Isothermal data of high molecular weight polyisobutylene obtained by mechanical measurements with a spectral range over eight decades and additional photon correlation measurements have found that there are three distinct viscoelastic mechanisms in the glass-rubber transition zone. Theoretical considerations have helped to identify these three mechanisms to originate separately from local segmental (alpha) modes, sub-Rouse (sR) modes, and Rouse (R) modes. The temperature dependences of the shift factors of these mechanisms, a(T,alpha), a(T,sR) and a(T,R), determined over a common temperature range are found to be all different. The differences in temperature dependences are explained quantitatively by the coupling model. The local segmental motion contributes to compliances ranging from the glassy compliance, J(g), up to 10(-8.5) Pa-1. The sub-Rouse modes contribute in the compliance range, 10(-8.5) less than or equal to J(t) less than or equal to 10(-7) Pa-1. The Rouse modes account for the compliances in the range of 10(-7) Pa-1 less than or equal to J(t) less than or equal to J(plateau), where J(plateau) is the plateau compliance. The magnitudes of the bounds given here are only rough estimates. Shift factors, a(T), obtained by time-temperature superpositioning of viscoelastic data taken in the softening transition over a limited experimental window are shown to be a combination of the three individual shifts factors, a(T,alpha), a(T,sR), and a(T,R). Consequently, care must be exercised in interpreting or using the WLF equation that fits the shift factors of the entire softening dispersion, because the latter do not describe the temperature dependence of any one of the three viscoelastic mechanisms.