Appendix B.3 Custom sample system types

The following table reports the summarized results from the custom sample. ‘ID’ = a unique identifier for each system (there may be multiple systems per language/role). ‘Form’ = formal system type, which for conditional indexing systems is symmetrical or asymmetrical, and for unconditional indexing systems is ‘full’, or ‘none’ where no indexing exists. Systems are ‘rounded up’ to asymmetrical if a zero index occurs anywhere in the system, though some internal splits may be symmetrical; these are not indicated. In the ‘Features’ column, ‘pers’ = person, ‘num’ = number, and ‘other’ = mostly gender or noun class. The feature is recorded if any form in a paradigm expresses it; hence, paradigms that only have number distinctions for some person values are still recorded as ‘pers/num’ systems. The occurrence of the summarized factors are indicated for each system: ‘ref’ = all referential factors; ‘event’ = all event-semantic factors; ‘co-occ’ = co-occurrence with conominal; ‘other’ = any other condition. The table is sorted by formal type and role.