This is a very beautiful and logical interface, but it's the wrong approach, because it's a very developer-centric approach to writing.
The problem with writing is you can't loop through it and find whether each sentence passes or throws an exception. A written work needs to be evaluated as a cohesive whole. That's what "bold and clear" writing means to me: a written piece of work that stands on its own and says what it means.
Computers are not smart enough yet to understand why "Lolita, light of my life, fire of my loins. My sin, my soul. Lo-lee-ta: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta." is a complete, perfect paragraph. It doesn't have a verb and it looks like Lola is misspelled multiple times, so it doesn't pass the subset of grammar rules set up in the backend. But the meaning, the essence of the paragraph is clear.
Writing may someday be able to be governed by algorithms, but not yet. I ran the second paragraph of David Copperfield through Hemingway [1], and it gave me too many adverbs, a misspelling of the British neighbourhood, and the use of passive voice. This is understandable, as Charles Dickens was a verbose writer who got paid by the word. And yet, it doesn't detract from the fact that he is one of the most-loved in the English cannon.
We can't measure good literature yet, because there is no straightforward formula, and although this is an interesting attempt, it can't teach good writing better than a human.
For a better, and still technical, approach to understanding how and why sentences and paragraphs work with us or against us, it's better to read Strunk and White, and even better to read "How Fiction Works" by James Wood.
If there is a way to incorporate at least those two books into conditional statements, I would be excited to see it.
> Computers are not smart enough yet to understand why "Lolita, light of my life, fire of my loins. My sin, my soul. Lo-lee-ta: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta." is a complete, perfect paragraph. It doesn't have a verb and it looks like Lola is misspelled multiple times, so it doesn't pass the subset of grammar rules set up in the backend. But the meaning, the essence of the paragraph is clear.
Eh, this isn't so infeasible. Any wide-coverage, statistical grammar will have a TOP --> NP production, as well as the usual TOP --> S. i.e. the grammar must accept NP-only sentences; they're too common to reject as mistakes.
I think it's reasonable to have a linter that throws a warning for that sort of sentence though. By all means dismiss them. But it can help you catch things you didn't intend.
The problem with writing is you can't loop through it and find whether each sentence passes or throws an exception. A written work needs to be evaluated as a cohesive whole. That's what "bold and clear" writing means to me: a written piece of work that stands on its own and says what it means.
Computers are not smart enough yet to understand why "Lolita, light of my life, fire of my loins. My sin, my soul. Lo-lee-ta: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta." is a complete, perfect paragraph. It doesn't have a verb and it looks like Lola is misspelled multiple times, so it doesn't pass the subset of grammar rules set up in the backend. But the meaning, the essence of the paragraph is clear.
Writing may someday be able to be governed by algorithms, but not yet. I ran the second paragraph of David Copperfield through Hemingway [1], and it gave me too many adverbs, a misspelling of the British neighbourhood, and the use of passive voice. This is understandable, as Charles Dickens was a verbose writer who got paid by the word. And yet, it doesn't detract from the fact that he is one of the most-loved in the English cannon.
We can't measure good literature yet, because there is no straightforward formula, and although this is an interesting attempt, it can't teach good writing better than a human.
For a better, and still technical, approach to understanding how and why sentences and paragraphs work with us or against us, it's better to read Strunk and White, and even better to read "How Fiction Works" by James Wood.
If there is a way to incorporate at least those two books into conditional statements, I would be excited to see it.
[1] http://imgur.com/k9hsHfj