When Birger Hjørland discusses the use of hierarchies of research methods to develop classification algorithms, he seems to be laying the groundwork for an exercise in targeted data linking. Some of the problems with classification systems that our readings consider can be boiled down to the concept that the criteria for classes can be vague, obscure, or archaic , especially in the longstanding enumerative approaches. The LCC is slow to adapt to new elements because each subject area is an “epistemological labyrinth” that defies hospitality. In a certain way this makes sense, because the paradigm of the users most likely to use the resources that fall into each subject heading should hold some control over how related information is organized. But as Hjørland says later on in the article, the ontologies that are accepted to structure these classifications are chosen at the expense of others. The systems these ontologies create are not objective and neither is the algorithm used by search engines like Google. Grounding relationships between research on a specialized subject in a kind of triple toward a machine-readable legibility could counteract the fixity of classification. It could also make it simpler to adapt classifications to shifting paradigms.