Lexical Complexity in Corporate Communication: A Corpus-Based Study of Translated and Non-Translated Chairman’s Statements

2.1 Simplification as a Translation Universal

The notion of translation universals was defined as “the features which typically occur in translated text rather than original utterances and which are not the result of interference from specific linguistic systems” (Baker 1993, p. 243). Some of the most well-known translation universals include simplification, explicitation, and normalization. Explicitation refers to translation’s tendency to “spell things out rather than leave them implicit” (Baker 1996, p. 180). It can be reflected both at the lexical or syntactic level. Normalization can also be termed “conventionalization” (Mauranen 2007) and means “the tendency to exaggerate features of the target language and to conform to its typical patterns” (Baker 1996, p. 183). Consequently, translated texts seem more normal than target texts (Xiao 2010). Among these proposed universals, simplification has received considerable attention in translation studies (Liu and Afzaal 2021). Simplification refers to the “tendency to simplify the language used in translation” (1996], p. 181).

Research on translation universals utilizes three primary methodological approaches, each of which requires specific types of corpora. The first is the comparable corpus approach, which compares translated texts with non-translated texts in the target language to identify universal features of translated language (Baker 1993). The second is the parallel corpus approach, which involves comparing source texts with their translations to examine translational patterns (Olohan 2004). The third is the multi-parallel corpus approach, which analyzes multiple translations of the same source text to explore translation variability (Granger 2003). Each of these approaches requires distinct combinations of corpora, including translated texts, original texts in the target language, and source texts. Regardless of the approach, certain essential criteria must be met for the corpora used: sufficient size for statistically significant results (Zanettin 2014), balanced representation of authors, topics, and source languages (Olohan 2004), and comparability in terms of genre, register, time period, and subject matter (Laviosa 2002).

Simplification has been explored at various linguistic levels, including the lexical level (e.g., Laviosa 1998) and the syntactic level (e.g., Wang, Liu, and Moratto 2023). For example, Blum-Kulka and Levenston (1983, p. 119) examined translations from Hebrew to English and identified lexical simplification, defined as “the process and/or result of making do with fewer words.” Vanderauwera (1985) examined syntactic simplification, discovering that translated texts often replace finite clauses with non-finite ones to streamline complex syntactic structures. In addition to syntactic changes, Vanderauwera also investigated stylistic simplification in translated texts. However, the generalizability of these findings is limited due to the small sample sizes and the absence of robust statistical methods in these earlier studies.

Both lexical and syntactic indicators have been explored in studies of translation universals, with lexical features receiving more frequent attention. Among the key contributions in this area is Laviosa’s (1998) study, which examined lexical density, sentence length, and the frequency of commonly used words in English translations of narrative prose. Using data from the English Comparable Corpus (ECC), the study identified four significant ways in which translated English differs from native English, highlighting distinct lexical patterns in translated texts. First, compared to original texts, translated texts exhibit a lower lexical density and a greater percentage of grammatical terms. Second, translated texts employ more high frequency terms. Third, the most frequently-used words are repeated more often in translated texts. Fourth, fewer lemmas are included in the most commonly-used terms in translated texts than in original texts (ibid.). Olohan (2004) used lexical diversity as a metric to compare translated and native English fiction, finding that translated works employed fewer color synonyms than their native counterparts. Similarly, Pastor et al. (2008) explored simplification by utilizing natural language processing tools, readability formulae, and other indices. Their study revealed that non-translated texts exhibited higher lexical density and richness compared to translated texts.

While simplification has been the focus of numerous studies, it remains a contentious topic among translation universals due to conflicting findings that challenge its universality (Liu and Afzaal 2021). For instance, longer mean sentence lengths have been observed in translated texts (Laviosa 1998; Xiao and Yue 2009), which contradicts Malmkjær’s (1997) assumptions. Jantunen (2004) found a higher frequency of modifiers in translated texts, offering no consistent evidence to support translation universals. Similarly, Mauranen (2000) studied collocations and found that multi-word patterns in untranslated texts were clearer and more stable, while translated texts displayed unusual collocations. Contradicting the simplification hypothesis, Ferraresi et al. (2018) found that translated texts were more complex and exhibited higher lexical density. It is also important to note that research on lexical simplification has predominantly focused on isolated or selectively chosen lexical features, lacking systematic and comprehensive analyses of broader lexical complexity indices. Moreover, most studies on simplification have concentrated on literary translation (Blum-Kulka and Levenston 1983; Laviosa 1998), leaving non-literary translation, such as financial texts, largely unexplored. To address this gap, it is crucial to investigate the lexical complexity of translated annual reports to determine whether simplification occurs in this genre. Following Baker’s (1993) comparable corpus approach, this study compares translated chairman’s statements from mainland Chinese companies listed on the Hong Kong Stock Exchange with non-translated statements from Hong Kong companies listed on the Hong Kong Stock Exchange and U.S. companies listed on the New York Stock Exchange and NASDAQ. This design provides a diverse representation of non-translated texts, enhancing the robustness of the investigation into translation universals in corporate communication. By analyzing the lexical complexity of translated and non-translated chairman’s statements, we can deepen our understanding of whether simplification universals apply to this context. This research can contribute to the broader study of translation universals, highlighting similarities and differences between genres and registers in translated texts.

2.2 Lexical Complexity in Language-Related Research

Lexical complexity, also referred to as lexical richness, is a measure of how diverse and sophisticated the vocabulary is in a text or by a language user. It involves several key dimensions, including lexical density, lexical sophistication, and lexical variation (Bui 2021; Lu 2012; Shi and Lei 2022). Lexical density is defined as “the ratio of the number of lexical words (Nlex) to the number of words (N)” (Lu 2012, p. 196). It is a measure that determines the proportion of meaningful words or content words in a given text, relative to the total number of words (Ure 1971). Lexical diversity, also referred to as lexical variation or lexical range, assesses the range of different words employed in a text (Lu 2012). Lexical diversity can be easily measured using various methods, such as counting the number of unique words in a text or employing the type-token ratio (TTR) (Jarvis 2013). However, TTR can be influenced by the length of the text (Engber 1995). To address this issue, scholars have put forward alternative calculations of TTR, such as corrected TTR (Carroll 1964) and root TTR (Guiraud 1960). Although the type-token ratio and its alternative calculations have some controversies, some previous studies used them in assessing lexical diversity of texts (Daller, Van Hout, and Treffers-Daller 2003; Lu 2012). Lexical sophistication refers to lexical rareness, which measures “the proportion of relatively unusual or advanced words” in the texts (Read 2000, p. 203). In contrast to lexical diversity, which focuses on the variety of words, lexical sophistication evaluates the complexity and rarity of the vocabulary. It is typically measured by calculating the proportion of less frequent words, as determined by reference frequency lists or corpora. Researchers compare the average word frequency in a text to that of a reference corpus (Lu and Hu 2022). These frequency-based measures indicate how much a text relies on uncommon or advanced vocabulary. Texts with a higher proportion of common words are generally considered less sophisticated (Crossley et al. 2007).

Lexical complexity has been extensively studied in various linguistic domains, including second language proficiency assessment, second language writing, academic writing, and translation studies. Researchers have explored its use in gauging second language learners’ proficiency (e.g., Zareva, Schwanenflugel, and Nikolova, 2005) and examined its relationship with writing quality in language testing (Laufer and Nation 1995; Yu 2010). Lei and Yang (2020) investigated its role in academic writing across different levels of expertise, finding that academic proficiency had a greater impact on research article writing than nativeness. In translation studies, lexical density and variation have been analyzed to test the simplification hypothesis at the lexical level (Laviosa 1998; Pastor et al. 2008; Volansky, Ordan, and Wintner 2015). These diverse applications highlight the significance of lexical complexity as a linguistic measure across multiple areas of language research and assessment.

2.3 Research Gaps and Questions

Based on the preceding review, two research gaps emerge: (1) Previous studies have largely concentrated on isolated lexical features in translated texts, lacking systematic and comprehensive analyses of lexical complexity to investigate lexical simplification. (2) Research has primarily focused on simplification in literary translation, leaving non-literary translation, especially financial translation, underexplored. This study aims to address these gaps by applying a comprehensive set of lexical complexity measures to examine the differences between translated and untranslated chairman’s statements. Specifically, the study seeks to explore the following research questions:

RQ1: In what ways do translated chairman’s statements differ from untranslated ones in terms of lexical complexity?

RQ2: What factors might account for any identified similarities or differences between the two?

3.1 Data Description

For this study, we compiled a corpus of chairman’s statements (COCS) from three regions: Mainland China, Hong Kong, and the United States. The data were sourced from the annual reports of listed companies. The chairman’s statements from Mainland Chinese and Hong Kong companies were obtained from companies listed on the Hong Kong Exchanges and Clearing Market (HKEX), while the statements from American companies were drawn from firms listed on either the New York Stock Exchange (NYSE) or the National Association of Securities Dealers Automated Quotations (NASDAQ). Unlike companies in Mainland China, which often produce annual reports first in Chinese and then translate them into English, companies in Hong Kong typically prepare their reports directly in English (Wang 2014). This longstanding practice is attributed to Hong Kong’s bilingual environment and the strong influence of Western stock exchange systems (Setter, Wong, and Chan 2010). As a result, Hong Kong offers a unique perspective on a variety of English that reflects both Western and Eastern linguistic and cultural influences (Wang 2014). Hong Kong’s position as a bridge between these two worlds provides a rich comparative framework, revealing subtleties in English usage within a bilingual and bicultural context. The inclusion of Hong Kong English texts enriches the diversity of non-translated corporate communications, aligning with Laviosa’s (2002) emphasis on comparing translated texts with a broad spectrum of non-translated texts. By integrating both U.S. and Hong Kong English texts, this study captures a more comprehensive representation of original English corporate communications, thereby strengthening the robustness of comparisons with translated texts.

The compilation of the three corpora followed a two-phase process. In the first phase, companies were selected based on their sectors to ensure balance across the corpora. The sector distribution was aligned with that of stocks listed on the HKEX, with 100 companies selected from 10 sectors^[1] for each corpus. A corresponding number of companies from each sector were chosen for the Mainland Chinese (CMCS), Hong Kong (HKCS), and U.S. (USCS) corpora. This method ensured a balanced and comparable representation of industries across regions. Table 1 provides a detailed breakdown of the descriptive data of the compiled corpora. In addition, the original Chinese versions of the chairman’s statements from Mainland Chinese companies were also collected as a reference, enabling a more comprehensive analysis of translation and its impact on the linguistic features of corporate communications.

3.2 Data Analysis

The chairman’s statements in the three corpora were analyzed using the Lexical Complexity Analyzer, a computational tool developed by Lu (2012) to assess lexical complexity in written English. The tool offers 25 different measures of lexical density, variation, and sophistication, as detailed in Table 2. Its reliability depends on the accuracy of the POS tagging and lemmatization, which are managed by the Stanford tagger, known for its 99 % accuracy (Toutanova et al. 2003). Therefore, the overall reliability of the analyzer is strengthened by the strong performance of the tagger and lemmatizer (Jin et al. 2021).

This study employed statistical techniques to compare the means or mean ranks of the 25 lexical complexity measures across the three corpora in order to investigate potential differences. The three corpora were treated as the fixed factor, while the 25 lexical complexity measures served as the dependent variables. To assess the normality of the dependent variables, the Shapiro–Wilk test was conducted, revealing that most variables did not follow a normal distribution. Consequently, the Kruskal–Wallis test was applied to identify statistically significant differences among the three corpora. For pairwise comparisons, Mann–Whitney U tests were conducted to assess significant differences between groups. The initial significance level (alpha) was set at 0.05, and to account for the increased risk of Type I error due to multiple comparisons, a Bonferroni correction was applied. Since three Mann–Whitney U tests were performed, the adjusted alpha level was set at 0.05/3 = 0.0167. The use of these statistical methods enabled the identification of significant differences in lexical complexity among the corpora. By applying the Bonferroni correction, we minimized the likelihood of Type I errors, thereby ensuring the reliability and validity of the study’s findings.

5.1 Strategic Simplification and Sophistication in Translation

The findings of this study reveal a dual trend in the translated chairman’s statements (CMCS), challenging traditional assumptions about translation universals. On one hand, the CMCS texts exhibit lower complexity in most lexical measures compared to the original English chairman’s statements (USCS and HKCS), including the Number of Different Words (NDW), Type-Token Ratio (TTR), and variations in verbs, nouns, and adverbs. This reduced complexity suggests a simplification process, where the vocabulary becomes more limited and repetitive. These results align with the simplification hypothesis, a well-established concept in translation studies, which posits that translated texts tend to use more common words and reduce variability to maintain clarity (Baker 1993). Our findings align with previous studies (Laviosa 1998; Pastor et al. 2008) that have shown translated texts tend to be less complex than non-translated texts. The directionality of translation may influence these results. In our study, the chairman’s statements from mainland Chinese companies were likely translated from Chinese (A language) into English (B language). Although translating into a B language is less conventional, it is increasingly common in international business communication (Pokorn 2005). The lower complexity observed in NDW-related metrics, TTR-related metrics, the Uber index, and various word class variations (lexical word, verb, noun, adverb) aligns with expectations for translation into a non-native language. Translators working into their B language often have a smaller active vocabulary and may choose more familiar, frequently used words to ensure clarity and avoid errors (Campbell 2014), which could explain the observed vocabulary simplification.

However, the translated texts also exhibit higher lexical density and verb sophistication, revealing a deliberate complexity in specific areas. This paradoxical finding can be attributed to two primary factors. First, the Chinese language structure inherently employs more verbs to express actions and processes, which may result in a higher proportion of verbs in the translated texts (Liu 2010). Second, translators may opt for sophisticated word choices to accurately convey culturally specific or abstract concepts that exist in Chinese but lack direct English equivalents. This reflects Jakobson’s (1959) concept of “equivalence in difference,” which proposes that creative and nuanced linguistic solutions are essential when confronting the challenges of translating between languages with distinct grammatical and lexical structures. To maintain the original meaning and nuance, translators utilize more advanced vocabulary, contributing to an increase in both lexical density and verb sophistication. This complexity is not accidental but rather a manifestation of the translator’s need to bridge the gap between the source and target languages while ensuring that the text remains comprehensible and preserves its original intent. The translator’s choices are guided by the necessity to strike a balance between faithfully conveying the source text’s content and making it accessible to the target readership. Consequently, while certain facets of the translation process lead to simplification, others require a more sophisticated approach to capture the nuances of the original text. This delicate equilibrium between simplification and sophistication illustrates that translation is not merely a cognitive activity occurring within the translator’s mind but also a form of cross-cultural social interaction (House 2014). As a result, translators must carefully navigate these dual objectives, tailoring their strategies based on the specific genre and communicative purpose of the text (Nord 1997). Simplified language aligns with modern digital reading preferences (Liu 2008) and regulatory requirements (Securities and Exchange Commission 1998). Ultimately, the interplay between simplification and sophistication in translated chairman’s statements reflects the translators’ efforts to produce a text that is both faithful to the original and effective within its new cultural and linguistic context.

5.2 Cultural Influences on Lexical Complexity in Corporate Communication

Chairman’s statements serve to communicate key information about a company’s performance, opportunities, challenges, and future plans in non-quantitative terms (Hu and Tan 2020). However, cultural differences influence how this information is presented, especially in translated chairman’s statements, which must align with their source texts in Chinese. According to Wang and Liu’s study (2024]) using Biber’s multidimensional analysis, translated chairman’s statements from Mainland Chinese companies are more informational, narrative, context-independent, and less spontaneous compared to those from American and Hong Kong companies.

This study’s findings emphasize the significant role of cultural factors in shaping the lexical complexity of translated chairman’s statements. The higher lexical density observed in CMCS, compared to USCS and HKCS, reflects a cultural preference in Mainland China for information-rich communication. Mainland Chinese corporate discourse tends to favor more indirect, contextualized, and writer-oriented texts, with a preference for providing extensive background information and details (Zhu 2005). This approach contrasts with Western corporate communication norms, which prioritize directness, concision, and reader-orientation, emphasizing clarity and efficiency.

Hong Kong’s business communication follows Western norms, reflecting its history as an international financial hub and former British colony. This is evident in the common practice of preparing annual reports in English rather than Chinese (Courtis and Hassan 2002). By contrast, many Chinese Mainland companies listed on the Hong Kong Exchange maintain a more authoritative and formal communication style, likely influenced by the political structure, where CEOs of state-owned enterprises often resemble government officials more than business professionals (Yang, Wang, and Nie 2013). The differences in lexical complexity between translated and non-translated chairman’s statements, therefore, reflect not only linguistic but also political and cultural distinctions. Moreover, the similarities between CMCS and HKCS highlight Hong Kong’s bilingual and bicultural influence on corporate communication. As a bridge between Eastern and Western business cultures, Hong Kong’s corporate discourse seemingly blends Western linguistic norms with Eastern rhetorical traditions. This cross-cultural context results in corporate texts that balance simplification with complexity, aligning with both local and international expectations.

These findings underscore the importance of considering regional linguistic norms and cultural preferences when analyzing translated corporate texts. Cultural factors shape both the language used and the broader communicative objectives, influencing how translators balance clarity and simplicity with professionalism and credibility. Ultimately, the level of lexical complexity of translated chairman’s statements reflects a nuanced interaction between strategic linguistic choices and cultural influences, requiring translators to adapt their approach to the specific cultural and communicative context of each text.