Incertae sedis is necessary as a link type. It would be better to implement as a modifier (attribute) of hierarchical relationshipf. Incertae seids also can be implemented as a status of a taxon concept, but it introduces an assumption that placement of a taxon concept in hierachy affects on the concept itself. It is model dependent because there can be models defining a concept by its members but not with its higher taxa, so it shouldn't be a status of taxon cencept in the exchange schema. -- JamesYtow 12-Sep.-2004

Is a reasonable translation of 'incertae sedis' - 'possibly included in'?
I think it would be overly complex to incorporate modifiers for relationships - though clearly some people might want to record (or represent historical) informatin which is uncertain.
We already have the relationship 'doubtful' (i.e. 'uncertain relationship between C1 and C2'): 'incertae sedis' ('possibly included in') seems a bit more specific than this.
would the inverse relationship 'possibly includes' also be required??? -- TrevorPaterson 130904

No, 'included in, but not confident its placement inside'; it points to at least two or more higher taxon as higher taxon. For example, "I'm not confident in family of this genus, but confident in its order". It is not doubtful at all by skipping at least one rank. Only inconfident in its complete placement. I don't think we need 'possibly includes' because I'm seeing positive side of incertae sedics: whereever its precise place inside, it must belong to this one. -- James Ytow 13-Sep.-2004

Therefore I can't see any need for a new relationship - we can record concept 1 (rank x) 'is included in' concept 2 (rank z). If nothing is recorded about which concept of rank y that concept 1 is included in, that can be interpreted as 'incertae sedis'. 'Incertae sedis' seems to be a statement that the data is incomplete, so is not necessary if we record what data there is - the incompletion is self-evident.... -- TrevorPaterson 130904

How can the TCS pass a loop-back test with a database recording incertae sedis? Incertae sedis does not mean incompleteness of data; if author of a publication stated as incertae sedis, it should be recorded as so but it is complete data. -- JamesYtow 13-Sep.-2004

As we understand it - an example of incertae sedis is 'I classify this genus as belonging to order x, but dont know what family it is in.' I am not sure why this cannot be captured by recording that the genus is included in the order - why do we specifically need to record the uncertainty about intervening ranks - does the fact that nothing is recorded not capture this? I think that the rules of nomenclature do make assignment to some ranks mandatory, is it just these ranks that you would mark 'incertae sedis', as otherwise many classifications would be full of 'i.s.' notes for the intervening ranks that they didnt use. I dont know whether any databases represent incertae sedis, and if people did want to record uncertainties in the future, the various relationships available for TCS should allow this in several ways: 'included in', 'not included in, 'doubtful' which, when used in combination may be a little more expressive than just incertae sedis. If data does exist in databases with 'i.s.' would we represent this as a type of relationship to the confident supertaxon (order), or as multiple relationships to the uncertain supertaxa (famillies). I think this is a difficult question and neither of these mechanisms would necessarily add any information to the option of not explicitly recording incertae sedis. --TrevorPaterson 130904

Incertae sedis is a fact that the author of the publication explicitly stated that assignment of the taxon is (or taxa are) to a higher taxon is unclear. If we map the fact to relationships can be used for other cases, e.g. 'doubtful- included in', we can't distinguish incertae sedis from cases where the author assigned the taxon/taxa with doubt to a higher taxon. Skipped rank does not help because there can be cases where rank is skipped without incertae sedis statement. Note that ICZN does not control taxa higher than family. -- James Ytow 13-Sep.-2004