The quote below from another forum discusses Arp's contribution.
Among the several thousand quasars known today (cf.
Hewitt & Burbidge 1993; VeŹron-Cetty & VeŹron 1993;
Hewett, Foltz, & Chaćee 1995) there are a number of cases
where a quasar is found in close angular proximity to a
galaxy (Monk et al. 1986; Arp 1987; Stocke et al. 1987;
Burbidge et al. 1990; Borgeest et al. 1991; Bowen et al.
1991; Womble 1993; Burbidge 1995), but where the redshifts
of the galaxy and the quasar are notably dićerent
from each other. This led Arp and others to conclude that
this points to a Doppler interpretation of the observed redshifts
of the quasars (Arp et al. 1990 and references therein).
In this hypothesis quasars are ejected from galaxies (cf.
Valtonen & Basu 1991) and, hence, do not lie at those
cosmological distances which are inferred from their measured
redshifts. This point of view has been criticized by
various authors (ôô The Redshift Controversy ¤¤ ; Weedman
1976). Serious arguments against the hypothesis of Arp et
al. are the agreement of the cosmological interpretation
with the observational data from gravitationally lensed
quasars (e.g., Dar 1991), the detection of the host galaxies
of some quasars (e.g., Bahcall 1995; Bahcall, Kirhakos, &
Schneider 1995; Disney et al. 1995; Hutchings & Morris
1995), the nondetection of tidal perturbations in the morphology
of quasar-galaxy associations (e.g., Sharp 1985,
1986), or other reasons (e.g., Newman & Terzian 1995).
Although the arguments for the cosmological interpretation
of the quasar redshifts are highly convincing, here I
discuss another observational test which could allow us to
check whether the apparent close angular proximity of
some quasars to galaxies is due to a spatial closeness of
these objects to each other.