A casual LinkedIn comment can sometimes detonate into a far more serious discussion than expected. That is what happened when Rose posted about her overnight Microsoft 365 migration and her husband waking up in panic because half her inbox had vanished. It was a funny, familiar moment, but it also exposed a problem I have been dealing with for years. A problem almost invisible to people outside the Japanese language world. A problem that affects sustainability, accessibility, sovereignty and digital freedom all at once.
That problem is furigana support. Or, more accurately, the almost total lack of proper, interoperable, standards‑compliant furigana support in software that claims to be sustainable, privacy‑respecting, or open‑source replacements for Microsoft Office.
Furigana are tiny hiragana or sometimes katakana printed above kanji in horizontal text or to the right in vertical text. They tell you how to pronounce the kanji. They are essential in children’s books, textbooks, language learning materials, municipal notices, newspapers, manga and fiction. They appear in technical documents, glossaries, and especially on names, where readings can be unpredictable even for native speakers. They are not decorative. They are not optional. They are foundational accessibility infrastructure for Japanese readers.
And almost every editor outside Word fails to support them correctly.
The Work That Absolutely Requires Furigana
I work in Japanese localization, teaching and linguistic engineering. I produce reading exercises, vocabulary lists, graded content, bilingual annotations, corrected manuscripts and technical translations. All of these require furigana, not occasionally but constantly. Not as superscripts. Not as weirdly positioned text boxes. Not as detached kana floating above lines like misplaced stickers. I need proper ruby text: bound to the underlying characters, stable under reflow, preserved across exports and copy‑paste, and linked to dictionary data so readings can be generated and corrected efficiently.
Only Microsoft Word provides this consistently.
This is not because Microsoft alone cares about Japanese typography. It is because proper furigana support demands a multilayered technical system. Most editors never implement that system at all, or implement only fragments of it. Developers often underestimate how complex it is until they collide with the real requirements.
What Proper Furigana Support Actually Involves
First, an editor must allow you to apply furigana to selected text. This sounds basic, but most editors treat text as a flat stream of characters with no structure. Without a clear separation between base text and annotation layers, furigana degrade into styled superscripts that break the moment the document reflows.
Second, the editor must support ruby text at the typographic level. That means ruby base, ruby text, grouping, fallback rules, spacing logic, alignment control and line‑height negotiation. LibreOffice can technically render ruby, but line spacing becomes inconsistent, paragraphs balloon in unpredictable ways, and multiple lines with ruby become visually irregular. Google Docs does not support ruby at all. Markdown has no standard ruby syntax. Apple Pages includes a hidden, inconsistent phonetic guide feature that breaks under copy‑paste.
Third, proper furigana means supporting both horizontal and vertical writing. Japanese uses both. Horizontal layout places ruby above the base text; vertical layout places ruby to the right of the base text. Many layout engines only support Western horizontal writing and cannot reposition ruby correctly.
The Need for Dictionary Integration
Fourth, dictionary integration is non‑optional. There is no algorithm that can guess the correct reading of a kanji or kanji compound with acceptable accuracy. Kanji are not phonetic. One character can have multiple readings. Multi‑kanji compounds can have several legitimate pronunciations. Japanese names can be arbitrary and historically idiosyncratic. Determining the correct reading requires a morphological analyzer, a dictionary of readings and a fallback model for unknown compounds.
Microsoft Word uses the system IME dictionary to guess readings, often correctly. LibreOffice cannot do this. OnlyOffice cannot do this. Google Docs cannot do this. Without dictionary‑based generation, adding furigana becomes a manual process, and manual processes become intolerable beyond a few dozen words.
Fifth, even the best dictionary integration needs human correction. Editors need a streamlined correction interface. Word provides a single dialog where you adjust readings, alignment and size. LibreOffice requires multiple modal windows and context switching that slows the workflow to a crawl. Most other editors do not offer correction at all.

Export and Interoperability Problems
Sixth, preserving furigana across formats is where almost every editor besides Word collapses. Export to PDF often flattens ruby into fixed glyphs. Copy‑paste from LibreOffice into a browser strips all ruby structure. Export to HTML from Pages produces styled spans that other editors cannot interpret. Excel stores furigana as phonetic fields, but Calc loses them or converts them into separate text boxes. Interoperability becomes a lottery where the losing ticket is the default.
Seventh, furigana must remain stable under reflow. If a paragraph wraps, the ruby must wrap with it. If you edit the base text, the link between ruby and base characters must remain intact. I have seen implementations where the ruby drifted, duplicated itself, detached from the wrong characters or persisted after the base text was deleted. This makes documents unreadable and untrustworthy.

Accessibility Considerations
Finally, furigana must integrate with accessibility systems. Screen readers must recognize that ruby is pronunciation, not content. Users should be able to toggle it on or off. Without this, furigana become noise for blind or low‑vision readers and a burden rather than a support.
It should now be clear why Word dominates this space. It is the only editor that implements the full stack of requirements. That is not good news. It means Word is mandatory for anyone working with Japanese educational materials, junior‑level reading content, name‑heavy documents, or accessibility‑conscious communication.
This is not cultural lock‑in. It is technical lock‑in.
Sustainability Hits the Same Wall
Rose and I were discussing her move toward sovereignty‑first infrastructure: self‑hosting, minimizing unnecessary compute, reducing dependence on hyperscale providers and choosing tools aligned with privacy and European regulatory values. She made a clean break from Microsoft 365 by migrating to a small, efficient VPS and rebuilding her productivity setup around Nextcloud and OnlyOffice. It was the kind of transformation that has become overdue for many people and organizations.
But my situation shows the other side of the story. If I tried to duplicate her move today, my workflow would implode. My students would lose accessible documents. My localization clients would receive text missing essential reading information. My learning materials would become harder to use. My spreadsheets would become unreadable. I would not be able to do my job without sacrificing essential functionality.
This is a sustainability barrier. Most discussions about sustainable software focus on energy, carbon, bandwidth and infrastructure choices. These matter deeply. But they are not the whole picture. Sustainability also means making it possible for people to switch to sustainable tools without being penalized for the language they speak or the script they use. If you cannot use a tool because its internationalization support is incomplete, then its sustainability benefits become irrelevant.
When sustainable tools fail to support essential linguistic or accessibility features, sustainability becomes selective. It becomes something that works well for English and a few European languages, but excludes the rest. That is not a sustainable future. A sustainable web must be a global web.

Why Furigana Belong in Sustainability Discussions
I told Rose that furigana are not an exotic edge case. They sit at the intersection of accessibility, internationalization and sustainability. If sustainable office suites lack furigana, then Japanese users remain trapped in high‑impact ecosystems. That has environmental, social and technological implications.
The Web Sustainability Guidelines explicitly ask contributors to identify barriers that prevent people from choosing sustainable options. Missing furigana support is the definition of such a barrier. It is not explicitly listed yet, but it fits the framework perfectly. It is a textbook example of how missing internationalization capabilities can make sustainable transitions impossible for entire populations.
Rose understood immediately. Even if the first publication cannot include detailed Japanese‑specific requirements without further i18n review, she said the guidelines can establish the principle: sustainability must include comprehensive language support. Missing features can create barriers as real as excessive energy use.
Longer‑term, the guidelines may reference specific examples like furigana to illustrate the principle. But that requires technical documentation—something I can contribute. I can provide the details: dictionary requirements, text segmentation rules, edge cases in name readings, vertical writing constraints, ruby grouping behavior, fallback strategies, export semantics and accessibility integration.
This is not about giving Japanese special treatment. It is about making sustainable software truly inclusive.
What Happens Next
If sustainable alternatives continue ignoring furigana support, then millions of Japanese users will remain tied to Microsoft Office regardless of their preferences or values. Teachers will remain locked‑in. Students will remain constrained. Localization teams will remain dependent on proprietary ecosystems. Businesses will pay licensing fees simply to produce readable documents. This is the true cost of incomplete internationalization.
A sustainable web cannot afford to ignore linguistic completeness. It cannot treat English as the default and every other language as an afterthought. Language support is not decorative. It is infrastructure.
For now, I continue using Word because it is the only editor that respects the practical needs of Japanese readers. I want that to change. I want open‑source office suites to implement real furigana. I want developers to understand that ruby text is not a niche feature but a central accessibility tool for a major world language. And I want sustainability conversations to include the people who cannot choose sustainable tools because those tools do not support their writing system.
This is not a small problem. It affects millions. It affects education, accessibility and digital equity. And it deserves attention from anyone serious about building a sustainable, global web.
Small characters. Small detail. Enormous consequences.


Leave a Reply