Microsoft Word Criticizes Homer - Impressions and Expressions of Ijon — LiveJournal
Microsoft Word Criticizes Homer|
|Date:||September 28th, 2004 10:42 pm (UTC)|| |
Of course it doesn't have an ancient Greek dictionary or grammar engine!
That's just the thing -- it presumes to check grammar in Greek text. Hell, it's not even transliterated Greek -- it's actual Unicode polyphonic Greek! Why would it even try? Just an oversight, I suppose.
And yes, I'm far from admiring its performance even on English text. I think grammar checkers are useless for people above intermediate level.
|Date:||September 28th, 2004 11:12 pm (UTC)|| |
Ahem... how many people are above intermediate level? Percentagewise?
The grammar checker would have some pointed comments about the Constitution, not to mention about the spelling of the Ten Commandments (to mention just a couple of texts basic to our culture).
For my part, I am tired of being “corrected” by Anglophones who have learned their contemporary English orthography and grammar from Word. And I expect that, if Word did begin to recognize Epic poetry, then it would also insist that it rhyme and that its meter be based upon a simple count of syllables.
|Date:||September 29th, 2004 04:29 am (UTC)|| |
Syllable count is not so simple.
The truth or falsity of that assertion depends, of course, on the metric for simplicity.
|Date:||September 29th, 2004 07:24 am (UTC)|| |
How about "you can code a prototype in a week and not have people laugh at the results".
The problem is that many words that have only one dictionary syllable interpretation sometimes get syllables dropped so they fit the meter. This "sometimes" is not usefully formalizable (Hi, avva
I was under the impression that MS coded prototypes in about a week, and had Zipf-Davis fawning over them; meanwhile, we can probably find someone to laugh at any code, no matter how well written.
But you seem to have confused the simplicity of the count with the ease of the counting.
|Date:||September 29th, 2004 01:37 pm (UTC)|| |
we can probably find someone to laugh at any code, no matter how well written.
Yes well, that was obviously an informally stated way of saying the result will be pitiful, and that I don't think you need a very precise metric. The point is that meter is quite convention-based. You could, I suppose, compose a survey of critical satisfaction over a machine-generated scan, but I bet you'd run into practical problems perfoming it.
What, again, am I confusing?
The point is that meter is quite convention-based.
Not a point to labor, since the force of my initial cynical claim (that MS would insist upon simple syllable count) comes from the fact that the convention for Epic poetry is radically different from that which is now typically used in English poetry.
What, again, am I confusing?
Simple count with easy counting. The task of counting syllables would not be easy; the result would be simple, and inappropriate to Epic poetry.
|Date:||September 29th, 2004 04:10 am (UTC)|| |
Why would it even try?
Either it thought it was Greek, in which case, saxtein that it does have a Greek engine, or else it thought it was English (or Hebrew) because it lacks Evan's language detection code, in which case yes, it's an oversight, but not something that's *trivial* to fix.
Hey, it comes with an off switch.