Academic publishers and sharing humanities data: the interface between theory and practice. This paper will examine the role of academic publishers as an interface, specifically in their support of research data sharing by humanities scholars. The expectation for researchers to share the data underpinning their research findings originates from stakeholders including funders, publishers, institutions and research communities themselves. The motivations for this are broad and suggest that sharing associated materials will elevate the research in a number of ways - for example, increased openness, transparency, reproducibility, trust, impact and integrity. In creating research data sharing policies to support these aims and motivations, theoretical best practices can be translated to action, transforming research objects into publicly-shared, reusable datasets which evidence the claims made by their associated publications. Regarding the publication and the linking of research objects, publishers act as an interface between drafts and the Version of Record, by minting persistent identifiers and creating links between articles/chapters and other research material. They also provide an interface between best practice data sharing ideals and published data outputs: by providing both notional “gateways” which facilitate or prevent publication, and technical gateways to support data sharing practices. In adopting the role of an interface, we acknowledge the necessity to translate current journal data sharing requirements (often associated with life sciences journals) to policies which reflect the working practices of humanities scholars. These include, but are not limited to, different research methods, definitions of data, data collection practices, ownership and licensing of data, data life cycle and longevity, repository coverage and features, as well as cultural attitudes and sharing practices. Also understanding that while some of these ideas may seem to be straightforward or procedural, in reality they can be messy or fuzzy. We will also introduce the work of the Humanities Sub-Group of the STM Association’s Research Data program, currently composed of representatives from six global academic publishers, and discuss the topics that this group is currently addressing. These include collating definitions of “research data” in the humanities, identifying examples of the benefits of humanities data sharing, and developing resources to assist humanities scholars to comply with journal data policies and increase the impact of their research.