The distribution attached to a parameter before certain data are obtained. The posterior distribution is the distribution attached to it after the data are obtained. The posterior distribution may then be considered as a new prior distribution before further data are obtained.