Information transfers occur if announcements made by one group of one or more firms contemporaneously affect the returns to shares of another group of one or more non-announcing firms. Some of the currently accepted definitions and motivations used in empirical investigations of earnings information transfers are explored. The difficulty in invoking directly the usual policy-based motivations for accounting research and the absence of startling results suggest that it may be useful to reconsider the views about the objectives of information transfer research. The definition of information transfers used in much empirical accounting research focuses on corporate announcements. Three kinds of motivation for information transfer research are considered: 1. those related to disclosure policies, 2. those that extend some existing research on information content, and 3. those that are statistical or descriptive.