Investigating the tense contrast in parasitic gap constructions

It has often been observed that parasitic gaps fare better in untensed adjunct clauses than in tensed ones. García Mayo & Kempchinsky (1994) argue that this contrast emerges in Romance languages, but not in English. We present an acceptability judgment experiment to investigate the tense effect in different varieties of French and Spanish, and American English. While the main effects of tense and parasitic gaps in the Romance languages were significant, we did not find super-additivity. We did find super-additivity in the English data, although the experimental design prevents us from comparing the datasets. Our results do not support García Mayo & Kempchinsky’s analysis. Furthermore, we compare our results with earlier experiments. We tested different language varieties for the sake of representation, but also because language contact with Germanic languages or Quechuan languages and Shawi may lead to transfer effects, reducing participants’ sensitivity to the tense contrast compared to Metropolitan French and European Spanish. The patterns in the Spanish datasets corroborate this hypothesis. Finally, we find variation between the adjunct types we tested in the Romance subexperiments (after-, before-, and without-clauses). Our findings thus provide an empirical basis for future research into the phenomenon.