Heckman (1980) recognized the class of emission-line galaxies he called LINERs as low-ionization AGNs. They seem to be objects photoionized by the same type of hard spectrum, extending to high energies, as in Seyfert galaxies, but with a lower mean ionization parameter (Ferland and Netzer 1983; Keel 1983a; Halpern and Steiner 1983). This can be due either to a lower-luminosity central source, or greater mean distance of the ionized gas from the source; probably both situations appear in nature. A careful study of a well defined sample of spiral galaxies by Keel (1983b) shows that approximately half of these objects have observed Seyfert (5%) or LINER (50%) spectra.
Among the LINERS he studied, Keel (1983a) found several cases in which very weak, barely detectable broad H emission components are present in addition to the narrow emission lines. Another example of this type of object is Mrk 883, whose spectrum was published by Osterbrock and Dahari (1983).
In another very interesting study, Filippenko and Sargent (1985) made a special search for such weak broad H emission components in a large number of LINERs. This necessitated accurate subtraction of the galaxy integrated-stellar absorption line spectrum, to detect the weakest possible emission components. They found that about a significant fraction of the LINERs analyzed in this way have a measurable broad H emission component. Of course these observational results strongly suggest that all AGNs have some BL gas within them, whether it is bright enough to be detected in their spectra or not.
Among the LINERS in which Filippenko and Sargent (1985) detected weak, broad H emission are the best known elliptical galaxies with emission lines in their spectra, NGC 1052 and the N1GC 4278. This is the most convincing evidence known to me to date that the elliptical-galaxy has LINERs belong to the same general type of AGNs photoionized by a hard spectrum extending to high energies, as in spiral-galaxy nuclei.