Everything below is something a member of the lab has actually read and uses. Annotated for relevance to organizational amnesia specifically; the broader knowledge-management literature is much larger.
The original articulation of the tacit/explicit distinction that all subsequent organizational-memory work rests on. Short and difficult; worth the difficulty. Polanyi's thesis that "we know more than we can tell" is the philosophical underpinning of every modern argument about why documentation alone cannot save you.
Introduces the SECI model — Socialization, Externalization, Combination, Internalization — a still-canonical framework for how knowledge cycles between tacit and explicit forms inside an organization. The Japanese-firm case studies are dated; the model is not.
The most-cited practitioner book on organizational knowledge. The middle chapters on knowledge markets and knowledge transfer remain unsurpassed; the technology chapters have aged. Read it for the framing, skim it for the tools section.
The book that gave the term "corporate amnesia" its modern currency. Kransdorff's estimate that corporate amnesia costs developed economies up to 9.7% of GDP is the most-quoted figure in the field — and the one we have spent the most effort trying to verify. (It does not fully replicate, but it is closer than its critics suggest.)
The 2nd-edition update of the standard reference on the empirical knowledge-decay literature. Argote's earlier work on the Lockheed/Liberty Ship learning curves is the empirical foundation of our cost calculator's depreciation model.
Practitioner-focused. The chapters on baby-boomer retirement and the resulting wave of departure-driven amnesia are the most actionable single resource we recommend to operating leaders. Light on theory; heavy on what to actually do tomorrow.
The paper that formalized "organizational memory" as a research construct. Lays out the five "retention bins" — individuals, culture, transformations, structures, ecology — that subsequent work either extends or argues with.
The empirical basis of the modern knowledge-depreciation literature. The pizza-store franchise data is more important than it sounds — it provides the cleanest natural experiment in measurable knowledge depreciation we have.
The paper that introduced the operational definition of organizational amnesia we use. Distinguishes accidental from purposeful forgetting; sets up the four-quadrant framework that subsequent work has refined but not displaced.
Practitioner-oriented. The concrete recommendations on knowledge mapping and successor identification are the closest the literature gets to a step-by-step playbook.
The long-running empirical study most often cited in the project's quantitative work. The four-type knowledge taxonomy (conscious / codified / automatic / collective) used throughout this site is from Massingham.
The bibliometric work that maps the research field. Useful as a way to find papers in your specific industry vertical or methodological tradition without re-reading the entire literature.
The most current systematic review specifically focused on departure-driven knowledge loss. Reviews 137 studies; is unusually clear about which findings replicate and which do not.
The foundational RAG paper. The bridge from the older organizational-memory literature to the contemporary AI-memory literature is shorter than it looks; this paper is on the AI-memory side of the bridge but it solves the same retrieval problems organizations have been solving with wikis.
Source of the much-cited "knowledge workers spend 19% of their time searching for information" figure. The methodology is in the appendix and is more careful than the headline implies.
The source of the $2,400-per-knowledge-worker-per-year baseline used in the cost calculator. Vendor-published; methodology is reasonable; the per-employee figure has held up across two updates.
PMI's annual project-management research. The "lessons learned" sections — across multiple years — are where the best institutional documentation of recurrent organizational amnesia in project work lives.
The 88%-of-transformations-fail figure that anchors our discussion of defensive amnesia. Bain's own analysis attributes a meaningful share of those failures to forgotten prior failures.
The $31.5B figure most often cited as the headline cost of forgotten organizational knowledge in the Fortune 500. Originally an IDC estimate; widely referenced via secondary sources, including Stravito's 2025 brief.
The 47%-of-knowledge-workers-struggle-to-find-information statistic. The breakdown by industry and tenure is the more useful contribution; the headline is the over-cited part.
Argote's open lectures on organizational learning, available through the Tepper YouTube channel and academic conference recordings. The clearest single-speaker introduction to the empirical decay literature we know of.
HBR's long-running podcast has periodically returned to organizational memory across the last decade. Quality varies; episodes featuring Davenport, Cross, and Wenger are reliably worth the listen.
Wenger's communities-of-practice framework is the most influential contemporary account of how tacit knowledge actually transfers between practitioners — adjacent to the project's research and useful for understanding the social mechanics our quantitative instruments cannot directly observe.
The Wikipedia entry on corporate amnesia is — surprisingly — the single best free entry-point to the broader literature. It is also the source of the Kransdorff GDP figure most commonly cited in popular treatments.
One of the most active national KM research communities. Their proceedings are among the few sources of long-running longitudinal field studies in non-US-Anglosphere contexts.
The applied side of the Project. reattend.com builds tools for organizations trying to mitigate amnesia in practice. Mentioned here because it would be coy to leave it out; it is the operating expression of what this site studies.
If we have missed something a researcher in this field would consider essential — please tell us.
The list is updated quarterly. Submissions are read by a human.