Chapter 4: Preservation Strategies

The Archaeobyte Lives—Now What?

You have done the work. You excavated the artifact from the digital strata, diagnosed its ontological state with your three-axis Microscope, and cleared it through the Custodial Filter. The Archaeobyte sits before you—rescued from oblivion, ethically approved, ready for preservation. And yet, a new crisis emerges: what form should this preservation take?

In traditional archaeology, preservation is often straightforward. A clay pot can be cleaned, stabilized with consolidants, and displayed in a climate-controlled case. The pot remains, materially, a pot. But digital artifacts exist at the volatile intersection of hardware, software, and data—what preservation theorist Jeff Rothenberg calls "a dance between content and machine."¹ When the machine stops dancing, the content becomes unintelligible.

Consider a 1985 Lotus 1-2-3 spreadsheet stored on a 5.25" floppy disk. To preserve this artifact, we face cascading dependencies:

• The physical disk (magnetic medium vulnerable to bit rot)
• The disk drive (hardware requiring vintage IBM PC architecture)
• The operating system (MS-DOS 3.x or compatible)
• The application (Lotus 1-2-3 v2.01 with its proprietary .WKS file format)
• The contextual ecosystem (printer drivers, macro libraries, user documentation)

Each dependency layer introduces vulnerability. As David Bearman notes, "Digital preservation is not a product but a process—one that must be continuously enacted against the tide of technological obsolescence."² This is the paradox of digital preservation: formats are mortal, but meaning must persist.

Two competing philosophies have emerged to address this challenge. Emulation argues we should recreate the original machine—build a virtual IBM PC, install MS-DOS, run Lotus 1-2-3 exactly as it was in 1985, preserving not just data but the authentic experience of interaction. Migration argues we should translate the spreadsheet into a contemporary format like .XLSX or .CSV, accepting some loss of authenticity in exchange for guaranteed readability on modern systems.

Neither approach is universally correct. Each carries philosophical assumptions about what constitutes "preservation"—what matters most, what can be sacrificed, who the future audience is. This chapter provides the conceptual foundation and practical frameworks to navigate these decisions. We begin with format obsolescence theory, explore the emulation-migration spectrum, examine hybrid strategies, and conclude with a preservation decision matrix calibrated to your artifact's diagnostic profile.

"To preserve is to choose what survives. To choose is to declare what matters."

I. The Problem of Format Obsolescence

Formats as Sociotechnical Contracts

A file format is more than a technical specification—it is a sociotechnical contract between creator and future reader, mediated by software. When you save a document as .DOCX, you are making an implicit bet that Microsoft Word (or compatible software) will exist in the future to render that document. As Matthew Kirschenbaum observes, "File formats are stabilized agreements about how to encode meaning in machine-readable form."³

But agreements expire. Vendors go bankrupt. Standards bodies dissolve. Software architectures evolve. What happens when the contract can no longer be honored? Format obsolescence occurs when:

The UK National Archives estimates that 40% of digital government records from the 1990s are already at risk of format obsolescence, with proprietary database formats and email systems particularly vulnerable.⁴ This is not a theoretical problem—it is an ongoing crisis.

The Spectrum of Format Risk

Not all formats are equally vulnerable. The Library of Congress Sustainability of Digital Formats project identifies key risk factors:⁵

This risk assessment informs preservation strategy. Low-risk formats may only require refreshing (copying to new media) and monitoring (watching for software drift). High-risk formats demand immediate intervention—either emulation to freeze the execution environment or migration to a more sustainable format.

The Ship of Theseus Problem

The ancient paradox haunts digital preservation: if you replace every plank of a ship over time, is it still the same ship? If you migrate a WordPerfect document through six format transformations (WPD → DOC → DOCX → ODT → HTML → PDF), is it still the same document?

Philosopher of technology Don Ihde distinguishes between instrumental preservation (preserving function—"can I still read it?") and hermeneutic preservation (preserving meaning—"does it still mean what it meant?").⁶ A spreadsheet migrated to CSV preserves the data (instrumental success) but loses formulas, formatting, and macros (hermeneutic loss). An emulated Lotus 1-2-3 session preserves the entire phenomenological experience—loading times, menu navigation, even the CGA color palette—but requires future users to learn obsolete interfaces.

There is no escape from this paradox. Every preservation decision is an act of interpretation—a declaration of what aspects of the artifact constitute its "essential character." The question is not how to avoid choosing, but how to choose consciously and transparently.

II. Strategy One: Emulation (The Museum Approach)

Philosophy and Implementation

Emulation preserves the artifact by recreating its original execution environment in software. Rather than translate the content, we build a virtual machine that mimics the hardware and operating system the artifact was designed for. Jeff Rothenberg, emulation's chief advocate, argues: "The only reliable way to preserve digital documents is to preserve the ability to recreate the original environment in which they were created."⁷

The emulation chain looks like this:

Modern OS (Windows 11/macOS/Linux)
    └─ Emulator (DOSBox, QEMU, MAME, Basilisk II)
        └─ Guest OS (MS-DOS, System 7, Windows 3.1)
            └─ Historical Software (Lotus 1-2-3, HyperCard, Director)
                └─ Original Artifact (the .WKS file, .HC stack, .DIR project)
                

Each layer introduces complexity, but the payoff is authentic preservation. When you open a 1991 HyperCard stack in a Macintosh System 7 emulator, you experience the artifact as its creators intended—complete with pixel-perfect rendering, original sound synthesis, and interaction timing.

Case Study: The Internet Archive's Emulation Infrastructure

The Internet Archive's Historical Software Collection demonstrates emulation at scale. Using Emularity (a JavaScript wrapper for emscripten-compiled emulators), they deliver in-browser experiences of MS-DOS games, early Macintosh software, and arcade cabinets.⁸ Users can boot The Oregon Trail (1985) or Prince of Persia (1989) directly in their web browser, experiencing pixel-perfect recreations of the original software.

The technical stack:

• Emscripten: Compiles C/C++ emulators (DOSBox, MAME) to WebAssembly
• EM-DOSBox: Browser-based DOS environment
• JSMESS: JavaScript port of MAME (Multiple Arcade Machine Emulator)
• Metadata Layer: JSON manifests describing boot sequences, required files, input mappings

As of 2024, the archive hosts over 100,000 emulated software titles spanning four decades of computing history. This is emulation as public infrastructure—a commitment to preserving not just files but computational experiences.

Advantages of Emulation

Disadvantages and Limitations

Emulation is the museum approach—it preserves artifacts in their original context, but at the cost of accessibility and long-term sustainability. It is ideal for high-value artifacts where authenticity outweighs usability (interactive fiction, digital art, software as cultural artifact).

III. Strategy Two: Migration (The Translation Approach)

Philosophy and Implementation

Migration preserves the artifact by translating it into a contemporary format. Rather than recreate the past, we bring the artifact into the present, accepting transformation as the price of survival. As digital preservation theorist Adrian Brown argues, "Migration acknowledges that change is inevitable—the question is whether we manage that change deliberately or let it happen through neglect."¹⁰

Migration strategies include:

The goal is not pixel-perfect reproduction, but semantic preservation—maintaining the artifact's informational content and functional purpose even as its material form changes.

Case Study: The UK Web Archive's Migration Pipeline

The UK Web Archive, operated by the British Library, captures over 1 billion UK web pages annually. Given the scale and format diversity (HTML, CSS, JavaScript, Flash, Java applets, video codecs), emulation is impractical. Instead, they employ a normalization-on-ingest migration strategy:¹¹

This hybrid approach balances accessibility (users see working pages) with scholarly integrity (researchers can audit transformations). The system processes ~40TB of web data annually, demonstrating migration at institutional scale.

Advantages of Migration

Disadvantages and Losses

Migration is the translation approach—it prioritizes usability and integration over authenticity, accepting that the artifact will change over time. It is ideal for large-scale collections where access outweighs perfect preservation (government documents, email archives, web crawls).

IV. Hybrid Strategies and the Middle Path

Beyond Binary Choices

The emulation-versus-migration debate presents a false dichotomy. In practice, most preservation projects employ hybrid strategies that combine both approaches, tailored to specific use cases and user communities.

Consider three increasingly sophisticated hybrid models:

Format-Specific Strategies

Some formats demand specialized approaches that blend emulation and migration:

The key insight: preservation strategy should follow from preservation intent. What aspects of the artifact are considered "significant properties"? Who is the intended audience? What resources are available? These questions must be answered before choosing a technical approach.

V. The Preservation Decision Matrix

Mapping Triage State to Strategy

Preservation strategy should align with your Chapter 2 triage diagnosis. The three-axis framework (Technical Legibility, Functional Integrity, Contextual Ecosystem) provides decision inputs:

The Preservation Strategy Scorecard

When triage state is ambiguous or resources are constrained, use this weighted decision scorecard:

The Preservation Action Report

Just as Chapter 3 required a Custodial Report, Chapter 4 requires a Preservation Action Report documenting your strategy decision. This creates an audit trail for future practitioners and ensures transparency.

Example: GeoCities Vienna (from Chapter 1)

Let's apply the framework to the GeoCities Vienna case study:

Conclusion: The Operational Anvil

You have now assembled the complete toolkit:

• The Trowel (Chapter 1): You can excavate digital artifacts from obscured strata
• The Microscope (Chapter 2): You can diagnose their ontological state with precision
• The Filter (Chapter 3): You can determine ethical clearance for preservation
• The Anvil (Chapter 4): You can forge preservation strategies that balance authenticity and accessibility

The Anvil is where theory becomes practice, where abstract debates about "what is preservation?" meet the concrete reality of storage budgets, staff time, and user needs. Preservation is not a technical problem with a universal solution—it is a hermeneutic problem requiring contextual judgment.

Emulation preserves the past at the cost of the present. Migration brings artifacts into the present at the cost of the past. Hybrid strategies attempt to serve both gods, but require twice the devotion. There is no escape from trade-offs. The question is not which strategy is "correct," but which trade-offs you—as custodian, not owner—are willing to accept on behalf of future communities.

As digital preservation pioneer Margaret Hedstrom wrote: "We preserve not things but the capacity to make meaning from things."¹⁴ Whether that meaning emerges from pixel-perfect emulation or semantically migrated data is secondary to the fact that meaning can still emerge at all. The worst preservation strategy is the one never implemented—the artifact left to rot because we demanded perfection instead of accepting pragmatism.

You are now equipped to act. Excavate, diagnose, clear, and preserve. The Archaeobytes wait. The Anvil is hot. Strike while the iron—and the bits—still hold form.

"Digital preservation is an act of optimism—a belief that the future will care about what we cared about, and that we can build bridges across the chasm of obsolescence to make that caring possible."
— Trevor Owens, The Theory and Craft of Digital Preservation, 2018