There's a version of this story that starts with a moral judgment, and then everything is explained through that lens. Either the groups were heroic pioneers who brought manga to the world, or they were copyright infringers who damaged the industry they claimed to love. Both framings are real, both are incomplete, and neither explains why the work they did looks, in retrospect, like a significant piece of cultural infrastructure.
So let's start somewhere else: the early 2000s, when most manga being published in Japan had no English edition, no license, no distribution outside Asia, and no realistic prospect of any of those things changing. The Western anime and manga community existed primarily through the same mechanisms that had built it through the 1990s — tape trading, convention bootlegs, and the kind of obsessive knowledge-sharing that happens when you're part of a small group with access to something most people can't see.
The IRC era
Before centralized sites, groups operated through IRC channels. You'd join a server, find the channel for a specific group, and request chapters through bots using commands like !list to see what was available and !get chapter-name.cbz to download. The files would transfer directly, peer to peer, at whatever speed your connection allowed.
This was slow by any modern standard. A single chapter might take forty minutes on a dial-up connection. Groups operated on volunteer schedules, and release days were events — people would sit in the channel waiting, chatting about the series, speculating about what the chapter would contain.
The structure of a group at this point was relatively flat. There was usually a project leader who coordinated which series the group was working on. There was at least one translator, often someone who had studied Japanese at university or learned through immersion. A cleaner handled the raw scan — erasing the original text, fixing scan artifacts. An editor or typesetter placed the translated text. A proofreader checked the final version. Sometimes these roles overlapped; sometimes a single person handled multiple jobs.
Quality varied enormously. Some groups were meticulous. Others moved quickly and prioritized speed over accuracy. The reader had no reliable way to know in advance which they were getting — you'd find out when you read it.
What defined a "good" group
By around 2005, certain groups had developed reputations. Not through any formal review system, but through reader experience over time. A group that consistently made careful translation choices, maintained visual quality, and didn't rush out chapters with obvious errors built trust. That trust was durable — people would follow a good group to new series, would wait for their release rather than downloading a faster but lower-quality version from somewhere else.
This reputation system had consequences. It created real quality competition. Groups that wanted to be known for accuracy had to actually be accurate. Groups that wanted to be known for speed had to actually be fast. The reader base was large enough and observant enough that pretending otherwise didn't work for long.
It also created what you might call a group culture. Each major group had a distinct character — particular standards, particular approaches to translation problems, particular ways of handling SFX or honorifics or cultural notes. Reading a chapter from one group versus another was a different experience, even for the same source material.
The shift to web distribution
The transition from IRC to web readers changed everything about how groups operated, and the effects took years to fully play out.
Web readers meant readers no longer had to download files. They could read in a browser, on the same page where chapters were listed and described. This removed a significant friction barrier, and readership numbers — always hard to estimate on IRC — became visible and large. A popular chapter on an aggregator site might be read hundreds of thousands of times in the week after release.
This scale created new dynamics. Publishers noticed. Some began sending takedown notices. Others, observing that series with active scanlations seemed to develop English readerships faster, made different calculations. The relationship between the industry and the groups was never simple, and it became more complex as the numbers got larger.
For the groups themselves, the aggregator era created a new problem: their work was being redistributed, often without credit, on sites they had no relationship with and no control over. A chapter that a group had spent two weeks carefully translating and typesetting might appear on a dozen aggregator sites within hours of release, sometimes with the group's name stripped from the credits.
The platform consolidation
The arrival of centralized platforms — sites that hosted chapters but gave groups direct accounts, credits, and some control over their work — didn't resolve these tensions. But it changed the landscape again. Groups now had a canonical place where their work lived, with version histories and comment sections and statistics.
This professionalized the field in certain ways. Groups that maintained active accounts were visible and accountable. Readers could follow specific groups, could flag errors, could see when a project was dropped or picked up. The social infrastructure that had existed informally on IRC forums and group websites moved to platforms with proper features for it.
It also created new pressures. The comment section is a challenging place for volunteer translators who are making difficult judgment calls about language. The reader who comments "why is it translated this way, the Japanese clearly says X" may be right, may be wrong, may have learned Japanese from a phone app three months ago, and the group has to decide how much energy to spend engaging with that feedback.
Where it stands
Today the scanlation landscape is harder to map than it was at any earlier point. Official digital releases have changed the economics — a title with a same-day English release on a subscription platform doesn't need a fan translation in the same way a completely unlicensed title does. Some groups have effectively retired; others continue; new groups form around specific genres or translators.
What remains constant is the basic structure: people who can read Japanese, who care enough about a specific story to spend significant unpaid time making it available to people who can't, working with a set of skills — translation, cleaning, typesetting, editing — that take years to develop properly.
The groups that did this well, over sustained periods, for titles that mattered to their readers, built something. That something is not easy to describe in terms that fit neatly into either the "heroic pioneers" or "copyright infringers" framing. It was cultural infrastructure. It was a labor of care. It was, for many readers, the reason they know what manga is.