How I use a reference manager in my PhD

A reference manager is portrayed as the missing infrastructure that lets researchers turn note-taking into verifiable, source-grounded scholarship—especially when writing scales up from short papers to a PhD dissertation. The core claim is that “tlec casting” (a slip-box style method for building connections between ideas and tracking their origins) only works as reliably as the documentation behind it. That documentation, in turn, depends on accurate citations, stable metadata, and the ability to trace every claim back to a specific source and even a specific version (translation, edited article, or page number). In an era of misinformation and AI-generated content, the ability to audit where ideas come from becomes a practical research skill, not just an academic formality.

The transcript links tlec casting’s purpose—developing original thought by connecting information from existing knowledge—to the bibliographic slip box that records sources separately from notes. The system’s value isn’t just storage; it’s lineage. By keeping sources distinct and traceable, gaps in knowledge become visible, fallible sources can be questioned, and biases can be surfaced by seeing what shaped an idea in the first place. From there, the discussion shifts to why a dedicated reference manager matters beyond keeping links or citations inside a notes app like Obsidian. For small assignments, manually entering citations and building a bibliography can be manageable. For a PhD dissertation—over 200 pages and potentially hundreds of citations—manual citation work becomes a cognitive and time burden that risks errors and steals time from writing.

Paperpile is introduced as the reference manager used to solve that problem. It centralizes research materials: it stores references and PDFs, supports searching within the library, and allows direct PDF annotation inside the app. Those annotations can be exported (including page numbers) so quotes and notes can flow into a separate note system like Obsidian without losing the citation trail. Paperpile also emphasizes metadata accuracy and version specificity. When adding sources, it can import from tools like Zotero or EndNote, search the web for references (and attach PDFs when available), and use an ISBN lookup to ensure the correct edition of a book. A browser extension enables saving metadata from web pages like Wikipedia directly into the library.

The workflow is demonstrated through organization features—folders for projects (e.g., comprehensive exams) and labels for topics (e.g., “climate Justice”)—and through reading behavior that avoids the “collector’s fallacy” by actually opening PDFs, highlighting, commenting, and exporting annotated results. Finally, the transcript shows how Paperpile integrates with Microsoft Word to generate and update citations and a work cited list in MLA 9. Instead of manually typing citations, the user inserts citations from the library, adds page numbers, and can suppress repeated author names when citing multiple works by the same author. If metadata changes in Paperpile, Word citations and the bibliography update automatically, reducing the risk of citation drift during long writing cycles.

Overall, the message is that tlec casting builds connections, but reference management preserves accountability—making research more transparent, ethically grounded, and resilient against misinformation as writing demands grow.