In GRAID, noun phrase subconstituents are marked with <ln> or
<rn> for elements respectively to the left or right of the phrase
head. As, however, GRAID functions are only marked on the phrase head (with
the exception of the <:poss> function), the function of the phrase can
only directly be read off its head, not any of of its other constituents. The
mc_subfunc function finds the heads of NPs and copies their GRAID
functions to potentially referential subconstituents that - crucially - are
continguous with the phrase head in linear order. That is, if a NP is broken
up by any elements not belonging to it, this function cannot associate any of
its non-contiguous subconstituents with it. Note also that only a subset of
subconstituent glosses are assigned functions, specifically those containing
the forms in the subforms argument. By default, these are <np>
for lexical nouns (e.g <ln_np> and <rn_np>, but also
<ln_pn_np> etc.), <pro> for free pronouns, <refl> for
reflexives, and <indef> for indefinites. This list of forms can be
replaced or expanded as necessary, or left empty to match all affixed
subconstituent glosses (i.e. <ln_*> and <rn_*>, but not simple
<ln> or <rn>).
mc_subfunc(data, subforms = c("np", "pro", "refl", "indef"))A data.table in multicastR format,
containing minimally columns for text, GRAID function gfunc,
and GRAID form gform.
A character vector of GRAID form symbols that should be assigned the function of the phrase they are subconstituents of.
The data with an additional column pfunc combining the
values of gfunc with functions copied to phrasal subconstituents.