinkle et al. 2012). However, the findings of the present study differ from Koźbiał’s (2020) observation in the translated English and Polish EU judgments and non-translated Polish judgments where the majority of instances of epistemic modality were of high-value, while only a very small proportion were of low-value. The differences highlight the uniqueness of the common law system which differs from the continental law system in both legal practice and cultural background (Cheng et al. 2008). Unlike Hong Kong’s common law system, in Poland’s civil law system, judges have the responsibility to interpret a system of law which is continuous, stable, and authoritative (Dainow 1966). As a common law jurisdiction, Hong Kong is clearly distinctive from such a system in terms of how judicial reasoning is conveyed.

5.3 Differences between non-translated and translated English court judgments

This study shows divergence in the distribution of epistemic modality markers between translated and non-translated English court judgments in Hong Kong, in terms of frequency, variety, and types of epistemic statements. Such divergence is related to the concept of “textual fit”, defined by Biel (2014a: 119) as the “linguistic distance between translations and non-translations of a comparable genre”. The concept of textual fit was inspired by research carried out by Baker (1996) and Toury (1995), showing that certain features of translation texts might be expected to be shared but that other variables such as translation strategies (Biel 2014a: 119) may show differences. A study by Biel (2014b) examined the textual fit of translated EU law by comparing deontic modality in translated Polish EU law and non-translated Polish national law. She concluded that the differences between the two corpora might be attributed to factors such as legal instruments, institutionalisation, translator’s choice, genre-specific features, and socio-cultural contexts (Biel 2014b: 349). In the present study, the divergent textual fit of non-translated and translated English judgments might be influenced by a combination of different factors, such as the different writing styles of original Chinese judgments and English judgments, the genre-specific features of epistemic modality in Chinese and English, and the different judicial thinking of judges using Chinese and English.

First, epistemically modalised statements are underrepresented in the translated judgments compared with non-translated judgments. Such a divergence in textual fit can be attributed to the different writing styles between the original Chinese judgments and the English judgments. Wong and Su (2016) pointed out that Chinese court judgments in Hong Kong feature the regular use of classical Chinese expressions which are highly succinct and formal. In such a writing style, epistemic modality occurs at a relatively lower rate in court judgments originally written in Chinese, and such a low ratio might be carried forward to the translated English judgments due to the “shining-through” effect of the source text (Teich 2003: 145). In addition, the use of “omission” strategies by translators might also contribute to the underrepresentation of epistemic modality in the translated texts. For example, in one Chinese judgment, the judge used 得內在不可信, literally ‘appear to be inherently incredible’, to cast doubt on the credibility of the evidence presented by the defendant. However, in the translated version, the lexical verb顯得(literally: ‘appear’) that expresses epistemic modality was omitted in the English translation. It should be noted that the translated court judgments used in this study are mainly used for informative purposes and have no legal status (Judiciary Administration of Hong Kong 2003), and hence the purpose of translation might be a factor in this case.

Second, the translated judgments are characterised by a lesser variety and frequency of words and expressions denoting epistemic modality, which might be due to a trend towards normalisation (Baker 1996) in the translation process. For instance, the less frequent use of lexical verbs, such as appearseem, in translated judgments compared to non-translated ones, results in a lower proportion of median-value objective epistemic modality markers in the translated corpus. The normalisation trend might be related to the institutional nature of translation of judgments, which is conducted by a specified group of translators, i.e., the Court Language Section, trained to follow particular norms and regulations (Ng et al. 2021). The divergent textual fit between translated and non-translated judgments might also be related to certain genre-specific epistemic expressions prevalent in Chinese judgments. For example, no instances of nominalised forms of modal adjectives were identified in the translated legal judgments, which clearly leads to a low representation of explicit objective epistemic modality in the translated corpus. Such a phenomenon might be associated with the original Chinese texts which seldom use the nominalised forms of modal adjectives, e.g.,可能性(literally: ‘possibility’), to denote degree of certainty. Due to the absence of such terms in the source Chinese texts, a lesser variety of epistemic modality markers appears in the translations. The frequently used expressions 本庭/本席認為 (literally: ‘this court/judge thinks’) in Chinese legal judgments clearly contributes to the prevalence of explicit subjective epistemic modality markers in the translated English judgments. 本庭/本席認為 is typically translated into a predicate or prepositional phrase with first-person pronoun, such as ‘we consider’ or ‘in our/my opinion’.

Third, the differences between translated and non-translated English judgments might also reflect the different judicial thinking when judges use Chinese and English (Ng 2009). Compared with the translation which frequently employed first person pronouns I and we, the use of 本席(literally: ‘this Chair’; formal use of the first-person pronoun ‘I’ by the judge) and 本庭(literally: ‘this court’) in original Chinese judgments carries a certain degree of aloofness and emphasises the authority of the judges and the court. From a jurisprudential perspective, the judge’s individual opinions conveyed using the singular first-person pronoun I in the non-translated judgments suggest that judicial power of the judge is foregrounded in the English judgments of Hong Kong, while the court as a collective entity is foreshadowed. In comparison, Cheng (2012) found that in non-translated Chinese appellate judgments in Hong Kong, both the court’s and the judges’ voices are prominent, and the judicial control from the court is even more emphasised, which shows a divergent thinking from the judgments originally written in English. Such a judicial thinking embedded in the Chinese judgments in Hong Kong clearly has an influence on the high representation of plural first-person pronouns to express epistemic modality in the translated judgments.

6. Conclusion

This study is a corpus-based contrastive study of the epistemic modality in translated and non-translated judgments of Hong Kong. Our study has advanced the developing field of legal translation studies by enhancing the interdisciplinary understanding of law, translation studies and linguistics, as well as breaking through the Eurocentric tradition of legal translation studies to other language pairs (Biel 2018; Božović 2022). In particular, our study has bridged the current research gap in translated judgments in the Hong Kong context and deepened our understanding of epistemic modality in translated and non-translated legal judgments. Translated judgments, as a relatively underexplored text variety, exhibit distinct patterns in epistemic modality compared to non-translated judgments, including reduced variety and an overall underrepresentation of epistemic modality. Through systematic categorisation of epistemic modality into different types, including orientation and value, we have obtained a more comprehensive understanding of its expression varieties and distribution patterns in both translated and non-translated judgments. Furthermore, as Hong Kong is the only common law jurisdiction where both Chinese and English are used in legal practice, our study contributes to a deeper comprehension of bilingual judgment writing and translation in this context.

Despite the findings, some limitations of the present study should be pointed out. First, we mainly analysed a comparable corpus consisting of translated and non-translated English judgments, and the original Chinese judgments were largely left out of the picture. Future research could incorporate Chinese-English parallel texts to dig deeper into the differences between translated and non-translated court judgments and how genre-specific features of epistemic modality in Chinese affect the use of epistemic modality in English translated judgments. In terms of corpus comparability, there are also variables that could be considered in future analyses, such as the nature of the case (civil or criminal) and the court (e.g., court of last resort, court of appeal, or court of first instance). In addition, although the lists of epistemic modality expressions were generated, our study did not provide detailed analysis of the pragmatic meanings of the expressions. Future studies can also link epistemic modality to genre moves (e.g., reviewing the evidence, making judicial decisions, and clarifying the reasons) to understand how different legal meanings are achieved with epistemic modality.

References
  • Abbuhl, Rebekha (2006). “Hedging and Boosting in Advanced-level L2 Legal Writing.” Heidi Byrnes, Heather Weger-Guntharp and Katherine A. Sprang (eds) (2006). Educating for Advanced Foreign Language Capacities: Constructs, Curriculum, Instruction, Assessment. Washington D.C.: Georgetown University Press, 152-164.
  • Aldisert, Ruggero J. (2009). Opinion Writing. Durham, NC: Carolina Academic Press.
  • Baker, Mona (1996). “Corpus-based translation studies: The challenges that lie ahead.” Harold Somers (ed.) (1996). Terminology, LSP and Translation. Amsterdam: John Benjamins, 175-186.
  • Bhatia, Vijay K., Langton, Nicola M., and Jane, Lung (2004). “Legal discourse: Opportunities and threats.” Ulla Connor and Thomas A. Upton (eds) (2004). Discourse in the Professions: Perspectives from Corpus Linguistics. Amsterdam: John Benjamins, 203-231.
  • Biel, Łucja (2010). “Corpus-based studies of legal language for translation purposes: Methodological and practical potential.” Carmen Heine and Jan Engberg (eds) (2010) Reconceptualizing LSP: Online Proceedings of the XVII European LSP Symposium 2009, 1-15.
  • Biel, Łucja (2014a). Lost in the Eurofog: The Textual Fit of Translated Law. Frankfurt am Main: Perter Lang.
  • Biel, Łucja (2014b). "The textual fit of translated EU law: a corpus-based study of deontic modality." The Translator 20(3), 332-355.
  • Biel, Łucja (2018). “Corpora in Institutional Legal Translation: Small Steps and the Big Picture.” Fernando Prieto Ramos (ed.) (2018). Institutional Translation for International Governance: Enhancing Quality in Multilingual Legal Communication. London: Bloomsbury, 25-36.
  • Biel, Łucja, Koźbiał, Dariusz, and Dariusz Müller (2022). “The judicial English Eurolect: A genre profiling of CJEU judgments.” Stanislaw Goźdź-Roszkowski and Gianluca Pontrandolfo (eds) (2022). Law, Language and the Courtroom. London: Routledge, 3-25.
  • Biel, Łucja, Koźbiał, Dariusz and Katarzyna Wasilewska (2019). “The formulaicity of translations across EU institutional genres: A corpus-driven analysis of lexical bundles in translated and non-translated language.” Translation Spaces 8(1), 67-92.
  • Bingham, Tom (2000). The Business of Judging: Selected Essays and Speeches: 1985-1999. Oxford: Oxford University Press.
  • Božović, Petar (2022). "How are metaphors rendered in legal translation? A corpus-based study of the European Court of Human Rights judgments." Journal of Specialised Translation 38, 277-297.
  • Chan, Clara Ho-yan (2007). “Translated Chinese as a legal language in Hong Kong legislation.” Journal of Specialised Translation 7, 25-41.
  • Channel, Joanna Mary (1994). Vague Language. Oxford: Oxford University Press.
  • Cheng, Le (2010). “Discourse and judicial thinking: A corpus-based study of court judgments in Hong Kong, Taiwan, and Mainland China.” International Journal of Speech, Language and the Law 17(2), 295-298.
  • Cheng, Le (2012). “Attribution and judicial control in Chinese court judgments: A corpus-based study.” International Journal of Speech, Language & the Law 19(1), 27-49.
  • Cheng, Winnie, and Le Cheng (2014). “Epistemic modality in court judgments: A corpus-driven comparison of civil cases in Hong Kong and Scotland.” English for Specific Purposes 33, 15-26.
  • Cheng, Le and Lianzhen He (2016). “Revisiting judgment translation in Hong Kong.” Semiotica 209, 59-75.
  • Cheng, Le, Sin, King-kui and Ying-Long Zheng (2008). “Contrastive analysis of Chinese and American court judgments.” Critical Approaches to Discourse Analysis across Disciplines 5, 49-58.
  • Dainow, Joseph (1966). “The civil law and the common law: Some points of comparison.” The American Journal of Comparative Law 15(3), 419–435.
  • Garner, Bryan A. (2001). A Dictionary of Modern Legal Usage. New York: Oxford University Press.
  • Greco, Paolo (2018). “Evidentiality and epistemic modality in witness testimony in the context of Italian criminal trials.” Journal of Pragmatics 128, 128-136.
  • Halliday, M. A. K. and Christian Matthiessen (2004). An Introduction to Functional Grammar. London: Hodder Arnold.
  • Hinkle, Rachael K. et al. (2012). “A positive theory and empirical analysis of strategic word choice in district court opinions.” Journal of Legal Analysis 4(2), 407-444.
  • Hobbs, Pamela (2007). “Judges’ use of humor as a social corrective.” Journal of Pragmatics 39(1), 50-68.
  • Hyland, Ken (1998). Hedging in Scientific Research Articles. Amsterdam: John Benjamin.
  • Koźbiał, Dariusz (2020). “Epistemic modality: A corpus-based analysis of epistemic markers in EU and Polish judgments.” Comparative Legilinguistics 41, 39-70.
  • Leung, Matthew (2004). “Assessing Parallel Texts in Legal Translation.” The Journal of Specialised Translation 1, 1-11.
  • Leflar, Robert A. (1961). “Some observations concerning judicial opinions.” Columbia Law Review 61, 810-820.
  • Lewin, Beverly A. (2005). “Hedging: An exploratory study of authors’ and readers’ identification of ‘toning down’ in scientific texts.” Journal of English for Academic Purposes 4(2), 163-178.
  • Liu, Kanglong and Linqing Zhu (2021). “An Investigation of Norms in Legal Translation: A Corpus-Based Study of Conditional Connectives on Company Law.” Riccardo Moratto and Martin Woesler (eds) (2021). Diverse Voices in Chinese Translation and Interpreting: Theory and Practice, Singapore: Springer, 361-381.
  • Maley, Yon (1994). “The language of the law.” John Peter Gibbons (ed.) (1994). Language and the Law. London: Routledge, 11-25.
  • Merryman, John Henry, and Rogelio Pérez-Perdomo (2007). The Civil Law Tradition: An Introduction to the Legal Systems of Europe and Latin America. Stanford, CA: Stanford University Press.
  • Meyer, Paul George (1997). “Hedging Strategies in Written Academic Discourse: Strengthening the Argument by Weakening the Claim.” Raija Markkanen and Hartmut Schröder (eds) (1997). Hedging and Discourse: Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts. Berlin/New York: de Gruyter, 21-41.
  • Nesson, Charles (1985). “The Evidence or the Event? On Judicial Proof and the Acceptability of Verdicts’.” Harvard Law Review 98, 1357-1392.
  • Ng, Kwai Hang (2009). The Common Law in Two Voices: Language, Law, and the Postcolonial Dilemma in Hong Kong. Stanford, CA: Stanford University Press.
  • Ng, Eva NS, Janny HC Leung and Kelley Loper (2021). A Study on a Potential Model for Accreditation and Regulation of Interpreters and Translators in Ethnic Minority Languages in Hong Kong. https://www.eoc.org.hk/Upload/files/research-report/20211225/Study%20Report.pdf (consulted 21.12.2022).
  • Palmer, Frank Robert (1986). Mood and Modality. Cambridge: Cambridge University Press.
  • Poon, Wai-Yee Emily (2006). “The translation of judgments.” Meta: Translators’ Journal 51(3), 551-569.
  • Posner, Richard A (1994). The problems of jurisprudence. Cambridge, MA: Harvard University Press.
  • Szczyrbak, Magdalena (2017). “Modal Adverbs of Certainty in EU Legal Discourse: A Parallel Corpus Approach.” Karin. Aijmers and Diana. Lewis (eds) (2017). Contrastive Analysis of Discourse-Pragmatic Aspects of Linguistic Genres. Cham: Springer, 91-115.
  • Teich, Elke (2003). Cross-Linguistic Variation in System and Text: A Methodology for the Investigation of Translations and Comparable Texts. Berlin: Mouton de Gruyter.
  • Tessuto, Girolamo (2011). “Legal problem question answer genre across jurisdictions and cultures.” English for Specific Purposes 30(4), 298-309.
  • Thompson, Geoff (2014). Introducing Functional Grammar. 3rd Ed. London: Routledge.
  • Toury, Gideon (1995). Descriptive translation studies: And beyond. Amsterdam: John Benjamins.
  • Vass, Holly (2017). “Lexical verb hedging in legal discourse: The case of law journal articles and Supreme Court majority and dissenting opinions.” English for Specific Purposes 48, 17-31.
  • Wugalter, R. A. (2016). Legal and judgment translation in the government of Canada: The urgent need for reform. ATIO. https://acjt.ca/wp-content/uploads/2019/09/Federal-Judgment-and-Legal-Translation-Response-of-R-Wugalter-to-OCOL-Report-to-Cabinet-Pi%C3%A8ce-no-26.pdf (consulted 12.05.2022).
  • Wong, Pui-Kwong and Jinzhi Su (2016). “Language uses and norms in Chinese Legal Judgments in Hong Kong.” Global Chinese 2(1), 91-102.
  • Wong, Pui-Kwong (2006). “A comparative study of the legal judgments of Hong Kong and the Mainland China.” Language Teaching and Linguistic Studies 2, 35-41.
  • Yang, An, Shu-yuan Zheng, and Guang-chun Ge (2015). “Epistemic modality in English-medium medical research articles: A systemic functional perspective.” English for Specific Purposes 38, 1-10.
  • Yeung, C. K. (2003). “Judicial bilingualization and Chinese translation of law.” Wenhui Lu (ed.) (2003). Legal translation. Beijing: Law Press, 347–359.
  • Zhang, Yifan and Andrew K.F. Cheung (2022). “A corpus-based study of modal verbs in Chinese–English governmental press conference interpreting.” Frontiers in Psychology 13, 1065077-1065077.
Websites
  • Atkinson, Roslyn (2002). “Judgment Writing” [Conference presentation]. Magistrates Conference, Gold Coast, Australia. https://www.fja.gov.pk/files/articles/JudgmentWriting.pdf (consulted 25.05.2022).
  • Judiciary Administration of Hong Kong (2003). Translation of Judgments (LC Paper No. CB(2)1856/02-03(01)). https://www.legco.gov.hk/yr02-03/english/panels/ajls/agenda/ajag0428.htm (consulted 20.05.2022).
  • Judiciary Administration of Hong Kong (2017). General Guidelines for Preparation for Trial https://www.judiciary.hk/doc/en/court_services_facilities/guidelines_for_trial_06ed1_e.pdf (consulted 30.05.2022)
  • Data Availability Statement

    All relevant data in this study are available at https://osf.io/2j8n4/

    Biography

    一張含有 個人, 擺姿勢 的圖片  自動產生的描述Yanfang Su is currently a PhD student in the Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University. Her research interests include corpus linguistics, corpus-based translation studies, and computer-assisted language learning. She has previously published in linguistic journals such as Language Learning & Technology and System and contributed book chapters on corpus-based translation studies and language learning.

    ORCID: 0000-0002-4526-4187
    E-mail: yanfang.su@connect.polyu.hk

    https://www.polyu.edu.hk/cbs/web/en/img/?hash=eb39efa7f7d5542115fca1baec116124.jpg Kanglong Liu is Assistant Professor at the Department of Chinese and Bilingual Studies of the Hong Kong Polytechnic University. He specialises in corpus-based translation studies and his main interests include empirical approaches to translation studies, translation pedagogy and corpus-based translation research. He is currently Associate Editor of Translation Quarterly, the official publication of the Hong Kong Translation Society.

    ORCID: 0000-0003-3962-2563
    E-mail: kl.liu@polyu.edu.hk

    一張含有 個人, 牆, 室內, 握住 的圖片  自動產生的描述Andrew K. F. Cheung is Associate Professor at the Department of Chinese and Bilingual Studies of the Hong Kong Polytechnic University. He serves as a member of editorial boards of multiple journals. He is also a member of International Association of Conference Interpreters (AIIC). His research interests include corpus-based interpreting and translation studies, quality perception of conference interpreting and court interpreting.

    ORCID: 0000-0003-0848-1980
    E-mail: andrew.cheung@polyu.edu.hk

    Notes

    Note 1:

    For details of the official languages of Hong Kong, their status and use, please refer to the Official Languages Ordinance (https://www.elegislation.gov.hk/hk/cap5). Return to this point in the text