L-PERIPH does not get properly constrained

I have this type which is supposed to be working as a second position clitic:

ques-clitic-lex := no-hcons-lex-item &
  [ SYNSEM [ LOCAL [ CAT [ VAL [ SPR < >,
                                 COMPS < >,
                                 SUBJ < >,
                                 SPEC < > ],
                           HEAD adv &
                                [ MOD < [ LOCAL intersective-mod &
                                                [ CAT [ HEAD +nvrpd,
                                                        VAL.SPR < > ] ],
                                          L-PERIPH +,
                                          L-QUE - ] > ] ],
                     CONT.RELS.LIST < > ],
             NON-LOCAL [ SLASH.LIST < >,
                         REL.LIST < >,
                         QUE.LIST < >,
                         YNQ.LIST < *top* > ] ] ].

q-particle1-lex := ques-clitic-lex.

It clearly says it’s MOD is L-PERIPH +, and I do see that in the tree:

Screen Shot 2020-07-20 at 2.33.18 PM

But when I look at the local AVM of something it actually modifies in a tree, L-PERIPH is underspecified:

Why would this be?

Oh, that’s because it is its local AVM, right? (Still getting confused about that).

To clarify, I would like to see L-PERIPH - on that subj-head mother (and the tree ruled out). So the question is why L-PERIPH doesn’t get set to - on that subj-head… It should…

Oh I guess the problem is that the clitic is happy to attach to phrases, in this case. Forgot the LIGHT value.