"Just two, my son Johnny and me." generates invalid MRS

[EricZinda]

At least parse 39 and parse 71 (I stopped looking after that), generate MRS that can't be made into a well formed tree, because they have one two few holes. Their MRS is below:

Parse 39:

[ "Just two, my son Johnny and me."
  TOP: h0
  INDEX: e2 [ e SF: prop-or-ques ]
  RELS: < [ unknown<0:31> LBL: h1 ARG: x4 [ x PERS: 3 NUM: sg ] ARG0: e2 ]
          [ _just_a_1<0:4> LBL: h1 ARG0: e5 [ e SF: prop ] ARG1: e2 ]
          [ appos<5:31> LBL: h1 ARG0: e6 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x4 ARG2: x7 [ x PERS: 3 NUM: sg IND: + ] ]
          [ number_q<5:9> LBL: h8 ARG0: x4 RSTR: h9 BODY: h10 ]
          [ card<5:8> LBL: h11 ARG0: x4 ARG1: i13 CARG: "2" ]
          [ appos<10:31> LBL: h14 ARG0: e15 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x7 ARG2: x16 [ x PERS: 3 NUM: pl ] ]
          [ def_explicit_q<10:12> LBL: h17 ARG0: x7 RSTR: h18 BODY: h19 ]
          [ poss<10:12> LBL: h20 ARG0: e21 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x7 ARG2: x22 [ x PERS: 1 NUM: sg PT: std ] ]
          [ pronoun_q<10:12> LBL: h23 ARG0: x22 RSTR: h24 BODY: h25 ]
          [ pron<10:12> LBL: h26 ARG0: x22 ]
          [ _son_n_of<13:16> LBL: h20 ARG0: x7 ARG1: i27 ]
          [ udef_q<17:31> LBL: h28 ARG0: x16 RSTR: h29 BODY: h30 ]
          [ proper_q<17:23> LBL: h31 ARG0: x32 [ x PERS: 3 NUM: sg IND: + ] RSTR: h33 BODY: h34 ]
          [ named<17:23> LBL: h35 ARG0: x32 CARG: "Johnny" ]
          [ _and_c<24:27> LBL: h37 ARG0: x16 ARG1: x32 ARG2: x38 [ x PERS: 1 NUM: sg IND: + PT: std ] ]
          [ pron<28:30> LBL: h39 ARG0: x38 ]
          [ pronoun_q<28:30> LBL: h40 ARG0: x38 RSTR: h41 BODY: h42 ] >
  HCONS: < h0 qeq h1 h9 qeq h11 h18 qeq h20 h24 qeq h26 h29 qeq h37 h33 qeq h35 h41 qeq h39 > ]

Parse 71:

[ "Just two, my son Johnny and me."
  TOP: h0
  INDEX: e2 [ e SF: prop-or-ques ]
  RELS: < [ unknown<0:31> LBL: h1 ARG: x4 [ x PERS: 3 NUM: pl ] ARG0: e2 ]
          [ _just_a_1<0:4> LBL: h1 ARG0: e5 [ e SF: prop ] ARG1: e2 ]
          [ udef_q<5:31> LBL: h6 ARG0: x4 RSTR: h7 BODY: h8 ]
          [ appos<5:23> LBL: h9 ARG0: e10 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x11 [ x PERS: 3 NUM: sg ] ARG2: x12 [ x PERS: 3 NUM: sg IND: + ] ]
          [ number_q<5:9> LBL: h13 ARG0: x11 RSTR: h14 BODY: h15 ]
          [ card<5:8> LBL: h16 ARG0: x11 ARG1: i18 CARG: "2" ]
          [ appos<10:23> LBL: h19 ARG0: e20 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x12 ARG2: x21 [ x PERS: 3 NUM: sg IND: + ] ]
          [ def_explicit_q<10:12> LBL: h22 ARG0: x12 RSTR: h23 BODY: h24 ]
          [ poss<10:12> LBL: h25 ARG0: e26 [ e SF: prop TENSE: untensed MOOD: indicative PROG: - PERF: - ] ARG1: x12 ARG2: x27 [ x PERS: 1 NUM: sg PT: std ] ]
          [ pronoun_q<10:12> LBL: h28 ARG0: x27 RSTR: h29 BODY: h30 ]
          [ pron<10:12> LBL: h31 ARG0: x27 ]
          [ _son_n_of<13:16> LBL: h25 ARG0: x12 ARG1: i32 ]
          [ proper_q<17:23> LBL: h33 ARG0: x21 RSTR: h34 BODY: h35 ]
          [ named<17:23> LBL: h36 ARG0: x21 CARG: "Johnny" ]
          [ _and_c<24:27> LBL: h9 ARG0: x4 ARG1: x11 ARG2: x38 [ x PERS: 1 NUM: sg IND: + PT: std ] ]
          [ pron<28:30> LBL: h39 ARG0: x38 ]
          [ pronoun_q<28:30> LBL: h40 ARG0: x38 RSTR: h41 BODY: h42 ] >
  HCONS: < h0 qeq h1 h7 qeq h9 h14 qeq h16 h23 qeq h25 h29 qeq h31 h34 qeq h36 h41 qeq h39 > ]

[danflick]

This bug, which could leave the handle of an embedded apposition NP unconnected, was fixed in the 2023 version of the ERG. Using that version, I can't reproduce the broken MRS for the two parses whose MRSs are given above as #39, where "Johnny and me" is a conjoined NP and this NP is in an apposition with the number "two", or for #71, where the embedded apposition is "my son, Johnny". If you were already using the 2023 version, please also report the parse trees for the problematic readings.