Around 50 planetary nebulae (PNs) are presently known to possess "small-sca
le" low-ionization structures (LISs) located inside or outside their main n
ebular bodies. We consider here the different kinds of LISs (jets, jetlike
systems, symmetrical and nonsymmetrical knots) and present a detailed compa
rison of the existing model predictions with the observational morphologica
l and kinematical properties. We find that nebulae with LISs appear indisti
nctly spread among all morphological classes of PNs, indicating that the pr
ocesses leading to the formation of LISs are not necessarily related to tho
se responsible for the asphericity of the large-scale morphological compone
nts of PNs. We show that both the observed velocities and locations of most
nonsymmetrical systems of LISs can be reasonably well reproduced assuming
either fossil condensations originated in the asymptotic giant branch (AGB)
wind or in situ instabilities. The jet models proposed to date (hydrodynam
ical and magnetohydrodynamical interacting winds or accretion disk collimat
ed winds) appear unable to account simultaneously for several key character
istics of the observed high-velocity jets, such as their kinematical ages a
nd the angle between the jet and the symmetry axes of the nebulae. The line
ar increase in velocity observed in several jets favors magnetohydrodynamic
al confinement compared to pure hydrodynamical interacting wind models. On
the other hand, we find that the formation of jetlike systems characterized
by relatively low expansion velocities (similar to those of the main shell
s of PNs) cannot be explained by any of the existing models. Finally, the k
nots that appear in symmetrical and opposite pairs of low velocity could be
understood as the survival of fossil (symmetrical) condensations formed du
ring the AGB phase or as structures that have experienced substantial slowi
ng down by the ambient medium.