When the first mass-market ebook readers appeared in the mid-2000s, many assumed the printed book was entering its twilight. But the shift wasn’t a replacement, it was a redistribution of functions. Physical books retained qualities that digital formats could not replicate: the stability of a fixed page grid, the tactile index of wear on a spine, the unmediated chain from printer to reader. Digital delivery, by contrast, excelled where print struggled most: moving text across continents in seconds, indexing millions of volumes by phrase or concept, and scaling storage without shelf space. Yet those gains came with quiet costs, not in reading experience alone, but in how we locate, verify, and inherit written work.

These changes weren’t uniform across formats. An EPUB file behaves differently from a PDF. A library loan behaves differently from a purchase. And a DRM-locked title behaves differently from one sold without restriction. Understanding what shifted, and what merely migrated, requires looking past the screen and into the structure of the files themselves, the contracts governing them, and the physical habits they displaced.

Search and distribution: where digital delivered

Before digital, searching inside a book meant flipping pages, scanning indexes, or relying on memory. Even academic libraries required card catalogues, microfiche, or interlibrary loans for obscure references. Today, full-text search across tens of thousands of titles is instantaneous. A reader can locate every instance of ‘copper sulphate’ in a 19th-century chemistry manual without opening the cover, or even knowing the title. This isn’t just speed; it’s a change in epistemic access. The same applies to distribution. A small press in Buenos Aires can release an ebook simultaneously in Lagos, Helsinki, and Vancouver, bypassing customs, shipping delays, and regional licensing barriers. Print-on-demand services further narrow the gap, but they still require physical handling, inventory, and carrier networks. Digital files move as data: light-speed, frictionless, and independent of geography. That shift altered publishing economics, lowered entry barriers for independent authors, and enabled niche subjects to find audiences previously unreachable through bookstore placement alone.

Navigation: the lost architecture of the page

Printed books offer spatial anchors. Readers remember where an idea appeared, not just by chapter, but by position on the right-hand page, near a footnote, or opposite an illustration. Margins, running heads, page numbers, and even paper stock contribute to cognitive mapping. Digital formats disrupt this. EPUB files reflow text to fit screen size, font choice, and line spacing. A paragraph that falls at the bottom of page 87 in a paperback may appear mid-screen on a tablet, and shift again if the reader enlarges the type. Page numbers vanish unless artificially inserted as landmarks, and those landmarks rarely align across devices or editions. PDFs preserve fixed layout, but at the cost of legibility on small screens: zooming and panning replace turning pages, breaking continuity. Neither format restores the shared reference point of a physical page. Footnotes detach, illustrations float, and endnotes become harder to cross-reference. Navigation becomes functional, search, table of contents, progress bar, rather than embodied. You find content, but you don’t inhabit its structure.

Provenance and permanence: the quiet instability of digital text

A printed book carries its history in its materials: typeface, paper grain, binding method, marginalia, library stamps. These features allow scholars to date editions, trace provenance, and distinguish between revisions. Digital files carry far less intrinsic evidence. An EPUB contains metadata, author, title, publisher, but that data is editable, often incomplete, and rarely verified. Two identical-looking ebooks may differ in source text, editorial intervention, or encoding standard, with no visible cue. Worse, the file itself is fragile. Formats become obsolete: early PalmDOC or Rocketbook files are now unreadable without emulation. Even current standards evolve, EPUB 2 versus EPUB 3 support different scripting and accessibility features, and not all readers handle both equally. Publishers occasionally withdraw titles from sale, and purchased ebooks vanish from accounts without notice or recourse. There is no equivalent to the secondhand market’s role in preserving out-of-print works. When a digital edition disappears, it leaves no trace on a shelf, no catalogue entry in a municipal archive, just silence in a cloud locker.

Ownership, resale, and bequest: the DRM constraint

The word ‘purchase’ misleads when applied to most commercial ebooks. What buyers acquire is a licence to access content under terms set by the vendor. That distinction became stark with the introduction of DRM, digital rights management. DRM encrypts the file and ties it to a user account or device ecosystem. It prevents copying, sharing, or conversion to other formats. More consequentially, it blocks resale and bequest. A printed book can be gifted, donated to a library, sold at a flea market, or left to a grandchild in a will. Its legal status as a physical object grants transferable ownership under doctrines like first-sale. DRM overrides that. When a reader dies, their ebook library does not pass to heirs unless the platform explicitly permits account inheritance, a rare exception. Similarly, no secondary market exists for licensed ebooks. Attempts to create one, like the short-lived startup ReDigi, were blocked in court because courts treated the act of transferring a file as making a copy, not moving it. This isn’t a technical limitation but a legal and architectural one, baked into how DRM systems authenticate and authorise access.

EPUB vs PDF: two approaches, two compromises

Underlying many of these issues is the divergence between EPUB and PDF, formats designed for different purposes, now pressed into service as general-purpose ebook containers. EPUB is built on web standards: HTML, CSS, and XML. Its core strength is reflowability. Text adapts to screen dimensions and user preferences, supporting accessibility tools like screen readers and dynamic font scaling. But that flexibility sacrifices layout fidelity. Illustrations may resize unpredictably, tables break across columns, and poetry loses line integrity. PDF, by contrast, preserves exact positioning, type size, margins, image placement, as if capturing a printed page. That makes it ideal for academic journals, architectural manuals, or art books. Yet PDFs resist adaptation. On mobile devices, users must zoom and pan rather than scroll, and text selection can be inconsistent across scanned versus born-digital files. Neither format handles annotations portably: highlights and notes often remain locked inside a single app, unable to migrate with the file. And neither solves the deeper issue of long-term readability: EPUB depends on evolving reading systems; PDF depends on stable rendering engines that may degrade over decades.

What remains unchanged

The digital book did not abolish the printed one. It revealed which functions were tied to paper’s materiality, and which were simply logistical bottlenecks awaiting better infrastructure. Distribution and search improved because they were problems of scale and coordination, not essence. Navigation, provenance, and permanence deteriorated because they relied on fixed, physical referents that software deliberately discarded in favour of adaptability. DRM didn’t emerge from technical necessity but from commercial caution, attempting to replicate print’s scarcity in a medium defined by perfect reproducibility. As readers, our task isn’t to choose sides, but to recognise which capacities we’ve gained, which we’ve traded away, and which deserve deliberate restoration, not through nostalgia, but through design choices that treat text as both information and artefact.