Abstract
Since the discovery of dendritic spines by Ramon y Cajal2, their functional role has remained a matter of speculation. The early hypotheses all consider the establishment of physical contacts with presynaptic terminals as the main function of spines. In a similar spirit, Swindale20 has proposed in a recent issue of TINS that spines are primarily a morphological device for connecting axons and dendrites. Ideas of this type, as Swindale points out, do not preclude other functions of spines, nor do they explain why spines should be as plastic as a number of recent studies suggest6,7,8,19.
