Path: utzoo!utgpu!utstat!jarvis.csri.toronto.edu!mailrus!iuvax!ndcheg!uceng!dmocsny
From: dmocsny@uceng.UC.EDU (daniel mocsny)
Newsgroups: comp.ai
Subject: Re: Writing style analyzers
Summary: The Man. The Machine.
Keywords: Understanding and Comprehension, Reality and Modeling, Sentience
Message-ID: <688@uceng.UC.EDU>
Date: 13 Feb 89 19:15:59 GMT
References: <18464@santra.UUCP> <1241@arctic.nprdc.arpa> <904@ubu.warwick.UUCP> <44802@linus.UUCP>
Organization: Univ. of Cincinnati, College of Engg.
Lines: 57

In article <44802@linus.UUCP>, bwk@mbunix.mitre.org (Barry W. Kort) writes:
> Just for fun, Dan, I took the text of your posting and ran it through
> WWB (Writer's WorkBench).  Here are the condensed results...

I was tempted to run your post through Rightwriter. But first I
checked with my homunculus, and he said he's had enough of infinite
regress for one lifetime. Besides, it would be something like setting
up a dialog between Racter and Eliza.

>     WWB found no spelling errors, no punctuation errors, no split
>     infinitives, and no double words. 

So do I get the job, or what?

I'll admit that before I looked into this grammar-checking stuff, I
was not aware that the split infinitive was something "to really
avoid." :-)

>     READABILITY
>     
>          The Kincaid readability formula predicts that your text
>     can be read by someone with 10 or more years of schooling,
>     which is a low score for this type of document.

Rightwriter reports a similar index. When I first started using
Rightwriter, I was consistently scoring 17--20. Now I get the same
ideas across with 9--11's. Unfortunately, many people may misunderstand
the scoring, and mistakenly think that "writing on a higher grade level"
implies something other than useless semantic overhead. I wish these
tools could also report an "essential complexity index" to show how
complex the underlying factual content really is. The difference
between the Kincaid index and the essential index (when nonzero, and
thus positive) would measure accidental complexity.

Certain documents, such as insurance policies or legal briefs, would
typically have large accidental indices.

>          You have appropriately limited your use of passives and
>     nominalizations (nouns made from verbs, e.g. "description").

Or, e.g. "nominalization," or "use of." (Do as I say...) :-) I make a
game now of trying to avoid passives as much as possible while
simultaneously using "we" only when necessary. Sometimes this forces
me to think very hard about who or what is really doing what to whom
or what. This is the writer's job, not the reader's. Nominalizations
are now a pet peeve of mine. I only use them when I can't think of an
alternative.

> Although WWB did flag a few things differently than Rightwriter,
> they seem to be interchangeable.

WWB seems to do a bit more.

> --Barry Kort

Dan Mocsny
dmocsny@uceng.uc.edu