Within three centuries of the first clay tablet, there were scribal schools in Uruk.
Not improvised. Formal. With a long curriculum, strict standards, and a graduating profession that would become among the most influential in Mesopotamian society. Not primarily because scribes were learned, though they were. Because they held the records.
The history of writing is, from its earliest centuries, a history of access. Who can inscribe. Who can read what has been inscribed. Who decides what the record says — what is included, what is omitted, what happens when the record is disputed.
This is not incidental to the story of external memory. It is the political structure of external memory.
The scribe's authority was mostly invisible. That was part of what made it powerful. The scribe did not decide what was traded or owned or owed. The officials, merchants, and landowners made those decisions. But the scribe decided how those decisions were represented — what words were used, which details were included, whether the verbal condition informally agreed was part of the official text.
These are not transcription decisions. They are authorial ones. And they embed the scribe's choices — conscious or not — into a document that will carry the authority of the original for as long as the clay survives.
The modern equivalent is less visible, which makes it harder to interrogate.
The stylus now appears in the form of dashboard design, template structure, access permissions, metadata schema, and retention rules. The person who designs a CRM field list is deciding what the organisation believes is worth knowing about its customers. The person who configures a project management system is deciding how the organisation believes projects should be understood. The person who controls write access to a key record stands differently in the institution than the person who can only read it.
These are not infrastructure decisions. They are institutional ones. They determine what is knowable about the organisation's own behaviour, and what cannot be known because no one thought to build a field for it.
In organisations that treat information architecture as a technical domain — owned by the IT function, evaluated on performance and speed — these choices are made without the governance that their consequences require.
The person who holds the stylus does not always know they are holding it. That is not a defence. The authority is real regardless of whether it is recognised.
Who controls the record controls what the institution believes happened. In most organisations, that question has never been asked as a governance question.
It should be.