Rules for writing academic papers and checking them using LTex language server and LanguageTool
This is an incomplete, highly opinionated list of rules I use in LanguageTool for checking my papers. Feel free to use it, modify it, etc. The rules are geared towards programming-language papers.
The rules are written here as a literate org-mode file, but you’re welcome to work with the raw XML file rules.xml, which is also included in this repo.
Many of the rules are taken from The CAPRA Style Guide.
Pull requests and bug reports are always welcome!
Download the following patch to langtools.patch.
--- /tmp/b 2022-06-06 16:37:42.000000000 -0700
+++ /tmp/a 2022-06-06 17:00:59.000000000 -0700
@@ -29,6 +29,7 @@
-->
<!DOCTYPE rules [
+ <!ENTITY UserRules SYSTEM "file:///home/username/.language-tool-rules.xml">
<!ENTITY apostrophe "['’`´‘ʻ]">
<!ENTITY quote '["“”„]'>
<!ENTITY nbsp " "> <!-- no-break space -->
@@ -598,6 +599,10 @@
</phrase>
</phrases>
+ <category id="CUSTOM_USER_RULES" name="Custom User Rules" default="on">
+ &UserRules;
+ </category>
+
<!--
<category id="APRIL" name="April 1st">
<rule id="APRIL_1ST" name="good advice">
Run the following commands
sed -i -e "s|home/username|$HOME|" langtools.patch patch -p1 $PATH_TO_GRAMMAR langtools.patch
You’ll need to first set PATH_TO_GRAMMAR to something like $(ls /usr/local/Cellar/languagetool/*/libexec/org/languagetool/rules/en/grammar.xml) on macOS, or /usr/share/languagetool/org/languagetool/rules/en/grammar.xml on Linux.
This will automatically install the necessary lines into the right files, and LanguageTool will load the rules from $HOME/.language-tool-rules.xml for all English language grammars.
Install rules.xml to that path.
If you want to also install your own personal dictionary, you can use
ln -sf ~/.aspell.en.pws $PATH_TO_DICTIONARY
On macOS, PATH_TO_DICTIONARY for all English language dictionaries is
$(ls /usr/local/Cellar/languagetool/*/libexec/org/languagetool/resource/en/hunspell/spelling_custom.txt).
On linux, try
$(ls /usr/share/languagetool/org/languagetool/resource/en/hunspell/spelling_custom.txt).
This will load your aspell dictionary as a custom dictionary automatically.
- Run LanguageTool as a server, following this guide. This lets you modify your grammar rules.
- Add the rules to your grammar. Add a line that looks something like
this under
<!DOCTYPE rules [in LanguageTool’sgrammar.xml:<!ENTITY UserRules SYSTEM "file:///path/to/your/rules.xml">Then in some enabled rule-set, add
&UserRules;. For example, I have:<rules lang="en-GB" xsi:noNamespaceSchemaLocation="../../../../../../../../../../languagetool-core/src/main/resources/org/languagetool/rules/rules.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <category id="SEMANTICS" name="Semantics" type="inconsistency"> <!-- English rule, 2022-02-01 --> &UserRules;The full guide is here
- Configure LTex to use your local server. Make sure to restart the LanguageTool server each time you change the rules.
<!-- English rule, 2022-02-16 -->
<rule id="NO_CONTRACTIONS" name="No Contractions">
<pattern>
<token regexp='yes'>'m|'re|'ve|'d|'ll|'t</token>
</pattern>
<message>Do not use contractions in formal writing</message>
<short>No contractions</short>
<example correction=''>let<marker>'s</marker></example>
<example>let us</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="S_CONTACTION" name="S_CONTACTION">
<pattern>
<token postag='PRP|VBZ' postag_regexp='yes'>'s</token>
</pattern>
<message>In formal writing, 's should only be used for possessives, not as a contraction for "is".</message>
<short>Don't use 's as a contraction.</short>
<example correction=''>That<marker>'s</marker> the problem, so let's look at the car's engine. Let's look at the car's engine. That's the problem with the car's engine.</example>
<example>That is the problem, so let us look at the car's engine.</example>
<example correction=''>Let's look at the car's engine.</example>
<example>Let us look at the car's engine.</example>
<example correction=''>That's the problem with the car's engine.</example>
<example>That is the problem with the car's engine.</example>
</rule>
This is probably the most controversial rule. Even if there’s a single author, I refer to them as “they” instead of “he” or “she”. Realistically, the only place I can imagine using “he” or “she” in a paper is when directly quoting someone else.
<!-- English rule, 2022-02-16 -->
<rule id="NO_GENDERED_PRONOUNS" name="no gendered pronouns">
<pattern>
<token inflected='yes' regexp='yes'>he|she|his|her</token>
</pattern>
<message>There are very few cases where gendered pronouns are appropriate in a formal paper. You can never go wrong with "they"</message>
<short>"they"</short>
<example correction=''><marker>He</marker> proves this She proves this It is her proof It is his proof The proof is hers The proof is his</example>
<example>They prove this</example>
<example correction=''>She proves this</example>
<example>They prove this</example>
<example correction=''>It is her proof</example>
<example>It is their proof</example>
<example correction=''>It is his proof</example>
<example>It is their proof</example>
<example correction=''>The proof is hers</example>
<example>The proof is theirs</example>
<example correction=''>The proof is his</example>
<example>The proof is theirs</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="WE_HAVE" name="We have">
<pattern>
<token>we</token>
<token postag='VBP|VBN|VB' postag_regexp='yes'>have</token>
<token postag='VBN' postag_regexp='yes' negate_pos='yes'></token>
</pattern>
<message>Joey uses "we have" too much in his writing</message>
<short>Don't say "we have"</short>
<example correction=''><marker>We have a</marker> problem.</example>
<example>There is a problem.</example>
<example>We have seen it.</example>
</rule>
I always have a macro \lang that I use to refer to the language I’m developing, so I shouldn’t say “our language” a bunch
<rule id="OUR_LANGUAGE" name="OUR_LANGUAGE">
<pattern>
<token>our</token>
<token>language</token>
</pattern>
<message>
Joey writes
<suggestion>our language</suggestion>
too much
</message>
<example correction=''>
<marker>our language</marker>
</example>
<example>\lang</example>
</rule><!-- English rule, 2022-02-02 -->
<rule id="RUNTIME" name="RUNTIME">
<pattern>
<token>runtime</token>
</pattern>
<message>Unless you are talking about how long a program takes to run, use "run-time" as an adjective or run time as a noun.</message>
<example correction=''>
<marker>runtime</marker>
</example>
<example>run-time or run time</example>
</rule><!-- English rule, 2022-02-16, taken from the CAPRA guide -->
<rule id="RUN_TIME_ADJECTIVE" name="Run time adjective">
<pattern>
<token>run</token>
<token>time</token>
<token postag='N.*' postag_regexp='yes'></token>
</pattern>
<message>Use run-time as as a description of a noun.</message>
<short>Use run-time for adjective</short>
<example correction=''>There was a <marker>run time error</marker>.</example>
<example>There was a run-time error.</example>
<example>There was an error at run time.</example>
</rule>
We need a couple rules to catch all these cases, since the chunker is finicky
<!-- English rule, 2022-02-16, CAPRA style guide -->
<rule id="RUNTIME_AS_NOUN" name="Run-time as noun">
<pattern>
<token chunk_re='E-NP.*'>run-time</token>
</pattern>
<message>Use "run time" for the noun form</message>
<short>Noun form is "run time"</short>
<example correction=''>There was an error at <marker>run-time</marker> in the program This will cause problems at run-time.</example>
<example>There was an error at run time in the program</example>
<example>There was a run-time error in the program</example>
<example>This is a problem during run-time checks</example>
<example correction=''>This will cause problems at run-time.</example>
<example>This will cause problems during run-time analysis.</example>
</rule>
<!-- English rule, 2022-02-16, taken from CAPRA guide -->
<rule id="RUNTIME_SENTENCE_END" name="Run-time sentence end">
<pattern>
<token postag='SENT_END'>run-time</token>
</pattern>
<message>Use "run time" for the noun</message>
<short>Noun form is "run time"</short>
<example correction=''>There was an error at <marker>run-time</marker></example>
<example>There was an error at run time</example>
<example>There was a run-time error</example>
</rule>
<!-- English rule, 2022-02-02 -->
<rule id="TYPECHECK" name="Typecheck">
<pattern>
<token regexp='yes'>type-?check(er|ers|ing|s|ed)?</token>
</pattern>
<message>Use "type check" for consistency</message>
<example correction=''>
<marker>typecheck</marker>
typechecking type-check type-checking type-checks type-checker typechecks typechecker
</example>
<example>type check</example>
<example correction=''>typechecking</example>
<example>type checking</example>
<example correction=''>type-check</example>
<example correction=''>type-checking</example>
<example correction=''>type-checks</example>
<example correction=''>type-checker</example>
<example correction=''>typechecks</example>
<example correction=''>typechecker</example>
</rule><!-- English rule, 2022-02-02 -->
<rule id="CASTCALCULUS" name="CASTCALCULUS">
<pattern>
<token>cast</token>
<token>calculus</token>
</pattern>
<message>"cast calculus" should have a hyphen</message>
<example correction=''>
<marker>cast calculus</marker>
</example>
<example>cast-calculus</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="TRADEOFF" name="tradeoff">
<pattern>
<token inflected='yes'>tradeoff</token>
</pattern>
<message>use "trade-off" for consistency</message>
<short>"trade-off"</short>
<example correction=''><marker>tradeoff</marker> tradeoffs</example>
<example>trade-off</example>
<example correction=''>tradeoffs</example>
<example>trade-offs</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="JUDGEMENT" name="judgement">
<pattern>
<token inflected='yes' regexp='yes'>judgement(al)?</token>
</pattern>
<message>For consistency, use "judgment"</message>
<short>"judgment"</short>
<example correction=''><marker>judgement</marker> judgements judgemental</example>
<example>judgment</example>
<example correction=''>judgements</example>
<example>judgments</example>
<example correction=''>judgemental</example>
<example>judgmental</example>
</rule>
<!-- English rule, 2022-02-23 -->
<rule id="DEPENDENT_TYPE" name="Dependent type">
<pattern>
<token regexp='yes'>dependent(ly)?-type[sd]?</token>
</pattern>
<message>No hyphen in "dependent type"</message>
<short>No hyphen</short>
<example correction=''>The constructor builds a <marker>dependent-type</marker> The language supports dependent-types This is hard in dependently-typed programming languages</example>
<example>The constructor builds a dependent type</example>
<example correction=''>The language supports dependent-types</example>
<example>The language supports dependent types</example>
<example correction=''>This is hard in dependently-typed programming languages</example>
<example>This is hard in dependently typed programming languages</example>
</rule><!-- English rule, 2022-02-23 -->
<rule id="DEPENDENT_LANGUAGE" name="Dependent language">
<pattern>
<token>dependent</token>
<token>language</token>
</pattern>
<message>Use the full "dependently typed"</message>
<short>"dependently typed language"</short>
<example correction=''>This is a <marker>dependent language</marker></example>
<example>This is a dependently typed language</example>
</rule><!-- English rule, 2022-02-16, from CAPRA -->
<rule id="THIS_AS_SUBJECT" name="This as subject">
<pattern>
<token inflected='yes' regexp='yes' postag='DT'>this|these</token>
<token postag='VB.*' postag_regexp='yes' chunk_re='.*VP.*'></token>
</pattern>
<message>Be specific: don't use "this" as the subject of a sentence</message>
<short>Don't use this as the subject of a sentence</short>
<example correction=''><marker>This is</marker> problematic. These are problematic These are problematic</example>
<example>The problem is problematic.</example>
<example>This problem is problematic.</example>
<example correction=''>These are problematic</example>
<example>These problems are problematic</example>
<example>They are problematic</example>
<example>In this section we see the problem</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="WE_CAN_SEE_THAT" name="we can see that">
<pattern>
<token>we</token>
<token min='0'>can</token>
<token>see</token>
<token>that</token>
</pattern>
<message>"we can see that" adds nothing and eats your precious page budget</message>
<short>redundant</short>
<example correction=''><marker>we can see that</marker> this happens we see that this happens</example>
<example>this happens</example>
<example correction=''>we see that this happens</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="CLEARLY_OBVIOUSLY" name="clearly obviously">
<pattern>
<token regexp='yes'>clearly|obviously</token>
</pattern>
<message>Just eliminate these words, don't assume what is easy to the reader</message>
<short>redundant</short>
<example correction=''><marker>Clearly</marker> this is true Obviously this is true</example>
<example>this is true</example>
<example correction=''>Obviously this is true</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="EASY_TO_SEE" name="easy to see">
<pattern>
<token>it</token>
<token>is</token>
<token>easy</token>
<token>to</token>
<token>see</token>
</pattern>
<message>Don<suggestion>t assume what</suggestion>s easy to the reader</message>
<short>redundant</short>
<example correction=''><marker>it is easy to see</marker> that this is true it is easy to see this is true</example>
<example>this is true</example>
<example correction=''>it is easy to see this is true</example>
<example>this is true</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="SIMPLY" name="simply">
<pattern>
<token>simply</token>
<token negate='yes'>typed</token>
</pattern>
<message>You should only use simply to refer to the STLC</message>
<short>redundant</short>
<example correction=''>the result is <marker>simply three</marker></example>
<example>the result is three</example>
<example>this is the simply typed lambda calculus</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="NOTE_THAT" name="Note that">
<pattern>
<token regexp='yes'>note|notice</token>
<token>that</token>
</pattern>
<message>"Note that" or "notice that" adds nothing</message>
<short>Don't say "note that"</short>
<example correction=''><marker>note that</marker> this happens notice that this happens</example>
<example>this happens</example>
<example correction=''>notice that this happens</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="SEE_HOW" name="see how">
<pattern>
<token>see</token>
<token>how</token>
</pattern>
<message>"see how" is redundant and adds nothing</message>
<short>Don't use "see how"</short>
<example correction=''><marker>see how</marker> it happens</example>
<example>it happens</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="IT_IS_WORTH_NOTING_THAT" name="it is worth noting that">
<pattern>
<token>it</token>
<token>is</token>
<token>worth</token>
<token>noting</token>
<token>that</token>
</pattern>
<message>"it is worth nothing that" takes up your valuable space budget and adds nothing</message>
<short>redundant</short>
<example correction=''><marker>it is worth noting that</marker> this happens</example>
<example>this happens</example>
</rule>
<!-- English rule, 2022-02-16, taken from CAPRA guide -->
<rule id="LONG_LIST_NUMBERS" name="Long List Numbers">
<pattern>
<token regexp='yes'>firstly|secondly|thirdly</token>
<token>,</token>
</pattern>
<message>Use (1) and (2) instead of firstly, secondly, etc.</message>
<short>Don's use firstly, secondly, thirdly</short>
<example correction=''>Firstly, this is the problem.</example>
<example>(1)</example>
<example correction=''>Secondly,</example>
<example>(2)</example>
<example correction=''>Thirdly,</example>
<example>(3)</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="ALLOWS_TO" name="Allows to">
<pattern>
<token inflected='yes'>allow</token>
<token chunk_re='B-NP.*'></token>
<token min='0' chunk_re='I-NP.*'></token>
<token min='0' chunk_re='E-NP.*'></token>
<token>to</token>
</pattern>
<message>Use "lets" instead</message>
<short>Use "lets"</short>
<example correction=''>This <marker>allows the program to</marker> run</example>
<example>This lets the program run</example>
<example correction=''>The check allows our gradual language to execute safely.</example>
<example>The check lets our gradual language execute safely.</example>
<example correction=''>The check allows it to run safely.</example>
<example>The check lets it run safely.</example>
<example correction=''>These allow the program to run</example>
<example>These let the program run</example>
<example correction=''>These check allow our gradual language to execute safely.</example>
<example>These checks let our gradual language execute safely.</example>
<example correction=''>The checks allows them to run safely.</example>
<example>The check let them run safely.</example>
<example correction=''>This allows all the programs to run</example>
<example>This lets all the programs run</example>
</rule>
<!-- English rule, 2022-02-16 -->
<rule id="GIVE_ABILITY_TO" name="Gives ability to">
<pattern>
<token inflected='yes'>give</token>
<token chunk_re='B-NP.*'></token>
<token min='0' chunk_re='I-NP.*'></token>
<token min='0' chunk_re='E-NP.*'></token>
<token>the</token>
<token>ability</token>
<token>to</token>
</pattern>
<message>Use "lets" instead</message>
<short>Use "lets"</short>
<example correction=''>This gives the program the ability to run</example>
<example>This lets the program language run</example>
<example correction=''>The check gives our gradual language the ability to execute safely.</example>
<example>The check lets our gradual language execute safely.</example>
<example correction=''>The check gives it the ability to run safely.</example>
<example>The check lets it run safely.</example>
<!-- <example correction=''>These give all the programs the ability to run</example> -->
<!-- <example>These let all the programs run</example> -->
<example correction=''>These give them the ability to run</example>
<example>This lets them ruin</example>
<example correction=''>These give it the ability to run</example>
<example>These let it run</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="IN_ORDER_TO" name="In order to">
<pattern>
<token>in</token>
<token>order</token>
<token>to</token>
</pattern>
<message>Use the simpler "to"</message>
<short>Use "to"</short>
<example correction=''><marker>in order to</marker></example>
<example>to</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="SO_AS_TO" name="So as to">
<pattern>
<token>so</token>
<token>as</token>
<token>to</token>
</pattern>
<message>Use the simpler "to"</message>
<short>"to"</short>
<example correction=''><marker>so as to</marker></example>
<example>to</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="IS_BUILT_ON" name="Is built on">
<pattern>
<token inflected='yes'>be</token>
<token>built</token>
<token>on</token>
</pattern>
<message>Use "builds on" instead of "is built on" when describing systems/languages</message>
<short>"builds on"</short>
<example correction=''><marker>is built on</marker> are built on was built on were built on</example>
<example>builds on</example>
<example correction=''>are built on</example>
<example>build on</example>
<example correction=''>was built on</example>
<example>built on</example>
<example correction=''>were built on</example>
<example>built on</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="HAS_POTENTIAL" name="has potential">
<pattern>
<token inflected='yes'>have</token>
<token>the</token>
<token>potential</token>
<token>to</token>
</pattern>
<message>Use the simpler "could"</message>
<short>"could"</short>
<example correction=''><marker>has the potential to</marker> have the potential to</example>
<example>could</example>
<example correction=''>have the potential to</example>
<example>could</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="SUFFICIENT_AMOUNT" name="sufficient amount">
<pattern>
<token>a</token>
<token>sufficient</token>
<token>amount</token>
<token>of</token>
</pattern>
<message>Use the simpler "enough"</message>
<short>"enough"</short>
<example correction=''><marker>a sufficient amount of</marker></example>
<example>enough</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="UTILIZE" name="utilize">
<pattern>
<token inflected='yes'>utilize</token>
</pattern>
<message>Use the simpler "use"</message>
<short>"use"</short>
<example correction=''><marker>utilize</marker> utilized utilizes</example>
<example>use</example>
<example correction=''>utilized</example>
<example>used</example>
<example correction=''>utilizes</example>
<example>uses</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA -->
<rule id="MAKE_USE_OF" name="make use of">
<pattern>
<token inflected='yes'>make</token>
<token>use</token>
<token>of</token>
</pattern>
<message>Use the simpler "use"</message>
<short>"use"</short>
<example correction=''><marker>make use of</marker> made use of makes use of making use of</example>
<example>use</example>
<example correction=''>made use of</example>
<example>used</example>
<example correction=''>makes use of</example>
<example>uses</example>
<example correction=''>making use of</example>
<example>using</example>
</rule>
<!-- English rule, 2022-02-16, CAPRA-->
<rule id="WHICH_MEANS_THAT" name="which means that">
<pattern>
<token>which</token>
<token>means</token>
<token>that</token>
</pattern>
<message>Use "so" instead</message>
<short>Use "so" instead</short>
<example correction='so'><marker>which means that</marker></example>
<example>so</example>
</rule>
<!-- English rule, 2022-02-23 -->
<rule id="DESIDERATA" name="Desiderata">
<pattern>
<token inflected='yes'>desideratum</token>
</pattern>
<message>Use "goals", "criteria" or something similar</message>
<short>Overused</short>
<example correction=''><marker>desideratum</marker> desiderata</example>
<example>goal</example>
<example correction=''>desiderata</example>
<example>goals</example>
</rule><!-- English rule, 2022-02-23 -->
<rule id="THERE_IS_ARE" name="There is are">
<pattern>
<token>there</token>
<token regexp='yes'>is|are|were</token>
</pattern>
<message>Might be correct, but often it's stronger to use a more direct verb.</message>
<short>Use a direct verb?</short>
<example correction=''><marker>There is</marker> a cat There are some cats</example>
<example>A cat is sitting there</example>
<example correction=''>There are some cats</example>
<example>Some cats are sitting there</example>
</rule>Not a hard rule, but 99% of the time I use the future tense I want the present tense
<!-- English rule, 2022-02-23 -->
<rule id="WILL_FUTURE_TENSE" name="Will (future tense)">
<pattern>
<token postag='MD'>will</token>
</pattern>
<message>In academic writing, you very seldom want the future tense.</message>
<short>Use present tense</short>
<example correction=''>The term <marker>will</marker> reduce It will be a problem It will be a problem Will the term reduce?</example>
<example>The term reduces</example>
<example>I have a strong will.</example>
<example>I tried to will them to arrive.</example>
<example correction=''>It will be a problem</example>
<example>It is a problem</example>
<example correction=''>Will the term reduce?</example>
<example>Does the term reduce?</example>
</rule>
