Li, Kadiri and Horstemeyer [1] recently studied twinning in titanium by atomic-scale computer simulation and proposed a new mechanism in which elementary twinning dislocations (TDs) are nucleated and glide in an extended fashion on adjacent planes. In this comment, we argue that the interpretation of the simulations is in error for several reasons. First, the Burgers vector of the TDs seen in the simulations was not determined correctly. Second, these TDs do not produce the twin mode known to occur in titanium. Third, the experimentally observed mode occurs under c-axis compression, whereas the motion of the twin boundary in [1] was in response to c-axis tension. The former mode cannot be simulated with the MD model used in [1]. Fourth, the temperature dependence of twinning found experimentally was misunderstood in [1]. We conclude that the TD responsible for this deformation mode is the one long-established by classical twinning theory and studied at the atomic level by computer simulations performed more than 20 years ago.