1STYLE(1)                         User commands                        STYLE(1)
2
3
4

NAME

6       style - analyse surface characteristics of a document
7

SYNOPSIS

9       style [-L language] [-l length] [-r ari] [file...]
10       style [--language language] [--print-long length] [--print-ari ari]
11       [file...]
12       style -h|--help
13       style --version
14

DESCRIPTION

16       Style analyses the surface characteristics of the writing  style  of  a
17       document.   It prints various readability grades, length of words, sen‐
18       tences and paragraphs.  It can further locate  sentences  with  certain
19       characteristics.   If  no  files  are  given, the document is read from
20       standard input.
21
22       Numbers are counted as words with one syllable.  A sentence  is  a  se‐
23       quence  of  words,  that starts with a capitalised word and ends with a
24       full stop, double colon, question mark or exclamation mark.   A  single
25       letter  followed by a dot is considered an abbreviation, so it does not
26       end a sentence.  Various  multi-letter  abbreviations  are  recognized,
27       they  do  not  end  a sentence as well.  A paragraph consists of two or
28       more new line characters.
29
30   Readability grades
31       Style understands cpp(1) #line lines for being able to give precise lo‐
32       cations when printing sentences.
33
34       Kincaid formula
35              The  Kincaid  Formula  was  developed for Navy training manuals,
36              that ranged in difficulty from 5.5 to 16.3.  It is probably best
37              applied  to  technical  documents,  because it is based on adult
38              training manuals rather than school book text.   Dialogs  (often
39              found  in  fictional  texts)  are usually a series of short sen‐
40              tences, which lowers the score.  On the other  hand,  scientific
41              texts with many long scientific terms are rated higher, although
42              they are not necessarily harder to read for people who  are  fa‐
43              miliar with those terms.
44
45              Kincaid = 11.8*syllables/wds+0.39*wds/sentences-15.59
46
47       Automated Readability Index
48              The Automated Readability Index is typically higher than Kincaid
49              and Coleman-Liau, but lower than Flesch.
50
51              ARI = 4.71*chars/wds+0.5*wds/sentences-21.43
52
53       Coleman-Liau Formula
54              The Coleman-Liau Formula usually gives a lower grade  than  Kin‐
55              caid, ARI and Flesch when applied to technical documents.
56
57              Coleman-Liau = 5.88*chars/wds-29.5*sent/wds-15.8
58
59       Flesh reading easy formula
60              The  Flesh  reading  easy formula was developed by Flesh in 1948
61              and it is based on school text covering grade 3 to  12.   It  is
62              wide  spread, especially in the USA, because of good results and
63              simple computation.  The index is usually between 0  (hard)  and
64              100 (easy), standard English documents averages approximately 60
65              to 70.  Applying it to German documents does  not  deliver  good
66              results because of the different language structure.
67
68              Flesch Index = 206.835-84.6*syll/wds-1.015*wds/sent
69
70       Fog Index
71              The  Fog  index was developed by Robert Gunning.  Its value is a
72              school grade.  The ``ideal'' Fog Index level is 7 or 8.  A level
73              above  12 indicates the writing sample is too hard for most peo‐
74              ple to read.  Only use it on texts of at least hundred words  to
75              get  meaningful  results.   Note  that  a correct implementation
76              would not count words of three or more syllables that are proper
77              names,  combinations  of  easy words, or made three syllables by
78              suffixes such as –ed, –es, or –ing.
79
80              Fog Index = 0.4*(wds/sent+100*((wds >= 3 syll)/wds))
81
82       Lix formula
83              The Lix formula developed by Björnsson from Sweden is very  sim‐
84              ple and employs a mapping table as well:
85
86              Lix = wds/sent+100*(wds >= 6 char)/wds
87
88
89              Index         34   38   41   44   48   51    54    57
90              School year      5    6    7    8    9    10    11
91
92       SMOG-Grading
93              The  SMOG-Grading  for English texts was developed by McLaughlin
94              in 1969.  Its result is a school grade.
95
96              SMOG-Grading = square root of (((wds >= 3 syll)/sent)*30) + 3
97
98              It has been adapted to German by Bamberger &  Vanecek  in  1984,
99              who changed the constant +3 to -2.
100
101   Word usage
102       The  word  usage  counts are intended to help identify excessive use of
103       particular parts of speech.
104
105       Verb Phrases
106              The category of verbs labeled "to be" identifies  phrases  using
107              the passive voice.  Use the passive voice sparingly, in favor of
108              more direct verb forms.  The flag -p causes style  to  list  all
109              occurrences of the passive voice.
110
111       The verb category "aux" measures the use of modal auxiliary verbs, such
112       as "can", "could", and "should".  Modal auxiliary verbs modify the mood
113       of a verb.
114
115       Conjunctions
116              The  conjunctions counted by style are coordinating and subordi‐
117              nating.  Coordinating conjunctions join grammatically equal sen‐
118              tence  fragments,  such  as  a noun with a noun, a phrase with a
119              phrase, or a clause to a clause.  Coordinating conjunctions  are
120              "and," "but," "or," "yet," and "nor."
121
122       Subordinating conjunctions connect clauses of unequal status.  A subor‐
123       dinating conjunction links a subordinate clause,  which  is  unable  to
124       stand  alone, to an independent clause.  Examples of subordinating con‐
125       junctions are "because," "although," and "even if."
126
127       Pronouns
128              Pronouns are contextual references to nouns  and  noun  phrases.
129              Documents with few pronouns generally lack cohesiveness and flu‐
130              idity.  Too many pronouns may indicate ambiguity.
131
132       Nominalizations
133              Nominalizations are verbs that are changed to nouns.  Style rec‐
134              ognizes  words  that  end in "ment," "ance," "ence," or "ion" as
135              nominalizations.  Examples are  "endowment,"  "admittance,"  and
136              "nominalization."   Too  much  nominalization  in a document can
137              sound abstract and be difficult  to  understand.   The  flag  -N
138              causes  style  to  list all nominalizations.  The flag -n prints
139              all sentences with either the passive voice or a nominalization.
140

OPTIONS

142       -L language, --language language
143              set the document language (de, en, nl).
144
145       -l length, --print-long length
146              print all sentences longer than length words.
147
148       -r ari, --print-ari ari
149              print all sentences whose readability  index  (ARI)  is  greater
150              than ari.
151
152       -p passive, --print-passive
153              print all sentences phrased in the passive voice.
154
155       -N nominalizations, --print-nom
156              print all sentences containing nominalizations.
157
158       -n nominalizations-passive, --print-nom-passive
159              print  all sentences  phrased in the passive voice or containing
160              nominalizations.
161
162       -h, --help
163              Print a short usage message.
164
165       --version
166              Print the version.
167

ERRORS

169       On usage errors, 1 is returned.  Termination caused by lack  of  memory
170       is signalled by exit code 2.
171

ENVIRONMENT

173       LC_MESSAGES=de|en|nl
174              specifies  the  default document language.  The default language
175              is en.
176
177       LC_CTYPE=iso-8859-1
178              specifies the document character set.  The default character set
179              is ASCII.
180

AUTHOR

182       This  program  is  GNU  software,  copyright  1997–2007  Michael Haardt
183       <michael@moria.de>.
184
185       It contains contributions by Jason  Petrone  <jpetrone@acm.org>,  Uschi
186       Stegemeier <uschi@morwain.de> and Hans Lodder.
187
188       This program is free software; you can redistribute it and/or modify it
189       under the terms of the GNU General Public License as published  by  the
190       Free  Software Foundation; either version 3 of the License, or (at your
191       option) any later version.
192
193       This program is distributed in the hope that it  will  be  useful,  but
194       WITHOUT  ANY  WARRANTY;  without  even  the  implied  warranty  of MER‐
195       CHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU  General
196       Public License for more details.
197
198       You should have received a copy of the GNU General Public License along
199       with this program.  If not, write  to  the  Free  Software  Foundation,
200       Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
201

HISTORY

203       There was a style command on old UNIX systems, which is now part of the
204       AT&T DWB package.  The original version was bound to roff by  enforcing
205       a call to deroff.
206

SEE ALSO

208       deroff(1), diction(1)
209
210       Cherry,  L.L.;  Vesterman, W.: Writing Tools—The STYLE and DICTION pro‐
211       grams, Computer Science Technical Report 91, Bell Laboratories,  Murray
212       Hill,  N.J. (1981), republished as part of the 4.4BSD User's Supplemen‐
213       tary Documents by O'Reilly.
214
215       Coleman, M. and Liau,T.L. (1975). 'A computer readability  formula  de‐
216       signed  for  machine  scoring',  Journal  of Applied Psychology, 60(2),
217       283-284.
218
219
220
221GNU                            January 29, 2014                       STYLE(1)
Impressum