Over the last five years several experimental groups have reported anomalies in the temperature dependence of the period and amplitude of a torsional oscillator containing solid 4 He. We model these experiments by assuming that 4 He is a viscoelastic solid-a solid with frequency dependent internal friction. We find that while our model can provide a quantitative account of the dissipation observed in the torsional oscillator experiments, it only accounts for about 10% of the observed period shift, leaving open the possibility that the remaining period shift is due to the onset of superfluidity in the sample.PACS numbers: 67.80. -s, 67.40.-w, 46.35.+z In 2004 Kim and Chan [1,2] reported anomalies in the resonant period of a torsional oscillator (TO) containing solid 4 He. With exquisite sensitivity they detected a reproducible decrease in the oscillator period upon lowering the temperature below 200 mK. Subsequent experiments in several laboratories [3,4,5,6,7,8,9,10,11] A natural interpretation of the TO anomalies is the onset of the elusive and long-anticipated supersolid phase of matter [14,15,16]. In a supersolid, superfluidity coexists with the crystalline order of a solid; one expects a supersolid to exhibit superflow, much like a superfluid. Leggett [16] proposed that the superflow is best detected by searching for "nonclassical rotational inertia"; a superfluid condensate would remain at rest and not participate in rotation, and the resulting mass decoupling would reduce the rotational inertia and decrease the resonant period of oscillation. While compelling, the supersolid interpretation of these experiments has yet to be corroborated by other measurements, such as the response to pressure differences [17]. Moreover, Day and Beamish [18] reported a pronounced increase in the shear modulus of 4 He at temperatures below 200 mK, with a dependence on measurement amplitude and 3 He impurity concentration similar to the TO anomalies. Their results suggest that changes in the shear modulus might be intimately related to the TO anomalies.This Letter presents a detailed discussion of the mechanical response of a TO containing a viscoelastic solid. We build on earlier work by Nussinov et al. [19], who correctly identified a "back action" term in the TO equation of motion that represents the dynamical effect of the solid helium on the torsion cell. However, in contrast to Nussinov et al. we find no need to assume that the solid helium behaves as a glass. Instead, with a few carefully stated assumptions, we find that we can model the solid helium as a classical viscoelastic solid-i.e., a solid with internal friction. The TO period shift and dissipation peak are naturally related to the real and imaginary parts of the frequency-dependent shear modulus of the solid. We use our results to fit the dissipation peak reported in Clark et al. [5], and extract a temperature dependent time scale τ (T ) from the data. With all of the phenomenological parameters determined, we find that we are only able to account for ab...