The training layer: a meta-aggregator / HathiTrust Digital Library
HathiTrust Digital Library
infrastructureThe corporaactive
The archive's reading. The archive reads HathiTrust in three layers kept apart: holdings (volumes digitized via Google Books, the Internet Archive and the libraries), access (restricted where copyright status cannot be established), and computational availability (the Research Center's, now ended). Authors Guild v. HathiTrust, as Wikipedia reports it, is the record's precedent for a machine reading of whole books whose product is a finding (search) while the pages stay closed; the saved sources tie none of its volumes to model training.
| Relation to this census | infrastructure |
|---|---|
| Record type | library |
| The question it answers | holding and access: is a volume held, can it be found by full-text search, and may its pages be shown? |
| Kind | large-scale collaborative repository of digital content from research libraries (Wikipedia's description) |
| Operator | HathiTrust, a university consortium of more than 219 research libraries (2024) under shared governance, administered by the University of Michigan; executive director Mike Furlough (per Wikipedia; the own about page was refused) |
| Evidence | digital library |
| Unit of record | a volume: a digitized book or other publication, viewed in the PageTurner application |
| What it records | per Wikipedia, content from research libraries digitized via Google Books and the Internet Archive digitization initiatives and digitized locally by libraries: more than 18 million volumes, 6.7 million of them public domain in the United States (September 2024) |
| In the layer | not recorded: the saved sources say nothing of training use; per Wikipedia, the courts held its use of books scanned by Google to be fair use, with search and accessibility for the visually impaired grounds to find the service transformative |
| Under whose name | not recorded |
| Found and read back | per Wikipedia, full-text search across the entire repository; volumes viewed and downloaded as PDF in PageTurner where access allows; access restricted where copyright or permission status cannot be determined, and for users outside the United States to works published outside the United States from 1896 |
| What the apparatus makes from it | per Wikipedia, a full-text search index across the repository, which the courts held transformative and fair use; computational analysis of the volumes was the separate service of the HathiTrust Research Center, ended 30 September 2026 |
| Method | per Wikipedia, aggregation of library digitization (Google Books, Internet Archive, local digitization) in a shared repository; a Shared Print Program retaining almost 18 million monograph volumes for 25 years |
| Coverage | geography: member libraries across the United States, Canada and Europe (Wikipedia); period: since October 2008 (Wikipedia) |
| Count | value: more than 18 million volumes, including 6.7 million in the public domain in the United States; as_of: September 2024 (Wikipedia) |
| Holdings | more than 18 million volumes (September 2024, Wikipedia); 6.7 million volumes in the public domain in the United States (Wikipedia); full-text search across the repository (Wikipedia) |
| Stated limits | the own about page was refused (http 403), so no limit is quoted from the resource itself; Wikipedia quotes its copyright policy: 'if we cannot determine the copyright or permission status of a work, we restrict access to that work until we can establish its status' |
| Data access | per Wikipedia, full-text search; PageTurner viewing and PDF download where access allows; the Emergency Temporary Access Service for users of member libraries in special situations |
| Licence | per Wikipedia's infobox, 'Public domain (with restrictions on Google scans), various' |
| Lifecycle | state: active; detail: per Wikipedia, launched October 2008, current status active; the HathiTrust Research Center closes at the end of 2026, and access to the Digital Library remains open (HTRC home page) |
| Version and time | holdings: September 2024 (Wikipedia); members: 2024 (Wikipedia); read: 2026-11-01 |
| Founded | October 2008 (Wikipedia) |
What its record can settle
- whether a volume is held, by full-text search across the repository
- whether a volume can be displayed or only searched, under its copyright policy
- the scale of its holdings (Wikipedia figures)
What it cannot settle
- whether its volumes were carried into any training set (the saved sources do not say)
- anything in the resource's own words (own about page refused)
Kept apart
- NOT HATII (Wikipedia's disambiguation)
- NOT Google Books: per Wikipedia, its holdings include Google-digitized content alongside Internet Archive and locally digitized content
- NOT an open full-text library for in-copyright works: access is restricted where status cannot be established
- NOT recorded in the saved sources as a training corpus
- NOT the HathiTrust Research Center: computational analysis was the Center's separate service, which ended 30 September 2026
Relations
- has_part: HathiTrust Research Center Analytics (the HTRC home page describes the HathiTrust Research Center as established in 2011 as a partnership between Indiana University, the University of Illinois Urbana-Champaign and HathiTrust, for computational analysis of the HathiTrust Digital Library)
- uses: Internet Archive (Wikipedia: holdings include content digitized via the Internet Archive digitization initiatives)
Evidence
- self_description_wp (a third-party account):
HathiTrust Digital Library is a large-scale collaborative repository of digital content from research libraries , administered by the University of Michigan .
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - record_object_wp (a third-party account):
Its holdings include content digitized via Google Books and the Internet Archive digitization initiatives, as well as content digitized locally by libraries.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - founded_wp (a third-party account):
HathiTrust was founded in October 2008 by the twelve universities of the Committee on Institutional Cooperation and the eleven libraries of the University of California .
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - members_wp (a third-party account):
As of 2024, members include more than 219 research libraries
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - governance_wp (a third-party account):
across the United States, Canada, and Europe, and is based on a shared governance structure.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - director_wp (a third-party account):
The executive director of HathiTrust is Mike Furlough,
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - shared_print_wp (a third-party account):
The HathiTrust Shared Print Program is a distributed collective collection whose participating libraries have committed to retaining almost 18 million monograph volumes for 25 years
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - count_wp (a third-party account):
In September 2024, HathiTrust comprised more than 18 million volumes, including 6.7 million in the public domain in the United States.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - readback_wp (a third-party account):
HathiTrust provides a number of discovery and access services, notably, full-text search across the entire repository.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - fair_use_wp (a third-party account):
A federal court ruled against the Authors Guild in October 2012, finding that HathiTrust's use of books scanned by Google was fair use under US law.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - second_circuit_wp (a third-party account):
which found that providing search and accessibility for the visually impaired were grounds to consider the service transformative and fair use
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - limit_restrict_wp (a third-party account):
if we cannot determine the copyright or permission status of a work, we restrict access to that work until we can establish its status.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - limit_international_wp (a third-party account):
access is also restricted for users outside the United States to works published outside the United States after and including 1896.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - pageturner_wp (a third-party account):
PageTurner is the web application on the HathiTrust website for viewing publications.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - etas_wp (a third-party account):
for users of HathiTrust member libraries to obtain lawful access to copyright digital materials in place of the corresponding physical books held by the same library.
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - license_wp (a third-party account):
Public domain (with restrictions on Google scans), various
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - status_wp (a third-party account):
Current status Active
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - disambiguation_wp (a third-party account):
Not to be confused with HATII .
https://en.wikipedia.org/wiki/HathiTrust · read 2026-11-01 · http 200 · sha256 cf7c92313dd725f8… - htrc_open (the resource's own statement):
Access to the HathiTrust Digital Library remains open.
https://analytics.hathitrust.org/ · read 2026-11-01 · http 200 · sha256 5e7d8be238f03fa7…
Unchecked
the own about page (hathitrust--about): http 403, saved text is a challenge page ('Just a moment...'), unread and unquoted; HathiTrust's Access and Use Policies and copyright policy (known here from Wikipedia's quotation); full-text search itself (no query run)
Cite: Lee Sharks, ed. "HathiTrust Digital Library." The training layer: a meta-aggregator, traininglayerliterature.org, compiled 2026-11-01. https://www.traininglayerliterature.org/aggregators/hathitrust