Relation sumo#causes (sumo#process,sumo#process)  the process in the 1st argument brings about the process in the 2nd argument
  type:  pm#binary_predicate_type  the class of predicates relating two items - its valence is two
  type:  pm#asymmetric_relation_type  an antisymmetric and irreflexive relation
  supertype:  pm#cause (pm#situation,pm#situation)  see also rst#cause
     supertype:  pm#before_situation (pm#situation,pm#situation)  in WebKB, do not use these relations, use their inverses
        supertype:  pm#relation_from_situation_to_situation (pm#situation,pm#situation)
           supertype:  pm#relation_from_situation (pm#situation,*)
              supertype:  pm#relation_from/to_thing_of_common_kind (*)  this type permits to categorize relations according to their signatures and hence offers (i) a concise way to set essential exclusion relations, and (ii) a systematic and easy-to-follow categorization
                 >part of:  pm#relation__related_thing__relatedthing___related_with  type for any relation (unary, binary, ..., *-ary) and instance of pm#relation_type
           supertype:  pm#relation_to_situation (*,pm#situation)  in WebKB, prefer using relations "from" a situation
              supertype:  pm#relation_from/to_thing_of_common_kind (*)  this type permits to categorize relations according to their signatures and hence offers (i) a concise way to set essential exclusion relations, and (ii) a systematic and easy-to-follow categorization
           supertype:  pm#when_relation (?,?)
              supertype:  pm#wh-/how_relation (*)  this type permits to categorize relations according to the usual who/what/why/where/when/how questions ; this is a traditional but very subjective and ineffective way of categorizing relations 
                 >part of:  pm#relation__related_thing__relatedthing___related_with  type for any relation (unary, binary, ..., *-ary) and instance of pm#relation_type
        supertype:  nsm#after (?,?)
           supertype:  pm#binary_relation (?,?)  in WebKB, most relation types are binary and some have a variable number of arguments (as in KIF), hence this type is currently only specialized by types that I do not want to see as direct subtypes of pm#relation
              supertype:  pm#relation_with_particular_property (*)  this rather fuzzy type permits to group categorization schemes less common than those covered by the previous sibling categories
                 >part of:  pm#relation__related_thing__relatedthing___related_with  type for any relation (unary, binary, ..., *-ary) and instance of pm#relation_type
     supertype:  pm#why_relation__whyrelation (*)
        supertype:  pm#wh-/how_relation (*)  this type permits to categorize relations according to the usual who/what/why/where/when/how questions ; this is a traditional but very subjective and ineffective way of categorizing relations 
  supertype:  pm#asymmetric_relation (?,?)  this category only serves structuration purposes: it is instance of pm#asymmetric_relation_type which is not instance of pm#class_of_inheritable_relation_type
     supertype:  pm#irreflexive_relation__irreflexiverelation (?,?)  this category only serves structuration purposes: it is instance of pm#irreflexive_relation_type which is not instance of pm#class_of_inheritable_relation_type
        supertype:  pm#binary_relation_with_particular_mathematical_property (?,?)
           supertype:  pm#relation_with_particular_mathematical_property (*)
              supertype:  pm#relation_with_particular_property (*)  this rather fuzzy type permits to group categorization schemes less common than those covered by the previous sibling categories
     supertype:  pm#antisymmetric_relation__antisymmetricrelation (?,?)  this category only serves structuration purposes: it is instance of pm#antisymmetric_relation_type which is not instance of pm#class_of_inheritable_relation_type
        supertype:  pm#binary_relation_with_particular_mathematical_property (?,?)