Volume 48, Issue 4 (July 2003)
Commentary on: Srihari SN, Cha S-H, Arora H, Lee S. Individuality of handwriting. J Forensic Sci 2002; 47(4):856–72.
Although the proposition had never been tested in any systematic way, many forensic document examiners have long assumed that all writing is “unique”—that “no two people write alike and no one person writes the same way twice.” A recent article by Sargur Srihari and his colleagues (1) reported a study which aimed to prove, for the first time, the truth of at least half of that proposition: that the writing of every person is distinguishable from that of every other person. Srihari et al. drew handwriting exemplars from 1568 individuals, a sample they sought to make representative of the U.S. population. The study comprised two major analyses, each involving about 700 writers. Computer algorithms were used to extract features from scanned images of the handwriting. Attributes of the handwriting were compared at different levels (document, paragraph, word, character) to try to distinguish writers from each other.