The story of the rise of networked data through the evolution of archiving and digital storage.
Archiving Machines advances our understanding of memory, information, and data by charting the struggle between the computing technologies that archive data and the cultures of information that have led to platforms that assert control over its use. Amelia Acker examines the origins of data archives and the computing processes of storage, exchange, and transmission. Each chapter introduces data archiving processes that relate to the evolution of data sovereignty we experience from magnetic tape and timesharing computer models from the 1950s, to the establishment of data banks and the rise of database processing and managed data silos in the 1970s, to file structures and virtual containers in cloud-based information services over the past 40 years.
Archiving Machines is an unsettling book about contemporary archival work.
Grounded in history and theory, its arguments land very directly on the day-to-day realities of born-digital practice: storage choices, vendor platforms, automated pipelines, access controls, and the steady erosion of archival authority through technical default.
Acker’s central contribution is to show that the real archival stakes of digital systems lie not in whether data is stored, but in how access, context, and stability are governed. Repeatedly, the book demonstrates that archives can retain custody while losing control over how records can be interpreted, accessed, or defended as evidence.
This is not presented as a future risk, but as the baseline condition of modern systems in the present day.
What makes the book valuable from a practitioner’s perspective is its insistence that system design decisions are archival decisions.
1. Storage architectures function as access policies. 2. Automated pipelines replace appraisal with throughput. 3. App-based systems dissolve records into fragments that arrive without intelligible boundaries or provenance. 4. Infrastructure-generated data creates overwhelming scale without corresponding evidentiary clarity.
These arguments are made carefully, using concrete historical examples and contemporary systems rather than abstract critique.
Acker is clear that "bit preservation" alone is insufficient, and that archives risk preserving data they cannot render, explain, or trust over time. Equally important is the repeated warning that avoiding difficult governance decisions - around shared infrastructure, access tiers, or privacy - does not eliminate risk but simply shifts power to private (cloud) platforms.
This is not a book that offers easy solutions, and it assumes a reader already familiar with digital archival challenges. That said, its value lies in how precisely it names many failure modes that archives are already experiencing but may not yet have articulated clearly.
I would rate Archiving Machines at least 4.5 stars. It is a direct call for archives to reclaim their role as governors of access, context, and evidentiary stability - before those functions disappear into software.