Metadata are crucial for discovery and access of scholarly documents. However, while they provide contextual, technical, and administrative information in a standard form, metadata fields and values are also sites of tension between sociocultural representations. In metadata spaces, any input may be simultaneously interpreted as a data quality issue, political act to assert identity, or strategic curatorial choice to maximize visibility. In this context, we engaged with an international indexing service on a project to understand how metadata quality, consistency, and completeness impact individuals and communities around the globe. Working from a sample of records known to have erroneous, incomplete, or otherwise imperfect metadata, we first sought to identify and classify the issues stemming from how metadata and communities press up against each other to intentionally reflect (or not) cultural meanings. Making use of this typology, we then expanded this work to programmatically assess a random sample of 100,000 records for the prevalence of identified issues in the collaborating service’s corpus. This poster will provide an overview of the metadata quality issues we were able to identify and the typology we developed to better understand them, as well as our analysis of the prevalence of these issues, with a focus on differences found across multilingual and monolingual records.