Researchers generally study expository and narrative text separately, even though some theorists maintain that a text's discourse type has little to do with a reader's processing of that text and that other variables may account for more differences in the way texts are read. The research described here investigates the relation between text type, tone, and readers' responses. Fifty-eight students described their expectations about text type. Three weeks later, they wrote free written responses to informational text that was expository, narrative, and descriptive and had either a personal or an impersonal tone. The texts were similar in content, and familiarity with the content was statistically controlled. The results indicated that narrative and expository text were not responded to differently. However, although subjects made significantly more text-based than reader-based statements regardless of text type, they made fewer text-based and more reader-based statements when reading descriptive text than when reading expository or narrative text. Further, subjects who read text with an impersonal tone made significantly more improvement from pre-to posttest on knowledge measures than subjects who read text with a personal tone.Reading professionals generally accept that, before and during reading, readers develop expectations that guide their understanding of and reaction to text. In the absence of an externally imposed purpose for reading, such as to pass a test, readers may construct their own processing guidelines, relying on, among other things, certain features of the text. Some hypothesize, for example, that a text's genre or discourse type (e.g., narrative or expository) is an important consideration when deciding how text should be read (e.g., Langer, 1986). Other factors such as 281