Arabic–English Transliteration of Personal Names and Public Signages: A Systematic Review and Meta-analysis

Abstract

This study aimed to conduct a systematic review (SR) and meta-analysis (MA) of the author’s empirical studies published between 2021 and 2025 on Arabic–English transliteration of personal names on Facebook and public signages (shop names and linguistic landscapes). It aimed to synthesize evidence on orthographic anomalies, error patterns, and variation in English to Arabic and Arabic to English transliteration across social media, shop names, and linguistic landscapes. The fourteen studies share a unified methodological framework and provide quantitative data (percentages, frequencies, and error rates) that allow for statistical aggregation. The fourteen studies were categorized into 3 clusters: Shop names and linguistic landscapes, personal names, borrowed nouns, and ai generated transliteration. Results of the SR/MA revealed consistent patterns of inaccuracy and variation across human Arabic–English transliteration. In public signage, recurrent issues include vowel omission, inconsistent representation of consonants with no direct English equivalents, semantic and syntactic anomalies in compound names, and wide divergence from standard spellings. Personal names show similarly unstable patterns, with multiple transliterations for the same name, inconsistent rendering of the glottal stop and pharyngeal fricatives, variable spelling of the definite article /al /, and frequent gemination errors. Borrowed English nouns display phonological adaptation patterns shaped by Arabic orthography, especially in the representation of /g/ and other non native phonemes by Artificial Intelligence. Meta analytic pooling across studies shows high overall error rates, cross context variation, and tendencies toward under representation of vowels and over regularization of consonants. Subgroup analyses indicate that transliteration accuracy varies by domain, with signage showing the highest error density and personal names the greatest internal variability. Together, the findings demonstrate that human transliteration is shaped by sociolinguistic preference, orthographic habit, and contextual constraints rather than by standardized rules, establishing a coherent empirical profile of real world Arabic–English transliteration behavior. These results offer the first coherent map of human transliteration behavior and lay the groundwork for future research.