Context That Doesn't Translate

The audit team's email used one word three times. Inappropriate. On my desk, that word refused to be translated. What they saw was an honoraria schedule for a lecture, a roster of attending physicians, and a hotel invoice from the host city. What I saw was twenty years of trust that physician had built in his department, the text of the pharmaceutical regulation, and a code of speech that passes here without remark. We were reading the same documents in different languages.

One Email

Friday night, Tokyo already dark. Basel, where HQ sits, was in its early afternoon, and the auditor's voice on the call was bright with a finished lunch — but the written text was cold. "We have concerns about three of your affiliate's physician lecture programs from last year, regarding alignment with the global code of conduct." On the first line of the attached list I found a name I knew well. A physician at a regional university hospital who has taken in intractable patients to the very end. Inviting him to lecture had been my decision.

The concern ran like this. The honorarium exceeded the regional norm. The same physician had been engaged four times in one year. The venue was not a populous urban center but a regional city near where he worked. Under HQ's standard, these constitute "suspected inducement of prescribing through provision of benefit." They are correct. Measured against the standard, correct.

But I know things. That honorarium was a fair price for his expertise; four times a year simply tracked the conference calendar of his disease area; the regional venue was chosen because he carries too many patients to travel to Tokyo. Strip the context and it looks like deviation; restore it and it looks like integrity. The trouble was that the email had no field for context.

When Legitimate Practice Looks Like Deviation

What was happening here was not the concealment of wrongdoing. It was a failure of translation. Relationships with physicians, formed locally over long stretches of time, carry an unwritten code. Excessive entertainment is barred by both the pharmaceutical law and the industry's self-regulation. But honoring an expert, paying for their knowledge, and sustaining regional medicine is a different thing from bribery. The line dividing the two is drawn only inside the local context.

HQ's code tries to draw that line as a single line across the world. I understand why. If the line differs by country, a lax line somewhere becomes a loophole everywhere. Uniformity is the strength of governance itself. Yet a uniform line severs the context that has meaning only on the ground. On the severed side, legitimate practice is left stranded.

A rule cannot carry context. Only the person reading the rule can. And HQ did not write the rules in order to trust the people on the ground. It wrote them so the machine would run without trusting them.

Two Ways of Reading

HQ and the affiliate read the same three lectures like this.

Neither side is lying. Each reading is correct from where it stands. The hard part is that the two barely become a conversation. HQ speaks the language of standards; the affiliate speaks the language of context. The standard hears context as "excuse"; the context hears the standard as "rigidity that doesn't know the ground."

Worn Down in the Attempt to Translate

That night I rewrote the reply three times. The first version was a long defense explaining every local circumstance. I read it back and deleted it. In the language of standards, a defense sounds like weakness. The second version was a short note accepting HQ's standard wholesale and promising that all three "would not be done again." I deleted that too. It would have handed over twenty years with that physician for the sake of self-protection.

Translation admits three postures.

In the third version I chose ③. For each lecture I set out which clauses of the pharmaceutical law and the industry self-regulation made it lawful, placed the honorarium calculation beside the standard fees of the relevant society, and backed the choice of venue with patient-volume data. Not as a defense, but as a translation. I rebuilt it into a form where the reader could receive context as "evidence" rather than "excuse."

The Uniform Rule That Misfires

There is another, reverse failure of translation. A uniform rule handed down by HQ in ignorance of local conditions misfires here. Take the rule that every physician contact must be pre-registered in a dedicated system. It works in the West. But in regional Japan, hospital digitization lags, and there is a culture of physicians giving their time on personal judgment. Enforce pre-registration strictly, and the most conscientious, most cooperative physicians are the first to drift away, worn down by the procedure.

To HQ, the rule is a fence against wrongdoing. To the affiliate, the fence is a net that shuts out honest relationships too. Fence and net can share a designer and still serve different functions on different ground. My job was to keep redrawing the line between these two failures, surrendering completely to neither. I have never once drawn it cleanly.

Even after sending the reply, I could not be sure what I had protected. HQ took my translation, "accepted" two of the three, and asked that the remaining one go through "prior consultation in future." Not a win, not a loss. Between context and standard, a line was drawn for that one night, and erased again.

Context that doesn't translate does not vanish. It accumulates inside the translator. As sovereign I carry the local sense of right; as subject I restate it in HQ's language. What wears down most in that shuttling is neither the rule nor the physician, but the one person who can read both. As long as someone shoulders the wear, the misalignment never becomes a rupture. But on the morning that someone is finally spent, the organization learns, for the first time, what had been holding it up.