Conference thebay::joyoflex

    It's all to do with who you are writing your document for . It is
    an indicator test which should show you the level at which you
    have/should aim your text . 
    
    For instance , the Times would have a high index , and the Sun would
    need a fraction to several decimal places .
    
    Or vice versa .
    
    John J

    Rudolf Flesch, I think.  He thought that short sentences with short
    words in them were easier to read that long ones full of long ones.
    He's probably right.  The index is a way of expressing that quality.
    There are others, one called the Muddy River, or Clear Water, or
    something like that.  The resulting figure is usually stated   
    as years of education.  A corrolary of the theory is that people are
    most comfortable reading about two levels below their number of
    years of education.                                            
    
    I ran some of my stuff through one of these formulas and got a good
    grade, but I don't think a lot about it.  Still, in any organized
    system of thought and communication there are bound to be places
    where long Latinate words are needed to express exactly what is
    meant.  Technical jargon is no crime when used to talk to another
    member of the cult.  On the other hand, why not be clear?  I guess
    I think these theories are interesting toys without much practical
    value.
    

re. .0:

Hey Derek, could you give us a pointer to the program?  I'd like to play with 
it.

Thanks,

Larry

    I guess I'd have to argue the value of these systems, at least in
    theory.  Dredging back to the distant past, I once took some teaching
    of reading classes for teaching public school classes.  These systems
    are useful to get at least some feeling for the suitability of the
    text for the reader.  Also useful for copywriting of advertising
    and marketing materials, particularly direct mail.  If you want
    people to read a sales pitch for something they don't know they
    want :^), you have to make it as effortless as possible to get the
    message.  A lot of scientific and technical publications that are
    dry as dust might benefit from the systems, also, although interesting
    authors and non-pedantic editors might be more helpful!

    Anyway, they do have uses, and they are fairly well supported.
    
    Sherry
    

    G'day,
    
    Re -.a couple.. Hi Larry,
    I believe it was in toolshed library or somesuch... I got it back in
    March, so the memory fades  ...   :-(   (but I will try and find
    it again)
    
    However, I've the .exe available. If anyone wants it, drop a line...
    
    I certainly kind of find it useful to compare documents usingthe
    program. It gives a kind of neat measure. Foe example, as mentioned
    earlier, for good sales letters, there should be no more than seventeen
    words in a sentence, _they_ (whoever _they are_) say. Words should
    have as few syllables as possible. The program prints out all the
    3 or more syllabic words - which can be sort merged with no duplicates
    for the definitive list - . Makes for interesting comparisons too
    between engineering documents and marketing ones...
    
    derek
    

    G'day again,
    
    
    FOUND IT..... phew ! (took an hour!)
    in SUBWAY::SWSHARE 
    Hope Barry Dysert will not mind my extracting his note..
    
              <<< SUBWAY::DISK$D1:[NOTES$LIBRARY]SWSHARE.NOTE;1 >>>
                              -< Software/SHARED >-
================================================================================
Note 102.3             Tool that counts words in document.                3 of 5
DYO780::DYSERT "Barry Dysert"                        14 lines  19-FEB-1988 11:41
                           -< here's one that works >-
--------------------------------------------------------------------------------

    Feel free to copy:
    
    	DYO780::DISK$USER04:[DYSERT.PUBLIC]FLESCH.EXE
    
    It's a program I came across a long time ago that counts words, plus it
    does a whole lot more. It works on standard text files, though, so
    you'll have to extract your WPS file to something normal. 
    
    Invoke it via:
    
    	$ FLESCH:=$FLESCH	! must be invoked via foreign command
    	$ FLESCH input,output	! "output" can be sys$output
    
    Enjoy!
    
    ___________________________________________________________________
    
    Its written in PL/I (or appears to be) very unforgiving in the command
    line.
    
    derek
    

     Neat!  Example follows.  One mistake that it made was with
     "NODENAME" -- it only has two syllables.
     
     Dan
     
$ flesch login.com,z.z
$ type z.z
  WORDS IN TEXT WITH 3 OR MORE SYLLABLES  
******************************************
  
  
deramo
NODENAME
INTERACTIVE
terminal
terminal
application_keypad
DERAMO
logicals
gnu_emacs
DERAMO
BELINI
LISPUTIL
LISPUTIL
noverify
deramo
executed
conference
EDIT_MODE
     
     
SUMMARY STATISTICS OF ALL PARAGRAPHS 
***************************************
NUMBER OF SENTENCES     =        46.0
NUMBER OF WORDS         =       176.0
NUMBER OF SYLLABLES     =       274.0
AVERAGE SENTENCE LENGTH =         3.8
AVG. SYLLABLES PER WORD =         1.5
FLESCH INDEX            =        76.0
GRADE LEVEL EQUIVALENT  =         7.4

    If that's the same program we used in my old writing group -- and
    the part about PL/I errors in the command line makes me think that
    it probably is -- it gives a falsely high score for passages that
    contain sentences joining clauses with "and."   The Flesch system
    counts each such clause as a separate sentence, while the program
    counts everything up to the period. 
    
    If it's the same program, it counts words such as "joining" as two
    syllables when for readability purposes it's one, "join," with a
    verb ending.  Verb endings don't affect readability. 
    
    The Flesch readability theory doesn't say that you should never
    use long words, Latinate words, or long and complex sentences,
    only that your writing will be better understood if on the average
    the sentences are about 20 words long and average a bit under two
    syllables a word. The idea is that if you are faced with
    unavoidable complexity, you should do your best to avoid the
    complexities you can avoid. If you must use words the average
    person probably isn't familiar with, you can help the reader's
    comprehension by avoiding other unfamiliar words that aren't
    essential to your presentation. 

    Or, in other words, if you have to use big words to get your
    idea across, make sure your sentences are simple and your other
    words are concise and clear.  Otherwise, you'll be forcing
    the reader to deal with a lot of extraenous garbage...

    --bonnie

    I heard about the idea from my father about 20 years ago, only he
    called it a "fog index".
    
    Stephen

    Neat!  I just tried this with the following sentence ...
    
    This is a supercalifragilisticexpialidocious sample.
    
    It had a flesch index of -52, but a grade equivalent of 28.3 !
    
    At long last someone who appreciates all my years of education!!!
    
    stuart

    Re: .10
    
    That's the best laugh I've had all day.
    
    Thank you,
    
    twe

Here we go with another digression...

The Fleschmann index doesn't really deal with word length per se, but rather 
with the number of syllables in the word.  This started me thinking about long 
single syllable words.  With a little thought, I came up with "brougham"
(pronounced "brome", -  a closed carriage with an open driver's seat) which is
an eight-letter monosyllable. 

Anybody got any longer English language words of one syllable?  Any other 
eight-letter words?

Larry

    
    
    			strengths
                        
    
    					andrew
    

    re .12
    
    But Brougham is only pronounced "brome" in some areas ...
    
    where I came from it was "brogum" with a a long o and an aspirated
    g rather than a gutteral g (more like an aspirated h but not quite).
    
    other places it is "bro-am"
    
    so single syllable is moot.
    
    stuart

    8 letters : schmaltz,schnapps,schmooze,squawked,thoughts
    9 letters : squelched,stretched

    Note that "strengths", mentioned in a previous note, also has the
    highest ratio of consonants to vowels. 

    Sorry, but like my earlier comment on brougham, I don't by the
    idea that words like stretched are only one syllable.
    
    sure, the e goes silent, but that doesn't remove the written syllable.
    
    stuart

     Aaaaaaaaaarrrrrrrrrrgggggggggghhhhhhhhhh!!!!!!!!!!

Isn't that called Stretching the Agony ?
    
    

    An automated Flesch tester, assuming "brougham" to follow the same
    rules as most other English words, would probably figure it to have
    two syllables. 
    
    If we think "brougham" is as readable as the average two-syllable word,
    then that's the right behavior. 
    
    Now, would the tester calculate one syllable or two for words like
    "stretched"? Depends on how smart the algorithm is. I wrote a little
    Flesch tester once (in VAXTPU; it's in EVETPU::TPU_PROCS) that figures
    out whether a final "ed" is a separate syllable. (Like any algorithm
    for counting syllables, it's not perfect.) 
    
    The techniques of Flesch, Gunning (originator of the Fog Index) et. al.
    were designed to be easy to use in the days before anyone heard of word
    processing -- it's fairly easy to count up the syllables in a piece of
    prose by hand. 
    
    					Val

    You don't need to run your whole document through a Flesch
    test to measure its level, either.  Just pick a typical passage
    of 3 to 5 paragraphs from three to five random locations in
    your book and do a quick count by hand.
    
    --bonnie 

    G'day,
    
    Being born lazy and only able to count to 1023 on my fingers, its
    quicker to use the program! and I get a printable record that I
    can compare with other documents' figures ....
                                              
    derek
    

    Huh?  How's anyone supposed to understand the "flesch" program?
    Look at all this gobbledegook in its output!
    
    SUMMARY
    STATISTICS   (virtually unpronounceable)
                 (er, uh, I mean, yuh can hardly pronounce it)
    PARAGRAPHS 
    SENTENCES
    SYLLABLES
    AVERAGE
    SYLLABLES
    EQUIVALENT   (4OUR sllables, count 'em)
                       -------- ((notice, only 2 sllables in _this_ word)

    re .12
    
>   With a little thought, I came up with "brougham"
>   Anybody got any longer English language words of one syllable?
    
    Obviously, "broughammed".

	Readability formulas are a controversial area in writing
	research.  They were originally developed to come up with an
	"objective" and easy way to measure the simplicity of text.
	Here are some of the better known ones:

	o  Flesch	RE = 206.835 - .846 WL - 1.015 SL

		RE -- value between 0 and 100
	   	WL -- word length
		SL -- sentence length

	o  Gunning Fog Index	RGL = .4 (ASL = %PW)

		RGL -- reading grade level
		ASL -- average sentence length
		%PW -- percentage of words of 3 or more syllables

	Others include Dale-Chall, Coleman, Kincaid-Flesch, FORCAST,
	RIDE, and Cloze.

	Readability formulas have the following problems:

	o  Short sentences and short words do not good text make.  They
	   may be correlative with good text, but this is by no means
	   necessarily the case.  Gibberish, text entered in backwards,
	   etc. can get quite good results (Lewis Carrol's "Jabberwocky"
	   gets 90+ on the Flesch scale).  Further, readability
	   formulas say nothing about discourse, context, audience ...

	o  There is little research on how well good RF scores correlate
	   with actual readabilty -- and, thus, can predict readability.
	   Further, there is little agreement among the formulas themselves
	   as to predicted grade levels for the same text.

	o  RFs fail miserably as prescriptions.  RFs are indicative of
	   bad writing.  When they are used as prescriptions (i.e., run
	   your stuff through an RF program, get a bad score, shorten your
	   sentences and use shorter words, run it again, get a good score),
	   they tend not to improve readability.  In fact, the more they are
	   emphasized, the less improvement.  One researcher called this
	   phenomenon lighting a match under the thermostat in order to
	   warm up the room.  Making text readable is a difficult, time-
	   consuming task -- relying on RFs to make something "readable"
	   is the easy way out.

		-- Cliff

    Re: .12 and .24
    
    strengths
    
    It also has the highest ratio of consonants to vowels of any word.

Apropos of readability scoring, this article from todays Vogon News* is 
interesting:


>  Apple - "Is the MAC Making Applesauce of Student Prose?"
> 	{Business Week, 23-Apr-90, p. 120C}
> 	{Contributed by John Ellenburger}
>    The Apple Macintosh, with its easy-to-use graphics style, may represent a
>  victory of form over substance. That appears to be the finding of research
>  conducted at the University of Delaware. Marcia Peoples Halio, assistant
>  director of the English Dept.'s writing program, assigned the same five
>  teaching assistants to 10 freshman English classes. One student group used
>  Macs, the others used IBM PCs or IBM clones.
>    The instructors said the Mac's large type and graphics seemed to lead to
>  "sloppier writing and fluffier topics." A writing analysis program of a random
>  sampling of papers found that 30% of the Mac writers used complex sentences,
>  compared with 50% of IBM-clone writers. Sentence length averaged 16.3 words
>  for the Mac essays and 22.6 for those written on PCs. And the Kincaid Scale, a
>  measure of readability, showed Mac users writing at the 8th grade level, vs.
>  12th grade for the IBM-clone group. Her article "Student Writing: Can the
>  Machine Maim the Message?" concludes that the Mac's format seems to "encourage
>  a simple sentence structure and childish vocabulary."
> 

* As per usual, reprinted without permisision from Vogon News Service, edition 
  #2046

    G'day,
    
     in a lighter vein...
    re long words
    
    Smiles
    
    because
    
    
    Its a mile between the esses [or is that ss or 's's or....;-)   ]
    
    
    derek

	Well, now that this note has totally detiorated ...  The 
	longest word in existence is the one that comes after "and
	now a word from our sponsors."

		-- Cliff

>
>        Well, now that this note has totally detiorated ...  The 
>        longest word in existence is the one that comes after "and
>        now a word from our sponsors."
>

I've never heard an ad mention antidisestablishmentarianism ...

mind you I think most ad writers would think that the dissolution of the
monestaries was some chemical property of money!

Stuart

(Mind you, seeing what has been happening around the company these days, I
think the term could be applied to DEC workers ...  I certainly feel
antidisestablishmentarianistic as I see some management type decisions
that seem to be blowing away some of the company lore.)