The data article is a digital genre that has emerged in response to new exigencies, namely, to make data more transparent and research processes more trustable and reproducible. Following White's framework of intersubjective stance, this article draws upon statistical tools and collocational and discourse analyses to examine the linguistic resources deployed by authors to respond to both exigencies. The results show a high presence of dialogically contractive resources (above all, passive constructions and, only in one data article section, inanimate subjects) by which authors do not fully engage with dialogic alternatives (heteroglossic disengagement). Dialogically expansive resources (anticipatory "it"-subjects and "we"-pronouns) are extremely rare, corroborating that the authors' stance is neither monoglossic (undialogized) nor heteroglossically engaged. Further, the discourse functions and ensuing pragmatic effects of the prevailing intersubjective stance resources, significantly different between and among the data article sections, including their associated abstracts, reveal the construal of very distinct dialogic spaces for writer-reader interaction within this article type. Such intra-generic variation may be explained by the social (and rhetorical) action that the genre fulfills, namely, to describe and highlight the value of the research data.