I believe or trust you left.

[arademaker]

The first MRS I got was not solvable by https://github.com/coli-saar/utool/releases/tag/utool-3.4

[ TOP: h0
  INDEX: e2 [ e SF: prop TENSE: pres 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 ]
          [ _believe_v_1<2:9> LBL: h1 ARG0: e8 [ e SF: prop TENSE: pres MOOD: indicative PROG: - PERF: - ] ARG1: x3 ARG2: h9 ]
          [ _or_c<10:12> LBL: h1 ARG0: e2 ARG1: e8 ARG2: e10 [ e SF: prop TENSE: pres MOOD: indicative PROG: - PERF: - ] ]
          [ _trust_v_1<13:18> LBL: h1 ARG0: e10 ARG1: x3 ARG2: h11 ARG3: i12 ]
          [ pron<19:22> LBL: h13 ARG0: x14 [ x PERS: 2 IND: + PT: std ] ]
          [ pronoun_q<19:22> LBL: h15 ARG0: x14 RSTR: h16 BODY: h17 ]
          [ _leave_v_1<23:27> LBL: h18 ARG0: e19 [ e SF: prop TENSE: past MOOD: indicative PROG: - PERF: - ] ARG1: x14 ARG2: i20 ] >
  HCONS: < h0 qeq h1 h6 qeq h4 h9 qeq h18 h11 qeq h18 h16 qeq h13 > ]

[danflick]

The ERG's analysis of conjoined verbs that take a scopal argument, as with "believe or trust", produces an MRS which is inconsistent with our general assumptions about well-formedness, since both of the EPs for _believe_v_1 and _trust_v_1 have the same handle argument for their ARG2 (in fact the only available handle to fill both holes, namely the top handle of the embedded clause "you left"). One way to improve on this MRS might be to introduce a separate clausal "ellipsis" predication whose handle would be (qeq) the ARG2 of one of the two verbs, so the meaning of the sentence would be similar to "I believe you left, or trust so." (see for a more natural example "I believed that you left, and said so").