Addenda to features
Only restrictive relative clauses should be taken into account. Non-restrictive relative clauses may behave differently, but since they are not typical, they should be disregarded.
By adverb we mean any fixed expression for frequency that is not clearly an argument noun phrase or a verb (phrase). Thus, we include not only single word expressions like often and always, but also fixed phrasal expressions like all the time. However, freely created phrasal expressions (e.g. with high frequency) do not count as adverbs.
Only ordinary, non-echo questions should be taken into account. Echo questions may behave differently. Even in English, echo questions allow in-situ placement of the interrogative phrase ("You bought what?").
The criterion of elliptical answers that is given here as a sufficient criterion for classifying a pronoun as an independent pronoun is not the best criterion, because in some languages independent personal pronouns are not really felicitous in elliptical answers. While question-answer sequences such as English "Who did it? He" and French "Qui est arrivé hier? Lui" are not impossible, it seems that answers with an additional verb ("He did", "C'est lui", or similar) are more natural in English and French. However, everyone agrees that English "he" and French "lui" are independent personal pronouns. Thus, the criterion should be formulated in weaker terms: If a personal pronoun can be focused, it is an independent personal pronoun. If it cannot be focused, it is not an independent personal pronoun. Occurrence in an elliptical verbless answer is one focus context, but other focus contexts are contrastive situations of various sorts (e.g. "I will go, not you"), and occurrence in clefts ("It's you who will go").
There was an error in the annotation text of the March 2008 questionnaire:
Value 5 should be "Adpositional marking, preceding the possessum", and value 6 should analogically be "Adpositional marking, following the possessum" (instead of "possessor").
Please note that if the progressive marker is an auxiliary (e.g. Tok Pisin i stap), this should count as "particle" (as opposed to "affix"). And if there are two different progressive markers in the language, please count the more grammaticalized one, which in most cases will also be the one that is more frequently used. Finally, if the progressive marker consists of two parts (e.g. English is work-ing) that exhibit different link tightness, please count the one that is more tightly linked.
There was an error in the annotation text. It should be:
Here we are asking about object relative clauses, i.e. relative clauses in which the relativized element plays the role of object (and not subject as it was first stated).
The multiple choice note says erroneously "You may select several values, unless you select value 4". It should be "You may select several values, unless you select value 3".