Accent and authority

I have presented the same finding in two languages. The finding did not change and the reception did.

3 min read

Some months the review went through and some months I got picked at, and for a long time I assumed the difference was in the numbers.

The material was mostly the same shape every month. Here is what closed, here is what moved, here is the thing I want you to look at, in front of senior stakeholders at a company that was buying other companies at a fair pace.

It was partly in the numbers. It was also, more than I wanted to believe, in how the first ninety seconds sounded.

The signal being read is not the one people think

Nobody in a room like that is evaluating fluency. They would tell you, honestly, that they are evaluating the analysis. What actually happens is faster than that and mostly out of anyone’s control.

A listener has to decide, early and cheaply, how much scrutiny to apply. Full scrutiny of every claim is impossible, so people run something quicker. Does this person sound like they have done this before. Do the sentences arrive finished. Is the vocabulary the local vocabulary, used the local way.

Those are all real signals. They correlate with something. The problem is that they correlate with exposure rather than with correctness, and exposure is not distributed by ability.

I know this from the inside in a way I did not choose. I work in three languages, and I can put the same argument into more than one of them and watch the reception change while the argument stays exactly where it was. I am the control variable. There is no version of that experiment where I can pretend the content did the work, because the content was mine both times and I know it was identical.

The one thing I will say from the record rather than from feeling: when I sat the French diploma, my oral result came back two levels above the written one. Same language, same afternoon. Somebody who only ever heard me speak French and somebody who only ever read it would have formed two different views of my competence, and both would have been confident, and neither would have been about competence.

What writing changed, partly

The obvious repair is to move the argument into a channel that does not carry the signal, and that has genuinely happened, without anyone deciding it should.

A great deal of work that used to be decided in a room is now decided in a document that circulates before the room, or in a thread, or in a written proposal that gets read by people who never hear the author. That shift was made for scheduling reasons and time zones and nothing else. One of its side effects is a partial redistribution of who gets believed.

I have watched people who are unremarkable in a meeting become quite powerful in writing, in organisations that had no idea this was happening, because nobody sets out to reallocate authority and nobody notices when the tooling does it.

Where I had this too neat

Written channels have their own accent, and I underrated how strong it is.

Register, sentence length, how you open, whether you hedge, whether you use the industry’s current phrasing or last year’s. All of that is read, and it is read without the softening that a face provides. A written channel does not remove the scoring, it changes which fluency is being scored, and the new one is harder to acquire because nobody corrects you in the moment. You just get less agreement and no explanation.

There is a worse correction, which is about me. I spent a stretch attributing resistance to how I sounded when the actual problem was that the argument was weak. That attribution is comfortable and it is unfalsifiable from where you are standing, and it will keep you from fixing the thing that was fixable.

The tell I use now is whether the resistance survives the same claim arriving in writing from somebody else. Usually it does not, which tells me one thing. When it does, that is the more useful month.

All essays