Ambiguous Masculine Agreement. Ambiguous masculine agreement, i.e., where it may be the result of either the default agreement or single-conjunct agreement—MF (M = 100%), MN (M = 98%), FM (M = 10%), and NM (M = 5%). A t-test shows that there is a statistically significant difference between masculine agreement when Conj1 versus Conj2 is masculine (p < .001), with a weak effect of masculine in the distant or last-conjunct position. Considered together, the results of the experiment on conjunct- agreement production in postverbal position show a decrease in ConjP and Conj2’s potential to control the gender agreement on the predicate and conse- quently a higher and more stable overall dominance of the linearly closer con- junct (in this case Conj1) as the controller. While the low potential of the Conj2 controller in postverbal order is predicted by all accounts, the asymmetry in stability of ConjP as controller across two word orders is unexpected and will be addressed in section 4.2. Moreover, these data strongly argue in favor of analyzing agreement with Conj1 in the postverbal position as the result of two agreement strategies: highest-conjunct agreement and closest-conjunct agreement, which display an additive effect, as observed for feminine in FM (F = 90%) and neuter in NM (N = 95%). The decrease in ConjP’s agreement potential in postverbal position has to be incorporated in any comprehen- sive account of conjunct agreement. Finally, in the postverbal environment we do not observe the interaction of agreement controllers with different gender values. However, the dominance of linear agreement in postverbal position could be disguising the contribution of a conjunct’s gender value in boosting agreement in mixed-gender conditions.
