2) State all the givens. Things which I believed would be understood automatically and omitted to save time are much more likely to be picked apart as flaws, where the other person assumes I have not thought the matter through.
Yes—I have seen this so many times!
It’s particularly frustrating, because encountering it feels like discovering that you’ve overestimated your audience at the same time that they’ve underestimated you.
4) Don’t let your rationality slip due a sense of comradery. I feel that this community doesn’t treat commenters as friends; rather, it feels more like being treated as a coworker who is on the clock. As Morendil phrased it, “I wish someone had told me, quite plainly [...] this is a rationality dojo.”
I’ve noticed this too, and I long for the day when our rationality skills have advanced to the point where we can be rational and nice.
I haven’t really seen 3), and EY’s posts undermine 1) significantly, it seems to me.
EY’s posts undermine 1) significantly, it seems to me.
Of his most recently posted articles: Undiscriminating Skepticism scores at a Flesch-Kincaid grade level 17 and a Gunning Fog index of 17.9, You’re Entitled to Arguments is at 16 and 17.8, and Outside View as Conversation Halter is at 14 and 14.5. Note that a score of 15+ is considered academic writing by these measures. Tests of his recently upvoted comments show scores ranging from 7 to 20.
Here’s the Flesch-Kincaid calculator I used, and the Gunning Fog calculator. I would be surprised if other measures of readability, and tests of his other posts, did not show it to be academic-level writing.
Of his most recently posted articles: Undiscriminating Skepticism scores at a Flesch-Kincaid grade level 17 and a Gunning Fog index of 17.9,
“Undiscriminating Skepticism”—why, that’s ten (10) syllables right there in the title! My head is already spinning!
Seriously: tests like those do not control for the content or subject matter of the writing. There exists, furthermore, a significant subset of the (adult!) human population who would consider a phrase like “undiscriminating skepticism” itself to be difficult and unusually abstract. Needless to say, tests which heavily weight the judgements of such people are not very useful for the purpose of judging “readability” in most contexts here.
If you want to judge the readability of LW posts, I suggest spending some time reading typical articles published in academic journals.
“Undiscriminating Skepticism”—why, that’s ten (10) syllables right there in the title! My head is already spinning!
You’re right! It’s agonizing! Oh the pain of posting and reading here! My mouth is bloodied. You have defeated me, oh wise and amazing person who obviously knows better and is fully within their right to ridicule every attempt I make to explain the use of a single word in a sentence whose structure is still largely intact, as it was meant to be a frickin’ suggestion.
I would be surprised if other measures of readability, and tests of his other posts, did not show it [Eliezer’s postings] to be academic-level writing.
And yet I find his writing a model of clarity here, despite a few randomly chosen articles by other people having a far lower Fog Index. How useful are these indices? On the Gunning Fog page it says “The higher the Fog Index the trickier it is to read.” But the Wikipages for these tests reference no empirical studies.
I find his writing to be very readable as well. However, I consider myself highly educated, with excellent English skills, and I have been following his writing for some years now.
I was deferring to experts in the field of readability, and considered it likely that they would provide a better measure than self-reports of “looks fine to me.”
Further, it seems likely to me that Eliezer is very good at targeting his audience and maintaining interest despite the complexity of his prose. Academic doesn’t mean “boring” by necessity. One of the references from the Gunning Fog page states:
Although we have often given permission for reprinting the Fog Index, our means of measuring reading difficulty, we have sometimes cringed at the use made of it. In our work, we emphasize that the Fog Index is a tool, not a rule. It is a warning system, not a formula for writing. Testing without the support of analysis based on experience can be detrimental.
And yes, I do realize that this criticism can be applied to my own use of the tool, but point out that the measure directly supports my initial statement: “Use longer sentences and bigger words,” with the caveat that you should also be a good writer, to ensure the complexity doesn’t hinder the message. Or I could add, only do this if you can get away with it (still be a successful communicator).
I’d also like to point out that this feels like a good example of the dojo-style response to my clumsy use of a single word: academic.
For better or worse, [EY] enjoys a special status in this community.
A status earned precisely by writing posts that people enjoy reading!
If you’re suggesting that the ordinary academic/intellectual norm of only allowing high-status people to write informally, with everyone else being forced to write in soporific formal-sounding prose, is operative here, then I suggest we make every effort to nip that in the bud ASAP.
It’s particularly frustrating, because encountering it feels like discovering that you’ve overestimated your audience at the same time that they’ve underestimated you.
It feels the same way from the other end too! I.e. “Really? I have to explain this to you?”
Yes—I have seen this so many times!
It’s particularly frustrating, because encountering it feels like discovering that you’ve overestimated your audience at the same time that they’ve underestimated you.
I’ve noticed this too, and I long for the day when our rationality skills have advanced to the point where we can be rational and nice.
I haven’t really seen 3), and EY’s posts undermine 1) significantly, it seems to me.
Of his most recently posted articles: Undiscriminating Skepticism scores at a Flesch-Kincaid grade level 17 and a Gunning Fog index of 17.9, You’re Entitled to Arguments is at 16 and 17.8, and Outside View as Conversation Halter is at 14 and 14.5. Note that a score of 15+ is considered academic writing by these measures. Tests of his recently upvoted comments show scores ranging from 7 to 20.
Here’s the Flesch-Kincaid calculator I used, and the Gunning Fog calculator. I would be surprised if other measures of readability, and tests of his other posts, did not show it to be academic-level writing.
“Undiscriminating Skepticism”—why, that’s ten (10) syllables right there in the title! My head is already spinning!
Seriously: tests like those do not control for the content or subject matter of the writing. There exists, furthermore, a significant subset of the (adult!) human population who would consider a phrase like “undiscriminating skepticism” itself to be difficult and unusually abstract. Needless to say, tests which heavily weight the judgements of such people are not very useful for the purpose of judging “readability” in most contexts here.
If you want to judge the readability of LW posts, I suggest spending some time reading typical articles published in academic journals.
You’re right! It’s agonizing! Oh the pain of posting and reading here! My mouth is bloodied. You have defeated me, oh wise and amazing person who obviously knows better and is fully within their right to ridicule every attempt I make to explain the use of a single word in a sentence whose structure is still largely intact, as it was meant to be a frickin’ suggestion.
good jorb.
(addresses both of the posters above)
Wow, sarcasm. That’s original.
And yet I find his writing a model of clarity here, despite a few randomly chosen articles by other people having a far lower Fog Index. How useful are these indices? On the Gunning Fog page it says “The higher the Fog Index the trickier it is to read.” But the Wiki pages for these tests reference no empirical studies.
I find his writing to be very readable as well. However, I consider myself highly educated, with excellent English skills, and I have been following his writing for some years now.
I was deferring to experts in the field of readability, and considered it likely that they would provide a better measure than self-reports of “looks fine to me.”
Further, it seems likely to me that Eliezer is very good at targeting his audience and maintaining interest despite the complexity of his prose. Academic doesn’t mean “boring” by necessity. One of the references from the Gunning Fog page states:
And yes, I do realize that this criticism can be applied to my own use of the tool, but point out that the measure directly supports my initial statement: “Use longer sentences and bigger words,” with the caveat that you should also be a good writer, to ensure the complexity doesn’t hinder the message. Or I could add, only do this if you can get away with it (still be a successful communicator).
I’d also like to point out that this feels like a good example of the dojo-style response to my clumsy use of a single word: academic.
What works for EY may not work for everyone else. For better or worse, he enjoys a special status in this community.
A status earned precisely by writing posts that people enjoy reading!
If you’re suggesting that the ordinary academic/intellectual norm of only allowing high-status people to write informally, with everyone else being forced to write in soporific formal-sounding prose, is operative here, then I suggest we make every effort to nip that in the bud ASAP.
This is a blog; let’s keep it that way.
It feels the same way from the other end too! I.e. “Really? I have to explain this to you?”