Fw_seq often has unscoped `x` args

#1

Is it by design or a bug that the ERG 2020 will generate MRS that has fw_seq using x args that aren’t scoped by a quantifier?

I believe the i arguments in the fw_seq predications below are i type (and not x) because the ERG is trying to avoid having to scope them. I’m confused as to why this didn’t happen for the remaining one. (also: if it is a bug, is this still the best place to enter bugs?)

In the example below, fw_seq(x13,i15,i16) has the unscoped x13:

"I said "Hi there, friend!""
[ TOP: h0
  INDEX: e2 [ e SF: prop TENSE: past MOOD: indicative PROG: - PERF: - ]
  RELS: < [ pron<0:1> LBL: h4 ARG0: x3 [ x PERS: 1 NUM: sg IND: + PT: std ] ]
          [ pronoun_q<0:1> LBL: h5 ARG0: x3 RSTR: h6 BODY: h7 ]
          [ _say_v_1<2:6> LBL: h1 ARG0: e2 ARG1: x3 ARG2: x8 [ x PERS: 3 NUM: sg ] ]
          [ proper_q<7:26> LBL: h9 ARG0: x8 RSTR: h10 BODY: h11 ]
          [ fw_seq<7:26> LBL: h12 ARG0: x8 ARG1: x13 ARG2: i14 ]
          [ fw_seq<7:17> LBL: h12 ARG0: x13 ARG1: i15 ARG2: i16 ]
          [ quoted<8:10> LBL: h12 ARG0: i15 CARG: "Hi" ]
          [ quoted<11:16> LBL: h12 ARG0: i16 CARG: "there" ]
          [ quoted<18:24> LBL: h12 ARG0: i14 CARG: "friend" ] >
  HCONS: < h0 qeq h1 h6 qeq h4 h10 qeq h12 > ]

                                                 ┌──────── quoted(friend,i14)
                                                 │ ┌────── quoted(there,i16)
                                                 │ │ ┌──── quoted(Hi,i15)
                                     ┌────── and(0,1,2,3,4)
                                     │                 └─│ fw_seq(x13,i15,i16)
               ┌────── pron(x3)      │                   │
pronoun_q(x3,RSTR,BODY)              │                   │
                    │                │                   └ fw_seq(x8,x13,i14)
                    └─ proper_q(x8,RSTR,BODY)
                                          └─ _say_v_1(e2,x3,x8)
#2

You’re right that this is a bug in ERG 2020. It is fixed in the upcoming 2023 release, so no need to file a bug report. The intent of the foreign-word sequence construction is to allow a chain of any number of tokens delimited by quotes or italic-marks, with each token in the chain connected by a “fw_seq” predication, where the final fw_seq introduces a referential index bound by a proper_q quantifier. Each of the tokens introduces a “quoted” predication whose ARG0 is of type “i”, since we don’t know whether the token signifies an eventuality or an entity.

1 Like