Holistic scoring is widely used to assess writing proficiency in English-as-a-Second-Language (ESL) composition. Written recall protocols have recently been used to investigate the relationship between how much holistic scorers comprehend of a given text and how high they rate the quality of that text. This study investigated whether readers approach a holistic evaluation task differently when they are focused (i.e., aware they will also be tested for comprehension) from when they are naive (i.e., unaware), Subjects were 12 college ESL composition teachers experienced in holistic assessment of writing proficiency. Each read, rated, and recalled one of two texts, then repeated the cycle with the other text. One text was superior to the other. Focused raters recalled significantly more overall but were unable to recall more of the better-written text. It is concluded that recall studies of the holistic reading process should use naive subjects to avoid the interference of an ensuing recall task. It is also suggested that insights into the holistic reading process can support, extend, and, to some extent, reconcile various general reading process/comprehension models. (MSE)