Quantifying the meanings of words.

Whether in talking with each other, interacting with patients, teaching students, or writing reports, precision in communication is always desirable. Certainly as issues of "informed consent" have emerged—first on the medical practice scene and now on the medical research scene— care and carefulness in communication seem to be capturing more attention daily. The past several years have seen a variety of programs initiated to alleviate communication difficulties between physicians, between physicians and patients, between computers and users, and among various health professionals. The success of such efforts will always depend on consistent interpretation of the words used. In 1980, Bryant and Norman, 1 concerned about physician-to-physician communication through laboratory and x-ray reports, selected 30 terms used frequently in those reports—such as "probably," "sometimes," and "unlikely"—which seemed subject to repeated misinterpretations or to differences in interpretations by different physicians. They asked a group of hospital-based Canadian physicians to specify on a