Corpus-based translation and interpreting studies have gained prominence, aiming to characterize mediated language. However, criticism has emerged regarding over-reliance on "translation universals" and the manual selection of linguistic features, neglecting contextual influence and the holistic nature of language use. This study addresses these limitations by focusing on diplomatic discourse during Chinese Regular Press Conferences (CRPCs), which provides a high-stakes context for examining interpreted language. CRPCs offer a unique opportunity given the high-stakes nature of the events and the significant impact of the communication on international relations. Existing research on interpreted diplomatic language has often focused on general aspects or specific features, neglecting a comprehensive examination of co-occurring linguistic elements and their interplay with contextual factors. Therefore, the study proposes a multi-dimensional analysis (MDA) to uncover co-occurring patterns of linguistic features in interpreted diplomatic language compared to non-interpreted counterparts, addressing the following research questions: (1) What are the similarities and differences between interpreted diplomatic language and its non-interpreted counterpart in terms of the co-occurring patterns of linguistic features? (2) What contextual factors contribute to the co-occurring linguistic patterns in interpreted diplomatic language?

Early research on diplomatic language focused on stylistic features such as precision and formality. The linguistic turn led to investigations into interpreted diplomatic language, with studies exploring interpreting norms and agency, and examining specific lexical and syntactic features like lexical density, modal verbs, and hedging. However, limitations remain: insufficient attention to communicative context, a limited set of linguistic features, and the use of difference tests that fail to fully capture the complexities of language use. This study argues that a multi-dimensional analysis (MDA), which considers the interplay of various linguistic features within specific communicative contexts, is crucial. While MDA has been increasingly applied in translation and interpreting studies, its application to interpreted diplomatic language remains limited. Existing studies highlight the potential of MDA but often use pre-defined feature sets or insufficient statistical evidence, limiting their scope. This study aims to address these limitations by conducting a full MDA tailored to the specific discourse domain of interpreted diplomatic language during Chinese Regular Press Conferences.

This study compiled a comparable corpus using transcripts from Chinese-English simultaneous interpreting during CRPCs (SI corpus) and non-interpreted speeches from U.S. Department Press Briefings (NS corpus). The SI corpus contained 1,140,302 tokens, while the NS corpus had 1,062,135 tokens. Data cleaning focused on preserving spokespersons' replies. A full multi-dimensional analysis (MDA) was conducted following Biber's framework. This involved: (1) calculating and normalizing frequencies of linguistic features; (2) performing factor analysis to identify co-occurring patterns; (3) computing dimension scores; and (4) applying Mann-Whitney U tests to detect significant differences between the SI and NS corpora. The selection of linguistic features was inclusive, incorporating 113 variables initially, including grammatical features, n-grams, and mediated language-related features. After data cleaning and grouping similar features, 88 features were subjected to factor analysis using SPSS 26. Principal factor analysis (PFA) and Varimax rotation were used. A five-factor solution was determined optimal based on scree plot analysis and explained 42.80% of the variance. The Mann-Whitney U test, chosen for its suitability with non-normally distributed data, was used to compare the two corpora at multiple levels.

Factor analysis revealed five dimensions: (1) Involved vs. Informational Production (D1): SI showed a significantly higher informational focus (p<0.001), characterized by higher lexical density and longer words, compared to the more involved, less precise NS. (2) Objective vs. Addressee-focused Narration (D2): SI demonstrated significantly greater objectivity (p<0.001), using perfect aspect, adverbials, and downtoners, in contrast to NS's more addressee-focused approach using second-person pronouns, public verbs, and predictive modals. (3) Literate-Oral Continuum (D3): No significant difference was found (p=0.500) between SI and NS, although SI showed a slight tendency toward literateness. (4) Information Elaboration (D4): NS showed significantly higher elaboration (p<0.001), using more adjectives and amplifiers, than SI's more succinct style. (5) Narrative vs. Non-narrative Concerns (D5): SI displayed a significantly lower use of narrative elements (p<0.001), characterized by split auxiliaries and pp + prep + V-ing constructions, whereas NS utilized more past tense verbs and third-person pronouns. Overall style comparison using a weighted score revealed that SI showed a significantly more formal style compared to NS (p<0.05).

The findings reveal that interpreted diplomatic language in CRPCs exhibits distinct linguistic characteristics compared to its non-interpreted counterpart. The more informative and objective nature of SI can be attributed to communicative aims, interpreting norms emphasizing accuracy and completeness, and the structured nature of CRPCs. The absence of a significant difference on the literate-oral continuum suggests that both forms share characteristics, possibly due to the inherent formality of diplomatic discourse and the influence of written structures. The lower elaboration and narrative elements in SI may be explained by shared background knowledge among participants, interpreting constraints, and the focus on conveying essential information in diplomatic settings. The overall more formal tone of SI reflects the seriousness of diplomatic discourse, sensitivity of language, and awareness of audience.

This study provides a comprehensive analysis of linguistic variation in interpreted diplomatic language using a full MDA. The findings highlight the informative, objective, succinct, non-narrative, and formal nature of SI compared to NS. This research advances corpus-based interpreting studies and offers valuable pedagogical insights for training diplomatic interpreters by identifying linguistic patterns that can serve as benchmarks for enhancing skills and accuracy. Future research should consider using more robust methods for factor analysis, addressing the limitations of the study.

Methodological limitations include using a fixed threshold for factor loadings, the variance explained by the factorial solution not meeting the 50% threshold, and potential overlap between dimensions due to the multi-dimensional nature of language and the multifunctionality of linguistic features. Further studies should explore oblique rotation methods and conduct sensitivity analyses. The study's reliance on transcripts might not fully capture paralinguistic aspects of spoken language which might influence the linguistic features.