In this paper, we will give examples of errors involving subsequent reference found in our analysis of writing samples from ASL [American Sign Language] natives. We will discuss how these errors can be explained by differences between ASL and English. These differences motivate the use of a local focus tracking (or focusing) algorithm, which will also be described. Finally, we will show how we can use focus tracking to identify and correct the illegal NP omissions, and to flag inappropriate pronominalization. This work should be applicable to the correction of reference errors in other CALL systems for second language learners.