Something I disagree with: Writing advice often implores one to write in a “strong” way, that one should sound authoritative, that one should not sound uncertain.
While I agree that this can create a stronger reaction in the audience, it is a close sibling to dishonesty, and communication is best facillitated when one feels comfortable acknowledging the boundaries of their ability to know.
But perhaps I’m wrong- when one writes, one is not writing for an ideal Bayesian reasoner under the assumption of perfect honesty, since ideal Bayesian reasoners are not physically possible, and one cannot reliably prove that one is being honest, but one is rather writing for humans, and perhaps the most efficient way of transferring information given these constraints is to exaggerate one’s confidence.
Or perhaps the important bit is that humans care greatly about common knowledge- Charlie doesn’t care so much about the information that Charlie can derive from what David says, but rather how the people around Charlie will respond to what David says, and Charlie can more easily predict others’ reactions when David avoids indicating uncertainty, thereby making Charlie more comfortable taking a similarly strong position.
As it happens I came across this issue of strength (& its reverse, qualification) the very first time this morning, in Paul Graham’s essay How To Write Usefully. Here are his thoughts on the matter, FYI: http://www.paulgraham.com/useful.html
Something I disagree with: Writing advice often implores one to write in a “strong” way, that one should sound authoritative, that one should not sound uncertain.
While I agree that this can create a stronger reaction in the audience, it is a close sibling to dishonesty, and communication is best facillitated when one feels comfortable acknowledging the boundaries of their ability to know.
But perhaps I’m wrong- when one writes, one is not writing for an ideal Bayesian reasoner under the assumption of perfect honesty, since ideal Bayesian reasoners are not physically possible, and one cannot reliably prove that one is being honest, but one is rather writing for humans, and perhaps the most efficient way of transferring information given these constraints is to exaggerate one’s confidence.
Or perhaps the important bit is that humans care greatly about common knowledge- Charlie doesn’t care so much about the information that Charlie can derive from what David says, but rather how the people around Charlie will respond to what David says, and Charlie can more easily predict others’ reactions when David avoids indicating uncertainty, thereby making Charlie more comfortable taking a similarly strong position.
As it happens I came across this issue of strength (& its reverse, qualification) the very first time this morning, in Paul Graham’s essay How To Write Usefully. Here are his thoughts on the matter, FYI:
http://www.paulgraham.com/useful.html