[CWB] expanding positional attributes

[luigi.talamo at libero.it]

Dear list,
I was wondering whether is possible to make a better use of the tagset 
provided in an annotation.
For example, if I have two separate tags for a finite verb (VER:fin) and 
infinitive verb (VER:infi), as in La Repubblica corpus, it is possible to set a 
query as following:

[lemma="disegnare" & pos="VER.*"]

in order to catch both the finite and infinitive form of the verb disegnare.

We're making use of wildcards (and `colon-ed' values for representing 
morphosemantic features, as introduced by Marco Baroni for this tagset) as a 
workaround for a non-expandable, mono-dimensional value of the positional 
attributes: it would be nice to have a nesting-capable value of these 
attributes, as (AFAIK) in structural attributes. Here I'm proposing a sort of 
hierarchical structure for a token-level annotation, which could turn useful to 
people making, for instance, a morphological annotation, as in the example 
above.
What do you think?

Best regards,

Luigi