Person vs. locative agreement: Evidence from late learners and language emergence

Sign languages are frequently described as having three verb classes. One, 'agreeing' verbs, indicates the person/number of its subject and object by modification of the beginning and ending locations of the verb. The second, 'spatial' verbs, makes a similar appearing modification of verb movement to represent the source and goal locations of the theme of a verb of motion. The third class, 'plain' verbs, is characterized as having neither of these types of modulations. A number of researchers have proposed accounts that collapse all of these types, or the person-agreeing and spatial verbs. Here we present evidence from late learners of American Sign Language and from the emergence of new sign languages that person agreement and locative agreement have a different status in these conditions, and we claim their analysis should be kept distinct, at least in certain ways.