Syntactic Complexity in Legal Translated Texts and the Use of Plain English: A Corpus-Based Study

In applied linguistics research, syntactic complexity has been used extensively as a measure to capture the complexities of academic writing and texts produced by L2 English writers. Syntactic complexity focuses on the length of the unit of production, the extent of clausal embedding, types of structures, and the nuanced nature of the structures within the production units (Lu, 2011, p. 36). In his paper on Arabic-English bi-directional translation, Al-Jabr (2006, p. 206) directs our attention to three key features that are argued to make up syntactic complexity in the translation domain. These include the linguistic features of a given language (e.g., preference for coordination/subordination), the genre of the text (e.g., legal/journalistic writing) and the lexical or grammatical preferences of the translators. Legal translation is known as the subgenre of TS. Likewise, the complexity of the language of law warrants further investigation as legal texts have increasingly begun to find their way into the social arena. For instance, in recent years, legal translated texts have started to be published, circulated and read at international organizations, in public services and in the private sector (Biel et al., 2019). Legal translation, which is the subject of this paper, implies the "language of the law of England, America" (Tiersma, 2008, p. 7; Giczela-Pastwa, 2019; Afzaal, 2022), of other countries in which English is an official language and of countries that interact internationally within the arenas of business, international affairs and trade. The present study sought empirically informed insights into the legal translation of company laws while comparing Chinese-translated corpora with existing British corpus and Hong Kong Company translated laws in syntactic complexity. The focus was on the legal translation and plain structure used by non-native English translators of legal terms in Hong Kong and China. With the research gap and widespread usage of LT in mind, the study's goal was to identify the elements that impact legal translation, compare legal translated texts with non-translated materials and determine how far legal correspondence translation has made communication effective. Ramos and Cerutti (2022) suggest that "due to the expansive nature of the legal system and its various subfields and crossing conceptual networks, which address almost every facet of human existence, legal writings cover a wide range of topics conceptually", (p. 22). More broadly, this paper examines the syntactic structure of languages used in legal translated texts and investigates whether translated texts and translation norms through the application of the analytic lens of Kachru's (1985) Three Circles Model of World Englishes and norm orientation) interact and experience a transformation. Kachru's Three Circles model provides a theoretical underpinning to categorize the regions based on the English Language spoken in these countries. For instance, the inner circle offers insights from the region of English native speakers such as people of the United Kingdom or the USA. The outer circle focuses on the circle includes countries under the direct control of the UK as colonies where English is the official language. Lastly, the expanding circle focuses on the foreign language as norm-dependent and mainly depends on the inner circle. Therefore, the study undertakes plain language because plain English is an initiative that emphasizes the accessibility of communication. Even a person who is not trained in law should understand what is being said in a contract, not just the legal professionals who are drafting it. The study also highlights the difficulty of archaic legal jargon and complex syntax of legal texts and advocates using plain language, using words that everyone can comprehend (Mattila, 2016; Adler, 2012, Yu, 2021).

Studies of syntactic complexity in translation A large body of literature on syntactic complexity studies has investigated the academic writing of L2 and L1 learners in terms of syntactic complexity. Still, fewer studies have provided insights into the use of syntactic complexity in translated texts produced by non-native English speakers. As part of his research, Lu et al. (2020) looked at the indicators of syntactic complexity in three components that were employed in this study for instance, the Biber Tagger proposed by Biber et al., (1999), the Coh-Metrix proposed by McNamara et al., (2014), and the L2 Syntactic Complexity Analyzer initiated by (Lu, 2010). In addition, Lu (2010) created the L2SCA, which uses 14 metrics to computerize syntactic complexity analysis of L2 English texts in the target language. Pallotti (2015) considers complexity as polysemous in second language acquisition, and Bulté and Housen, 2012) define complexity as cognitive or linguistic. Lu (2017) defined complexity in syntactic structures 'as the diversity and level of sophistication of syntactic structures used in textual creation', (p. 5, 2017). Scholars in the field of legal translation have demonstrated a particular interest in the processes of multilingual legislative drafting that are utilized in various organizations on both the national and international levels (Dillion, 2022). According to Mancilla et al. (2015), non-native speakers employ more coordinating and complicated phrases but far less subordinated, than native speakers, even though expert non-native speakers writing approaches native speaker writing in terms of the quantity of compliance. Ansarifar et al. (2018) touched on the phrasal complexity within the abstract sections of theses produced by L1 learners. They (2018) further identified that "the phrasal complexity, measured using four phrasal modification indices, are the lowest in master's theses and highest in published RAs, and concluded that academic writing becomes more complex with writer expertize (p. 12)". In addition, another study by Song and Wang (2019) compared abstracts written in English of doctoral theses by L1 Chinese and L1 English students. The results of their study reported less use of subordination than L1 English students. According to Wu et al. (2020), "ELF writers utilize sentences and clauses that are longer than those used by L1 English writers, (p. 12)". On the other hand, a recent study by the authors of Yin et al. (2021) investigated differences in engagement with syntactic complexity between emerging and experienced international publication authors in seven research article (RA) part-genres and presented practical implications of syntactic complexity in second language writing programs. At the same time, this study is distinct in that it examines the length of legal texts at the word and sentence level to determine the complexity of legal writings. Liu and Afzaal (2021) investigated the simplification hypotheses in corpus-based translation studies that should be approached from the perspective of syntactic complexity. Two comparable corpora are used in this study: the English monolingual part of COCE (Corpus of Chinese English) and the native English corpus of FLOB (French Language and Literature Corpus) (Freiburg-LOB Corpus of British English). Syntactic complexity and legal translation Although the foci of the legal translation have not only proliferated in the last decade but also increasingly widened their focus due to diverse implications, the legal system comprises its own system, terminologies, axiology, and boundaries to shape their concepts (Łucja Biel, 2017). In this sense, it is necessary to investigate the legal translation of company laws in non-native English-speaking countries. Besides, Maaß & Rink (2021) propose that legal communication requires some linguistic and specific jargon knowledge to understand the legal terms. As legal communication is technical, it has several characteristic features useful for communication between experts. Also, when familiar persons are addressed in specific legal terms, such syntactic forms constitute accessibility barriers (Rink, 2020, 117). Perhaps, the language complexity reveals the ability to use a wide and varied range of sophisticated structures and vocabulary. Therefore, language plays a significant role in legal correspondence; however, on the other hand, translation is inextricably connected to language and translation. The languages contribute significantly to law correspondence, from legislation to legal documents and translation to interpretation of legal language, rules, and correspondence. A comparative approach to legal studies is undoubtedly not a new marvel. Biel et al. (2019) consider 'legal translation and interpretation an interdisciplinary area of linguistic practice', (p.7, 2019). Moreover, in corpus linguistics, the phenomenon of legal translation has already been expanded for the last four years. The research in legal translation has achieved a significant pace in recent years. Although some prior studies have discussed the simplification of legal translation (Biel et al., 2019; Biel, 2016; Bolton, 2009), there is a dire need to examine the complexity of translated texts, legal documents, and correspondence. Furthermore, as the legal discourse is always at par with the approach of an ordinary person, translation makes it easier for the public. In this sense, it is necessary to examine the syntactic complexity of legal translation to unveil the linguistic structure to maximize competency in legal translation works. Legal translation refers to translating legal documents and companies' laws to make the text understandable and clear. Previously, Biel's studies (2014: pp. 36-48) examined the phraseological continuum in the language of the law, which accounts for non-terminological categories that are statistically significant in the genre of legislation. Furthermore, she (2018) says that a few types of research have been reported while incorporating corpus-based methodology in the field of legal translation (Biel, 2018, p. 34). With the advent of corpus-based studies, applying corpora in legal translation brings fruitful results (e.g., Biel, 2015, 2018; Pontrandolfo, 2011; Trklja, 2017). In translation workflow, numerous small-scale corpus-based studies of legal translations have been conducted in Chinese-English legal translation (e.g., Li and Wang 2013), which have not been compared to other European languages. The legal translation helps readers understand the complexity and ambiguous nature of legal discourse practised in courtrooms. English is taken as an official and second Language in Hong Kong, whereas English is taken as a Foreign Language in China. Consequently, these investigations are worthwhile to pursue because they will aid in discovering translation conventions that function both in the networks of translation and in the cognitive act of translations (cf. Toury, 1995; Xiao and Hu, 2015; Eagleson, 2014). Laviosa's (2002) study, based on the analytical framework of linguistic variety, sentence length and information capacity to measure simplification, recorded that translated versions of texts have restricted lexical variety, and the ratio between content and function words remains low. The study is significant for two reasons: many studies conducted in the Chinese-English translation background followed qualitative approaches in translation despite some recent research using corpus-based quantitative techniques. This paper analyzes the syntactic complexity of legal translation corpora. Therefore, this study identifies the existing gap in Chinese-English legal translation research in several areas. Thus, the study fills the gap as no parallel contrastive study has been conducted in the legal discourse regarding syntactic complexity.

The present study contributes to the analysis of syntactic features of translated law texts by comparing Chinese-translated corpora with a British corpus and Hong Kong company laws in terms of syntactic complexity. It examines whether translated texts and translation norms, framed by Kachru's (1985) Three Circles Model (inner, outer, expanding circles), have affected the production of legal texts in plain English in Hong Kong and native English-speaking countries. Research questions: - RQ1: How do translated and non-translated legal texts, as reflected in the Hong Kong Government's websites, differ across China (Mainland), Hong Kong, and the UK regarding syntactic complexity? - RQ2: If the differences or similarities are significant, can they be used in conventions of the plain English movement? Corpus design: - Inner Circle (UK): monolingual datasets of the Companies Act 1948, 1985, and 2006. - Outer Circle (Hong Kong): bilingual, sentence-aligned sub-corpus of Companies Ordinance (old: 1984 Cap 32; new: 2014 Cap 622), with earlier execution in 1932 and modifications in 1984, 2012, 2014. - Expanding Circle (Mainland China): bilingual, sentence-aligned legal texts of Company Law (1993, 1999, 2004, 2006). The corpus reflects Kachru’s model to represent English use across regions and legal systems. Syntactic complexity analysis: - Tool: L2 Syntactic Complexity Analyzer (SCA) by Lu (2010), validated on 3,554 essays by Chinese EFL majors; reliability checked by dual annotation. - Measures: 14 indices across five categories: length of production (MLC, MLS, MLT), sentence complexity (C/S), subordination (C/T, CT/T, DC/C, DC/T), coordination (CP/C, CP/T, T/S), and particular structures (CN/C, CN/T, VP/T). Definitions follow Lu (2010) (see Table 1 in the paper). Analyses compute ratios such as clauses per sentence, dependent clauses per T-unit, complex nominals per clause/T-unit, verb phrases per T-unit, etc. Design and comparisons: - Contrast old vs. new versions within each region (e.g., UK 1985 vs. UK 2006; HK Cap 32 vs. HK Cap 622; Mainland earlier vs. later versions). - Statistical testing with p-value threshold p<0.005 to determine significant differences in each syntactic feature across versions/datasets. - Supplementary reporting of Z-scores, U values, and rank summaries for non-parametric comparisons (as shown in Tables 2 and 3).

- Overall significance: Across the 14 syntactic complexity measures, most differences between older and newer versions within regions are statistically significant at p<0.005, indicating meaningful shifts in complexity with newer legal drafting and translations toward plainer English. - Hong Kong (HK) company law (Cap 32 vs. Cap 622): Table 2 indicates significant simplification in many features (e.g., MLT, MLC, VP/T) in the newer Cap 622. Some features are not significant (e.g., MLS p=0.13104; C/S p=0.41222; C/T p=0.05238). The narrative notes that most features in the new dataset (Cap 622)—including MLS, MLT, MLC, C/S, VP/T, C/T, DC/C, DC/T, CT/T, CN/C, CN/T—are simpler than in the old version, reflecting plainer language in translations. - UK Companies Act (1985 vs. 2006): The newer 2006 Act is significantly simpler on 12 of 14 features; exceptions include MLC (p≈0.961) and CN/C (p≈0.900), which show no significant difference. Features with simplification include MLS, MLT, C/S, VP/T, C/T, DC/C, DC/T, T/S, CT/T, CP/T, CP/C, and CN/T. Word counts: UK 2006 contains 270,597 words; HK corpus contains 210,902 words. - Mainland China company law (older vs. newer versions): In older versions, only MLC, CP/T, and CP/C are simpler; other features (MLS, MLT, C/S, VP/T, C/T, DC/C, T/S) are more complex. In newer versions, MLS, MLT, C/S, VP/T, C/T, DC/C, and T/S become simpler, while MLC, CP/T, and CP/C remain more complex relative to the other features. - Cross-region patterns: UK (Inner Circle) and HK (Outer Circle) corpora generally show greater simplification than Mainland (Expanding Circle) corpora, aligning with the influence and adoption of plain English drafting norms. - Cognitive load: Complex nominal and hypotactic structures correlate with higher proposition density per sentence, increasing cognitive processing demands for readers; simplification via plain English in translations reduces these burdens and enhances accessibility. - Practical implication: Evidence supports that translation and plain English initiatives improve understandability of legal texts for lay audiences, with HK showing notable adoption and effects of plain English compared to Mainland.

The study addressed how translated and non-translated legal texts differ across the UK, Hong Kong, and Mainland China and to what extent plain English aids comprehension. Findings confirm that legal language typically departs from plain English, with complex nominal and hypotactic structures contributing to dense information packaging and higher cognitive demands. However, significant simplification is evident in newer versions, particularly in the UK and Hong Kong corpora, consistent with the plain English movement’s influence on legal drafting and translation norms. The Inner Circle (UK) sets language standards and shows measurable simplification in the 2006 Companies Act compared to 1985, while Hong Kong, influenced by the Inner Circle, has also advanced plain English in its translated legal texts (Cap 622 vs. Cap 32). Mainland corpora display mixed trends, with fewer features simplified in older versions but noticeable improvements in newer versions. These regional contrasts suggest that the adoption and institutionalization of plain English practices in legal drafting and translation can meaningfully reduce syntactic complexity and enhance accessibility for lay readers, including those with communication impairments. Overall, the results indicate that plain-language initiatives and translation practices can mitigate accessibility barriers inherent in traditional legal drafting. Nonetheless, the intrinsic complexity of law and the need for precision can limit the extent to which legal texts can be simplified without compromising legal adequacy.

Using a corpus-based contrastive design and syntactic complexity measures, the study demonstrates significant differences in the syntactic complexity of company law texts across the UK, Hong Kong, and Mainland China. Newer UK and Hong Kong versions exhibit broad simplification aligned with plain English principles, while Mainland texts show partial simplification, particularly in later versions. Despite such efforts, legal texts remain challenging due to the inherent complexity of legal concepts and the necessity for precise definitions. Contributions include: (i) a large-scale quantitative analysis of syntactic complexity in legal translations across regions framed by Kachru’s Three Circles; (ii) empirical evidence that plain English initiatives in legal translation and drafting correspond to reductions in syntactic complexity and likely improvements in accessibility; and (iii) a detailed feature-level view (14 indices) of where and how simplification occurs. Future research should further validate syntactic complexity measures for legal genres, explore quantity-based readability metrics alongside complexity indices, and investigate how these findings can inform legislative drafting policies to balance clarity with legal precision.