The linear excitation of straight-crested, capillary–gravity waves on the surface ( x > 0) of a deep, viscous liquid in response to the sinusoidal, vertical motion of a hydrophilic wall at x = 0 is calculated on the assumptions that: (i) the dynamical variation of the contact angle is proportional to (but not necessarily in phase with) the velocity of the contact line relative to the wall; (ii) the relative tangential velocity (slip) of the fluid below the contact line is proportional to the shear at the wall, (iii) k 0 l v [Lt ] 1 and k 0 l c = O (1), where k 0 is the wavenumber, l v is the boundary-layer thickness, and l c is the capillary length. The contact-angle and slip coefficients are complex functions of frequency that are found to be linearly related. Physical considerations suggest that the slip length l s (≡ slip velocity ÷ shear at wall) should be small compared with l v , which, in turn, implies that the motion of the contact line must be small in that parametric domain in which linearization provides a viable description of the wave motion near the wall; however, the analysis proceeds from (i) and (ii), qua phenomenological hypotheses, without a priori restrictions on the contact-line and slip coefficients. The present results include those of Wilson & Jones (1973), who assume that the amplitude and phase of the wave slope at the wall are prescribed, and those of Hocking (1987 a ), who assumes that the variation of the wave slope at the wall is in phase with the contact-line velocity and neglects viscosity. They also include a correction for the dynamical effects of the static meniscus, which is necessarily present for any static contact angle other than ½π but is neglected in the previous analyses, and have counterparts for the closely related problem (cf. Hocking 1987 b ) of the reflection of a plane wave from a stationary wall.