koto probably shouldn't use head-spr to combine with its argument

[emilymbender]

From the comments in pstruc.tex in Jacy book source files:

%% EMB: Cutting this because it does not make crucial use of
%% SPEC feature. Recommend recasting analysis as not using
%% HSR.

%% The second type of construction to use the \type{head-specifier-rule}
%% is nominalizing constructions. Here, a nominalizing head, such as
%% \kjpn[thing]{koto}{こと}, selects for a predicative constituent as its
%% specifier\index{specifier}. This contrasts with ordinary verb/adjective+noun
%% constructions which are built via the head-modifier rule. Furthermore,
%% not all

%% The \type{head-specifier-rule} is used by nominalizing constructions as
%% well. A predicative nominalization\index{nominalization} subcategorizes for a verb, while
%% the verbal endings on the other hand determine the SPEC behavior of
%% the verb. A negative ending for example states that it specifies for a
%% noun. This can be a regular noun, as in (\ref{ex10}) or a
%% nominalization\index{nominalization}, as in (\ref{ex11}). The same is valid for the plain
%% \jpn{ru} ending or the \jpn{tai} ending (\emph{want to}), but not
%% for polite endings like \jpn{masu}, as in (\ref{ex12}).