Magnum One

home *** CD-ROM | disk | FTP | other *** search

/ Magnum One / Magnum One (Mid-American Digital) (Disc Manufacturing).iso / d3 / style.arc / STYLED.DOC < prev next >

Wrap

Text File | 1988-08-15 | 46KB | 1,170 lines

STYLED.DOC The Manual for STYLED and STYLLIST Version 1.7 (c) 1987 by Louie Crew Chinese University, Shatin, Hong Kong --------------------------------------------------------------------- Copy freely. Pass it on. Do NOT sell. --------------------------------------------------------------------- To use: STYLED requires MS-DOS, version 2.1 or higher. 1) Put STYLED on a disk with at least one text file. 2) Type STYLED plus a carriage return. 3) Follow the menus. 4) STYLED automatically creates files to store 4 word lists. They could require as much space as your original text, but rarely do. Save plenty of space. or, for a report to disk only 1) Put STYLLIST on a disk with at least one text file. 2) Type STYLLIST plus a carriage return. or STYLLIST plus the name of a file to analyze The second choice automates everything and returns to MS-DOS after it has created REPORT. The first choice (slower) allows you to enter the filename from a prompt within the program. REPORT lists your long words, your forms of "to be," and your potential nominalizations, without some of the overkill in STYLED. It lacks the punctuation report and other screen graphs in STYLED. Why? STYLED and STYLLIST chart patterns in texts. I created them to help me revise. I now also use them to analyze what others write. They help me as I compare professionals with amateurs. Each program serves as a heuristic. It only describes: it cannot prescribe or remedy. It can help a clever person to identify and arrange insight. A dullard a dullard will remain. Each program abstracts several forms from texts. Writers must integrate form and content. Form alone lacks efficacy. Witness thousands of bad poems in good iambic meter, hundreds of bad symphonies in perfect 3/4 time, millions of discarded images painted with the finest of oils. Monsieur Jourdain, beware! Each program detects patterns. Clever writers will preserve some, expunge others. Each program can freshen the air for those who know little to do with a text except to check for "mistakes." At the least, each program orders dickering. Occasionally they prompt a reVision, a new way to see. Epiphanies come rarely, however. Only muses can program those. The programs offer no help with pre-writing and early drafts. They attend to matters that most writers notice only minimally in early on. The programs serve those who think they have said something worth the time to polish. Revision can gobble huge amounts of time. Many people revise for hours without improving. Many lack strategy. These programs can help, modestly. I sometimes spend an entire day revising one page. When I can't, I wish I could. I wish more of the writers I read, would. Warning: Do not let this program waste your time. I have appended samples of my own sessions with the program. You should find many other uses. I cringe when I imagine mindless ways to use the program. For example, the program quantifies transitionals. Transitionals often evidence cohesion, but only if the writers have used them accurately. For some writers, THUS=ALSO, MOREOVER=NEVERTHELESS, etc. This program will not detect their confusion. No program protects us from sophomores. Thank the goddess, most grow up. I smile when I imagine those who think these matters completely unimportant. At Breadloaf, Robert Frost used to talk for hours about a minute effect of meter. "But surely you don't think about such things when you write!" allegedly a poetaster exclaimed. Frost teased, "About little else." Too many people think that mystery abides only in abstracted ideas, that the package counts for little. Like choristers who sing in showers only, these compose not, and woo an audience of only one. Interpret, but distrust your interpretations. Features most important often occur but once. Each program identifies slots. You must imagine wise ways to fill them. ==== What? First you tell STYLED the name of the text you want to analyze. You can name the file from the command line in one of two ways: STYLED file.txt or STYLED #*.txt The second way loads STYLED and start with a list of *.TXT files for you to choose from. You can also simply load STYLED The program will then prompt you: Answer # (plus wild cards) if you you wish to see the directory. Hit just <ENTER> if you you wish to end. Which TEXT file would you like to analyze? Note that STYLED allows you to use MS-DOS wild cards (? or *) after the symbol `#' if you want to view a directory. You may ask to see a list of the files on a different drive or subdirectory as well, as when you are logged on A: and enter for the FILENAME #B:\mysubdirectory\*.DOC You will then see all the files with the extension .DOC on \mysubdirectory on the B: drive. The Main Menu Once you have told STYLED what text you want to analyze, the main menu gives you STYLED's menu initially gives eight choices: Regarding \l\problem.txt Analyze 1 = Word length 2 = Punctuation 3 = Syntax 4 = Nominalization ---------------------- (V)iew the file (N)ame another file (S)hell to MS-DOS (E)nd a session Choice: Any of the first seven will return to the menu. Use the E key to exit to the MS-DOS prompt in an orderly way. The V choice lets you view the text. You do not have to leave the program to return to your word-processor. The program shows the text to you one screen at a time. Hit ESC to exit, any other key to continue whenever the program pauses. The program numbers each screen when you view the text. The S choice lets you leave STYLED temporarily. While you work at the MS-DOS level, STYLED remains in memory. To return, type EXIT at the MS-DOS prompt. You may even run some other programs while you have Shelled, if your computer has enough memory. ======================================================================== Note: if you run STYLLIST, that program bypasses these the choices and directly prepares one file which reports the long words, the forms of TO BE, and the potential nominalizations. ======================================================================== At this point, I recommend that you stop. If you have not done so, run STYLED and STYLLIST. Investigate one of the samples or test a text file of your own. Return to this document only to reflect about how to interpret the data. xxxxxxxxxxxxxxxxxx The Reports 1. Word Length. This report replaces each word with a bar the length of the word, and it preserves the terminal punctuation. It prints in dark-on-light only words with 10 or more characters. It adds a sentence number, totals the words in each sentence, and calculates: Total word count: Words 10+ chars.: ( %) Aver. wds/per/sent: Then it lists all the long words with a sentence number. 2. Punctuation. (STYLLIST does not include this report.) All punctuation blinks. One diamond character represents each word. Afterwards, the program tallies the punctuation and computes the average interval. 3. Syntax. (STYLLIST does not include this report.) Structure words ( possible coordinators, subordinators, and transitionals) appear in high intensity. First and last words of sentences, as well as words before and after internal punctuation, appear in low intensity, unless reported as structure words. A dull bar replaces each other word. The program tallies the structure words and it lists them, each with its sentence number. 4. Nominalization. Forms of the verb TO BE appear in dark-on-light letters. Many words that may hide action appear in high intensity. All other words appear in low intensity. After the program tallies forms of TO BE and the words that may hide action, it lists them on screen with sentence number. How to Use These Data Your own goals should dictate how you use any statistics. Pattern to integrate content. What should you emphasize? What should you intensify. What junctures should you stress as such?.... What does your audience expect? How can you use their expectations to engage them on your terms? Or on theirs? If STYLED's graphs correspond in no way to what you intend, put a new Duracel in your inner ear and reprocess the words. If all sentences look alike, rearrange them to serve a purpose, possibly one you had not seen? If you detect monotony, make it serve you or delete it. A writer who strives primarily for style seems a bit like someone who arranges books by color. A writer with style, resembles someone who arranges some books by size. Thurber's editor warned him not to use the colon. Thurber persisted. One day the editor came to Thurber's desk, touched the colon, and ripped the key from Thurber's word-processor. Scholars have written about "Tennyson's scissors." Some verbal fetishes enrich; others impoverish. STYLED can help you only to spot them, not to evaluate them. The program will not zap any combinations on your keyboard or printer, but you should. Use the program to sample published texts which share your goals and your audience. Interpret carefully. Generalize reluctantly. Reverse a common practice if to reverse improves your manuscript. Norms should not imprison, but should liberate. Attend them consciously, cautiously. Visit the sculptor's private studio, not the gallery. Suspend reverence. Let the moment demystify. Ask how you might sculpt better. Models which deserve your praise will survive. In time you will discover your own norms. AUTHOR derived from AUCTOR, L., `originator,' which in turn derived from AUGERE, `to increase,' `to produce.' Those faithful to this etymology do not merely copy, but originate and augment. Word Length Readers understand and remember shorter sentences better than longer ones. Longer sentences suggest complexity. Shorter sentences suggest simplicity. Clever writers and readers master both. Some pack complexity into few words. Others uninterrupt simplicity beautiful for many lines. Many writers use short sentences to clarify the steps of a process. Some writers subordinate profusely to illustrate delicate logical relationships between abstract ideas. How does what you want to say constrain how you might say it? I like to require my "long words" to serve major rather than minor plans. When I see my long words lifted alone from the text in LONGWORD.REP, I do not welcome those for which I had no special purpose. A company will soon fail if its managers do not contribute to the company's net worth, especially if they ride in limousines. Shakespeare reserved limousines like INCARNADINE for matters hunky-dory. I try to identify clusters of long or short bars. If many short words cluster, do I dictate? Do I sound childish? If many long words cluster, do they cloud? What verbs can I use to unpack action? What clearer words can I use? How long is long? Each must decide. Another style-checker, Grammatik, (now released as Grammatik II, Reference Software, 330 Townsend, Suite 232, San Francisco, CA 94107) calls sentences under 14 words "short," sentences over 30 words "long." Just as arbitrarily, STYLED calls "long" all words with 10 or more characters. You decide. Some sociologists call all adults from 21-35 "young", from 36-50 "middle-aged", and from 50-65 "old". Since they give me only 7 more months of middle age, I dislike their arbitrary trifurcation of the work years, but I know what they analyze. Distrust categories; but respect them. Punctuation Punctuation often reveals whether I have stressed as I planned. For example, Sometimes I begin moderately and intensify later. At other times I begin dramatically (as with a short sentence, a long one, or a highly modified one) and then moderate, as with a stretch of sentences less varied. Sometimes I risk monotony and then surprise. Contexts influence the effects of patterns in prose just as much as in music. A series of short sentences from Hemingway will not likely function the way the short sentences function in a pamphlet to accompany aspirin. STYLED can help me spot places where I have deafened. Use the program to report how the translators punctuated STPAUL.TXT. Many consider this passage one of the most beautiful in English literature: perhaps only the style compels them? The substance cynically indicts everyone, especially in the Greek. Use STYLED as a stethoscope to monitor the pulse of any "set piece"--by Adrienne Rich, Lincoln, Sojourner Truth, Jefferson, Chinua Achebe, Maxine Hong Kingston..... Syntax Some words, like bones, structure the skeleton of a text. STYLED monitors three types of "structure words": coordinators, subordinators, and transitionals. When sentence after sentence appears with only the first and last words, the rest in bars, either the text does not cohere well or the writer has fleshed with other words to hide the bones. When the program shows structure words at frequent intervals, either the skeleton protrudes or supports a complex body of text. Compare the cohesion of the sample STPAUL.TXT. If I never subordinate by pattern, I might try to do so at the part of the paper that I think most important. Beware: the computer does not know what constitutes a subordinator, a transitional, or a coordinator. The word THAT, for example, can function in a variety of these ways, or just as a demonstrative pronoun. The favorite of grammar tests-- That that that that that boy used is wrong is obvious. won't confound this program. [We need no program to call this sentence terrible!] STYLED reports the 5 THATs as potential structure words, and reports the 2 forms of TO BE. Specifically the program reproduces any word which matches one of these 56: ACCORDINGLY, AFTER, ALSO, ALTHOUGH, AND, AS, BECAUSE, BEFORE, BESIDES, BOTH, BUT, CONSEQUENTLY, EITHER, ETC., FIFTH, FIRST, FOURTH, FURTHERMORE, HENCE, HOW, HOWEVER, IF, INDEED, INSTEAD, MEANWHILE, MOREOVER, NEITHER, NEVERTHELESS, NOR, OR, OTHERWISE, PROVIDED, SECOND, SIMILARLY, SINCE, SO, STILL, THAT, THEN, THEREFORE, THIRD, THIS, THOUGH, THUS, UNLESS, UNTIL, WHAT, WHEN, WHENEVER, WHERE, WHEREAS, WHETHER, WHICH, WHILE, WHO, WHOM The Cholesterol Report Forms of TO BE The program finds AM, ARE, AREN'T, BE, BEEN, BEING, IS, ISN'T, WAS, WASN'T, WERE, WEREN'T, I'M, YOU'RE, HE'S, SHE'S, IT'S, THERE'RE, THERE'S, WE'RE, THEY'RE, WHO'S, WHICH'S, THAT'S, and WHAT's--regardless of case. These forms almost always force a writer to bury action in less forceful words, usually in nouns. These forms locate all passives (except the lone past participle) and all expletives. What's at stake: When you fetch your transcripts, only the novice tells you: "I have lost your records." The experienced tells you, "Your records have been lost." (Passive: agent dropped.) Those ready for Washington say, "There has been a loss of your records"--with no hint of an agent who responds. In documents which I edit most thoroughly, I try to restrict TO BE to situations in which I name or identify--or to those sentences where all other alternatives sound silly. Nominalization Buried action can clog a text, like verbal cholesterol. When a noun contains an action, we call it a nominalization. NOMINALIZATION illustrates what it names: "-ize" buries the action "to make." Nominalizations make nouns out of other words. Often they take a part of a word, usually a verb, and remake it into a noun. Even when we can reverse the process and restore the action to a verb, should we? When? How often? Everyone uses nominalizations. I welcome them when: 1) They re-name old information Kwong acted well in the play.... Other ACTors in Hong Kong respect his talent, though he does not aspire to RESPECTability. 2) Alternatives strain, obscure, or seem silly and tedious. "The administration....." instead of "Those who administer" "Writer" instead of "One who writes" Nevertheless, I test as many nominalizations as time allows. They can violate the integrity of what anyone wants to say. If you need evidence, review the work of the Committee on Doublespeak at the National Council of Teachers of English. Bureaucrats and others who write for captive audiences like to mask most action. Review ADMINESE.TXT. The writer administers a department. He knew that his captives would read him closely. Presumably he wanted to impress them. He does not impress me. Read my version in REVADMIN.TXT. Which do you prefer? The program does not identify nominalization, but identifies words that MAY evidence it. Several English morphemes often turn verbs into nouns. For example, the morpheme MENT makes nouns out of hundreds of verbs, such as investment [invest+], endorsement [endorse+].... But MENT does not always make a noun. Witness DEPARTMENT, which buries no action, yet the program highlights it. The program is stupid: you should not be. Stay awake. Don't expect the program to find all buried action, or always to highlight words that hide it. Even after the program catches buried action, the user must choose whether to keep it. Version 1.5+ of the program highlights any word which contains one of the following strings: ING, MENT, ION, ANCE, ENCE, NCIES, but excludes DURING and any word that contains the string HING (e.g., sometHING). An earlier version highlighted words that contain ER, OR, at the end, to locate words like actOR, inventER, but sent more false alarms than true ones. If you find some of these matches annoy you more than others, or if you discover strings not included here which would help you edit more effectively, write me. If I incorporate those changes in later versions, I will acknowledge you and send you a free copy. Please share any other reactions. I don't mind the odd mismatch. STYLED saves me an enormous amount of time because it finds most of the cholesterol which previously I used WordStar to seek and replace. The more a writer names agents as subjects and puts action into verbs, the more the program will seem inaccurate. Fewer and fewer of its matches will hide action. That principle affects many computer checks. The better I spell, the more a good spelling program matches with words which I have not misspelled. Text Formats The program assumes text in ASCII (American Standard Code for Information Interchange). If you (V)iew your file first and it appears as you expected to, then you have a text in ASCII and can go on to my next paragraph. If, however, your text adds strange graphics, then your text does not conform to ASCII. Various programs to strip these characters exist. I like to use Gene Plantz' UNWS.EXE, copyrighted in the public domain. I share copies with the copies of STYLED I send out. UNWS means un-WordStar. WordStar uses the non-ASCII characters for its (D) option, the "document" mode from the main menu. You can create ASCII texts in WordStar if you use the "N" (Non-Document) option from the WordStar main menu. I prefer to use the D mode first, to have full editing power, then to strip with UNWS. STYLED reports sentences by number, consecutively, from the beginning to the end of the document. It ignores any line that begins with a period, since many programs format REMARKS thus. It recognizes all initials, plus a few abbreviations--Ms., Mrs., Dr., Prof., Mr.--as long as the period follows with no space between. The program will count the quotation marks (either " or `) at the left of words to compute the number of your quotes, the <,(, or [ to determine the number of parenthetical remarks. It does not detect whether you close these or whether you use them for any other purposes (such as for mathematics). The program will treat as an ellipsis any string of uninterrupted periods. If you put spaces between them, you will distort the punctuation report. The program recognizes spaces as the boundaries of words and tests the left of each word for no more than one punctuation mark, the right of each for no more than 2 punctuation marks. The program will treat any additional punctuation as part of the word: if it ever treats a quotation mark as part of the word, the program will abort. Although the extra quotation mark distorts the format of the report files, you can still view all the data if you type the name of the .REP file from the MS-DOS prompt. The program recognizes -- as a dash, - as a hyphen. It treats words on either side of the dash as separate, on either side of a hyphen as one. The program cannot determine whether a sentence continues after a quotation that includes terminals. It will treat as two sentences an item such as: "John, did you leave?" Mary asked. or When I asked "Could you come?" I was not sure you understood. Reports as Disk Files STYLED stores data in files to list after 3 of the 4 analyses. (It stores no punctuation report.) You can access those files when you edit: LONGWORD.REP [Choice 1 creates.] TRANSIT.REP [Choice 3 creates.] TOBE.REP [Choice 4 creates.] CHOLES.REP [Choice 4 creates.] STYLLIST uses the temporary extension .R$$ for each of these (except TRANSIT.REP, which it ignores). Later it combines the .R$$ files into a file called REPORT and then erases the .R$$ files. Do not use these names for any other files, not even the extension .REP or the extension .R$$. If you plan to access these any of these, rename them before you analyze another text, because each option overwrites the old file when it analyzes another text. You may rename files at the MS-DOS prompt with the syntax: REN oldfile newfile. You may freely erase the four .REP files, but you do not have to erase them, since the program automatically overwrites them. You must reserve space for these files with any new analysis. They vary, and all together could require as much space as the original text. Log onto a disk with plenty of space. A Tip about the Directory Option When the program prompts you for the file to view or analyze, it allows you to see request the directory first. If you enter just the pound sign plus a carriage return, you will see the full current directory of the logged drive. You may specify another drive. You may also specify the MS-DOS "wild cards" --asterisks and question question marks--to limit the clutter. For example: If you respond: #B:\ws\*.TXT the program will show you all files with the extension .TXT on subdirectory called \ws on Drive B:. How to Preserve the Reports Copies of the reports help when you edit. The entire session: To save the session with STYLED as a file, from the MS-DOS prompt type: STYLED >filetosave Replace "filetosave" with the name of the file in which to reserve the session. (Upper/lower case does not matter, but you must use the blank space and the greater-than sign exactly as shown.) If you trap the session in this way, you will slow the program slightly, and you must have enough disk space for the large new file. Some word-processors allow you to see the graphics in the new file; others, such as WordStar, do not. You can view the reports if at the MS-DOS prompt you enter TYPE sessionsaved If you hit Ctrl-P before you enter this command, you will print the entire session. The versions saved in this way will distinguish between high and low intensity. Smaller Portions of a Session: Those who have a dot matrix printer that supports MS-DOS graphics, can use PrtSc to print any screen. This method saves too much data for most purposes, however. Use PrtSc to print the word lists. Refer to those lists as you return to edit your text with your word-processor. Or you may print any of the four report files described above. You may edit them and print them from your word-processor or just TYPE them from the MS-DOS menu. Soft Copy I have explained earlier that the program stores four lists as disk files. When I edit, I use Sidekick to window these reports. How to Automate the Analyses If you use the program regularly, you will need not to retype all responses each time, especially to review only the word lists. You may use Superkey or ProKey or another macro program to monitor and store keystrokes which you use every time. I prefer to hit only one key to complete the full analysis from the MS-DOS prompt. The one key, S, starts a batch file which I call S.BAT. Content of My S.BAT UNWS <LETCODE1 styled <LETSTYL copy *.rep D:\multi\sidek\report del *.rep ws c.c First the .bat file runs UNWS.exe to convert my current WordStar file into ASCII. [See UNWS.DOC on this disk.] I do not have to answer UNWS's prompts from the keyboard, because I have stored my responses in the input file LETCODE1. "<" always tells MS-DOS to read an input file. Content of My Input File LETCODE1 c.c 1 For convenience, I always call my current document `c.c,' I rename it when I leave it for another current document. The first line here answers UNWS's prompt to "Name the file to strip." The second line answers UNWS's next prompt: it names the new file. (I reserve `1' for temporary files. I know that I may freely overwrite these later.) Beware: any new line in an input file sends a carriage return to the program. UNWS requires them where I placed them. Carriage returns not required, cause a program to abort. Note that `11' in LETSTYL below actually represents two different responses you would enter if you ran STYLED from the keyboard. STYLED does not expect a carriage return after the first "1" but does require one after the second. Before you write your input file, run the program it serves and monitor all of you your keystrokes exactly. Content of my input file LETSTYL 11 /////31 /////41 /////e The first number in each line specifies the analysis for STYLED to effect. The second number, "1", specifies my stripped file as the file for STYLED to analyze. I must have a carriage return here to match the one I would have to enter from the keyboard, to tell the program that I have completed the name of the file. The first slash after each carriage return disengages the pauses so that the program will not stop when it analyzes. The other slashes answer the program when it prompts "Hit any key" as it reviews its final lists. For long lists, I may need several slashes. Extra ones will not abort the program: when it it finishes the lists, the program waits indefinitely for a number (1 to 4) or for a V (to view) or for an E (to Exit). LETSTYL gives STYLED the E to Exit after it has requested all 4 disk reports. The third line of S.BAT copies these four reports to the directory where I keep my Sidekick files; the fifth line deletes them from my current directory (normally my WordStar directory). The final line in S.BAT returns me to my current file (c.c.). I use Sidekick to window the STYLED report while I continue to edit. ====== A Note on the versions of STYLED (and STYLLIST): Version 1.0 (May 1986) No STYLLIST. Version 1.1 (June 1986) Simplified the displays, to view more text with each screen. Added more forms of TO BE to check, specifically more contractions. Debugged for codes by word-processors that store ASCII text in lines (records) the full 255-character length. Saved the report files (formerly erased) so that writers can use when they edit. Added STYLLIST.EXE, the version that reports only to the disk file the REPORT of longwords, forms of to be, and a reduced version of nominalizations in which I eliminate most of the overkill. Version 1.11 (June 1986) Further reduced the overkill. Added more examples. Version 1.5 (September 1986) Reduced the size of both STYLED.EXE and STYLLIST.EXE by 18%. In STYLLIST, allowed you to bypass the menu: now you may name the file to analyze when you invoke the program at the MS-DOS prompt: styllist filename Version 1.6 (March 1987) Simplified the menus. Added a shell feature for STYLED. Version 1.7 (April 1987) Added the ability to specify a text to analyze when you evoke STYLED from the command line. Added more prompts to help you when you enter an incorrect file name. ====== I will gladly inform anyone of future updates or send copies. Send only enough money to cover my costs. The same applies to my other each for MS- DOS only. I will mail the next two new substantial updates at my own expense to anyone who documents a contribution of at least $10 to AIDS research. My other programs include: ADDRESS A program to manage addresses. It groups people in up to ten categories. APPLY A program which monitors applications (for jobs, grants, contests....), either for the applicant or the institution. It stores the relevant data and merges it with text files which you create. It allows you to edit your text for fine changes. I may order dossiers and print your letters and envelopes. APPLY reviews instanty any past experience you may have had with the person/institution. CANTONES A program to help you learn to speak Cantonese. An earlier version, "MailMerge Cantonese," won best-article-of-1985 by the Hong Kong Computer Society; reprinted in the Journal of the Chinese Language Teachers' Association, May, 1986. MUSES A program to help writers circulate manuscripts and prepare bibliographies of their publications. See review by Rochelle Ratner in CODA, November/December 1986. Louie Crew Director of the Writing Program Chinese University Shatin, N.T. Hong Kong telephone: 0-6066134 TELEX 50301 CUHK HK CABLE SINOVERSITY ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Appendix: STYLED Practice The program tests what we intuit about texts. For example, the program can chart a writer's "leanness" or "periodicity" with specific, visible evidence. The program helps writers test drafts. If we want a lean style, or a periodic one, the program charts how each new draft nears our goal. The program cannot provide a style; but it can measure several dimensions of style. We must know our own goals before we can use the program effectively. I value force and directness. I like to unpack nouns, to restore action to verbs. I eschew passives and most other forms of TO BE.... Joseph Williams' STYLE (Scott, Foresman, and Company) and Richard Lanham's REVISING PROSE (Scribner's) both preach the style I aspire to practice. I recommend them highly. Both help me winnow. A Sample Session with a Student's Paper I asked my second-year students to write about a book which they had hated. One student concluded: As far as science fiction is concerned, the story is not the least interesting. I can hardly read anything really interesting and mysterious: no fancy about the wonderful world on planets, no extremely mysterious, haunted rumors about the bronze statue, no need to puzzele, no way of amusement. Since the author only reports the plot, the story is quite boring. And I was dissatisfied after reading it. [This paragraph appears as STUDNESE.TXT on the disk.] Notice how she gropes. INTERESTING appears twice, as does MYSTERIOUS. But her text builds. Use STYLED to look at the punctuation, especially in sentence number 2, where she repeats more meaningfully. The final two sentences jab. STYLED charts the pulse. I argue that the student can make a strong paragraph here but needs skill to determine what to keep, what to excise. If she merely cuts the repetition, she could turn turn her pulsating indictment into a cadaver. With the "cholesterol" report, I found 4 forms I wanted to cleanse. Instead of the two forms of TO BE, compare The science fiction here does not engage. I tried to preserve the pulse of the second sentence, as a weighty indictment, but to repeat less, and to put the more vivid action into verbs: No fancy detail, no mystery, no wonder: no rumor haunts planets; the bronze statue neither puzzles nor amuses. In my version, verbs reinforce the jabs at the end. The author only plots. The book bores me. (The punctuation report of her text shows the jabs, but the report of her word-length marks them less clearly than does the report of my word length.) Some readers may prefer the less forceful original. Perhaps I have even changed her meaning, especially with "author only plots." I argue that I have discovered more meaning than she had yet freighted, as I often do when I revise my own texts. These exercises cannot prove one's taste better than another's. Some may prefer to write like the student here, or like the teachers and the administrator below. Which versions would you prefer to read? Which will you more likely remember? A Sample Session with Teachers' Prose In a 1986 issue of Computers and Composition, two teachers reported: In order to meet the second condition, that the analysis should lead to direct improvements in a revision, even the measurements which do correlate with essay quality need to be used with some caution. Research which studies the relationship of revised essays to holistic scores should provide more accurate guidance in this area, but our initial findings suggest some starting points. Reducing the percentage of abstract words and spelling errors can be accomplisehd directly and with measureable effect on essay quality; however, students should not expect that direct changes in word length, sentence length, or readability will improve an essay. The correlation of average word length to quality, for instance, cannot be directly addressed in revision. Substitution of lexical units for the mere sake of length would change the data but not, necessarily, the quality of the revised draft. Perhaps indirectly, through active experience of reading and practice in writing which increase active vocabulary, a student could meaningfully raise his or her average word length from the low mean (4.3) to the high limit of 5.0. The fact that these measurements can be improved only indirectly, suggests the overall importance of scribal fluency. Since sentence length, word length, and readability are all measurements of scribal fluency, the results of this study seem to suggest that more classroom time should be spent on improving scribal behavior than on practicing those discrete grammatical and stylistic elements which do not correlate significantly with essay quality. Whew! STYLLIST cites the following long words in this paragraph: 1 > improvements 1 > measurements 2 > relationship 3 > percentage 3 > accomplisehd 3 > measureable 3 > readability 4 > correlation 5 > Substitution 5 > necessarily 6 > indirectly 6 > experience 6 > vocabulary 6 > meaningfully 7 > measurements 7 > indirectly 7 > importance 8 > readability 8 > measurements 8 > practicing 8 > grammatical 8 > significantly Total word count: 242 Words 10+ chars.: 22 ( 9 %) They also practice what they preach. They value long words and want students to value them. In their last sentence they claim that writers waste time to remove those forms for which holistic graders do not lower the score. Elsewhere they specify nominalizations and forms of TO BE as examples. Again they follow their own advice: STYLLIST cites 6 forms of TO BE in their paragraph: 1 > be 3 > be 4 > be 7 > be 8 > are 8 > be Forms of TO BE: 2 % and the following words as possible nominalizations: 1 > condition 1 > improvements 1 > revision 1 > measurements 1 > caution 2 > relationship 2 > guidance 2 > findings 2 > starting 3 > Reducing 3 > spelling 3 > sentence 4 > correlation 4 > instance 4 > revision 5 > Substitution 6 > experience 6 > reading 6 > writing 6 > meaningfully 7 > measurements 7 > importance 8 > sentence 8 > measurements 8 > improving 8 > practicing 8 > elements Possible buried action: 27 ( 11 %) Total word count: 242 I used these lists when I revised, to express their ideas in my style: A student who expects to improve prose must apply our analysis cautiously. Even when our counts correlate with good grades, further research needs to define precisely how the correlation occurs. For example, we found that grades go up when students spell correctly and use few abstract words; but one should not exchange a word for another as mechanically as one alters spelling. Perhaps students should read more. Then they may meaningfully raise average word length from the low mean (4.3) to the high limit of 5.0. Sentence length, word length, and readability measure scribal fluency. Classes need to improve these rather than practice with those discrete grammatical and stylistic elements which do not correlate significantly with grades. Crusted with nominals, their prose sounds intelligent. But when I unclog its arteries, the teachers do not impress me. They discovered too little: Spelling helps grades. & "Perhaps" [their word!] students should read more. and claim too much: Why waste time with features graders do not notice? Holistically I too give a lower grade to the clearer version, because of its poverty, not because of its clarity. I agree that we need to search the data again to determine how quality correlates with style. I never finish a text, but halt. STYLLIST cites these long words in my version: 1 > cautiously 2 > correlation 3 > mechanically 4 > effectively 5 > meaningfully 6 > readability 7 > grammatical 7 > significantly Total word count: 125 Words 10+ chars.: 8 ( 6 %) I used no forms of the verb TO BE, but STYLLIST reports five words as possible nominalizatons: 2 > correlation 3 > spelling 5 > meaningfully 6 > Sentence 7 > elements Possible buried action: 5 ( 4 %) Total word count: 125 Since I reduced the total word length by almost half, I decided to live with the long words listed. I used no form of TO BE, but did use three nominalizations: CORRELATION and SPELLING name old information, introduced with the verbs CORRELATE and SPELL. I kept the ambiguous word MEANINGFULLY because I did not know what the writers MEAN. (SENTENCE and ELEMENTS hide no action; with them the program misfired.) Explore your own logic. I share mine to fortify you; examine me for weakness. A Sample Session With an Adminstrator's Report Option 4 of the program reveals many places an administrator buried action in nouns or used forms of TO BE: Implicit in what has been suggested above is the fact that the Department needs significantly to increase its expectations and its requirements. Both sections of the Graduate Division now have completed the pioneering phase of their programs. The programs are well established. A satisfactory international recognition for the programs has been achieved. The time now has come to build upon the initial successes of the Department's postgraduate program. During the forthcoming triennium, postgraduate coursework within the Department must be made more complex. Students must be required to operate at more sophisticated--at genuinely international--levels of commitment and skill. Compare: I have implied that the Department needs to expect more, to require more. We have pioneered long enough. Now we must sophisticate our graduate students. During the new triennium, our students must show more skill. They must commit themselves to more complex materials. The faculty must serve international, not parochial standards. When I showed the two versions to a graduate student, he exclaimed, "So much clearer! The second doesn't sound like a report." It does not have to. It IS [sic] one. Yes, I revolt. If you have read this far, maybe you will join me. Just how many more hours will unclear prose abuse? Re-Visers unite. ==============