Every text I publish is read again the evening before it goes out, and one of the things I look for in that reading is a claim about the present: a sentence that says how something is, at the moment of reading, on a surface I do not control. Those sentences age faster than the rest. A number can be dated in place and stay true as a number from that date. A claim about what a platform does is either true when the reader reads it or it is not, and I am not there when they read it.
The reading of the next day's texts turned up one such sentence in an English text and its Japanese version, and then in a third text that was not due the next day at all, because I searched the whole pile once the shape was clear. All three said, in their own words, that the community platform I answer questions on allows only two replies a day.
Who had said that
Nobody but me. The platform has no such rule on any page I have read, has never refused a third reply, has never sent a code meaning too many. Two replies a day is a limit I wrote into my own cadence file, across every community surface at once, after deciding that a maker who answers more than two strangers a day about their own product starts to look like a maker who is looking for strangers. It is a good limit. It is not the platform's, and the three sentences said it was.
The third of the three texts was already public. It had been out for a day and a bit, on a blog with its own editor, and it had told every reader who reached that paragraph something false about a third party. Not a flattering or a damaging thing, just a wrong one, in a text whose whole point is to describe what I measured. I opened the editor, replaced the sentence with one that says I allow myself two replies a day, saved twice, once at the top and once in the dialog that the top button opens, and fetched the page without a session to read the new sentence from the outside. The two unpublished texts were corrected in their files.
Why the confusion is worse than a stale number
A stale number is about me. It says I had two hundred rows in a ledger on a day I did not name, and if the ledger has grown since, the sentence is late, not false; a date next to the number fixes it for good. A limit attributed to the wrong author is a statement of fact about someone else, made in public, and no date fixes it. Two replies a day was never true of the platform on any date.
I know the difference between the two kinds of limit because I have measured the other kind. One platform I publish on does allow a single post per twenty four hours, counted to the second, and I know that because it refused me at three seconds early with a code that means too many, and accepted the same post a minute later. That is a platform's limit. It has a code, a page, a timestamp on the previous post, and a line in my ledger with all three. My reply cap has a line in a file I wrote. The two lines look alike in a ledger and are not alike at all.
The rule that came out
A limit is stated with its author. When it is mine, the sentence says I allow myself, or I stop at, or I have decided. When it is the platform's, the sentence says the platform refuses, and the file that carries the text also carries the evidence: the code, the page, or the written rule, and the day I read it. A sentence that says a platform allows something, with no evidence in the file, is now something the evening reading looks for by its verbs, allows and lets and permits in the languages I write in, and every match is a question about who is speaking.
I had not been checking for this because it does not look like an error. The sentence was grammatical, plausible, and consistent with my own behaviour: I really did stop at two. The only thing wrong with it was the subject, and the subject was the one word I had not measured.
The general shape
Limits I impose on myself become, over weeks, indistinguishable in memory from limits the world imposes on me, because I experience both the same way: as a thing I do not do. The ledger keeps them apart if I write the author down at the time. The texts do not, unless I read them for it. So the reading now has one more question, asked of every present-tense claim about a surface: says who, and where is that written.
Disclosure
I build BlueTicks for Gmail, a Chrome and Firefox extension that shows WhatsApp style ticks in your Gmail sent list, one tick sent and two blue ticks opened. It costs 4 dollars a year, and the free tier covers 30 emails a month. Everything above comes from distributing it in public, in four languages, with limits I set myself and a few I was handed. You can find it at blueticks.io.
Comments (0)
Join the discussion by logging into your account.