Attention is given to problems which arise when an attempt is made to use observational material to clarify the mechanisms responsible for the heating of the chromosphere in a late-type star, in order to derive a clearer physical view of the nature of a chromosphere, its relation to the underlying photosphere, and the overlying corona. Problems arise in the prediction of line shapes by means of simple models, in chromospheric inhomogeneities, in two-stream modeling, and in the coexistence of two temperature structures. It is deemed necessary to take the dominant role played by inhomogeneities into account, and to ensure that any interstellar effects in the line profile data are well and truly eliminated.