I review the radio to X-ray properties of GHz Peaked Spectrum (GPS)
and Compact Steep Spectrum (CSS) sources, the current hypotheses for
their origin, and their use to constrain the evolution of powerful
radio galaxies.
The GPS and CSS sources are compact, powerful radio sources with
well defined peaks in their radio spectra (near 1 GHz in the GPS,
and near 100 MHz in the CSS). The GPS sources are
entirely contained within the extent of the narrow-line region
(
1 kpc) while the CSS sources are contained entirely within the
host galaxy
( 15 kpc).
The peaks in the spectra are probably
due to synchrotron self-absorption, though free-free absorption through an
inhomogeneous screen may also
play a role. The turnover frequency varies with linear size l as
m 
l-0.65
suggesting a simple physical relationship between
these parameters. The radio morphologies are strikingly like
those of the large scale classical doubles, though some sources
can have very distorted morphologies suggestive of interactions. Radio
polarization tends to be low, and in some cases the Faraday Rotation Measures
can be extremely large.
The IR properties are consistent with similar stellar populations and active
galactic nuclei (AGN) bolometric luminosity as the 3CR classical doubles.
The optical host galaxy properties (absolute
magnitude, Hubble diagram, evidence for interaction) are consistent with
those of the 3CR classical doubles. CSS sources at all redshifts
exhibit high surface brightness optical light (most likely emission line gas)
which is aligned with the radio axis. The optical emission-line
properties suggest (1) interaction of the radio source with the emission-line gas
and (2) the presence of dust towards the emission-line regions.
X-ray observations of high-redshift GPS quasars and a couple GPS galaxies suggest
the presence of significant columns of gas towards the nuclei. Searches for cold
gas in the host galaxies have revealed large amounts of molecular gas and smaller
amounts of atomic gas in several sources; though probably not enough to confine
the radio sources. The main competing models for the GPS
and CSS sources are that they are (1) frustrated by interaction with dense gas
in their environments and (2) young and evolving radio sources which will become
large scale sources. Combining the bright GPS and CSS samples with the 3CR results
in a sample spanning a range in source size of
105 which can be used to study
source evolution. The number density vs linear size is consistent with a
picture in which the
sources expand with constant velocity and the radio power drops with linear size
l as P l-0.5. This strong evolution suggests that at least some of the
GPS and CSS sources evolve to become lower luminosity FR1 radio sources.
The GPS and CSS sources are important probes of their host galaxies and will
provide critical clues to the origin and evolution of powerful radio sources.
