Which feature often marks an unsound textual inference?

Determining the Intended Meaning of Texts and Recognizing Weak Inferences