But more interesting is the 
likelihood of interaction 
among cues. Interaction in 
science means that a 
combination of two factors 
produces an outcome (positive 
or negative) that neither alone 
nor the mere sum of the two 
would yield. As noted earlier, 
some kinds of information are 
absolute, referring to the 
distance of an object from us; 
other kinds of information are 
relative, referring to the depth 
between objects. One serious 
problem not yet considered is 
that few cues are suitable for 
giving us absolute distance 
information. Only 
convergence, accommodation, 
and familiar size qualify. For 
one reason or another already 
discussed, the second and third 
of these are questionable, thus 
leaving only convergence. But 
it seems unlikely that the angle 
at which the eyes converge on 
one object--which is potential 
information only about its 
distance and only for limited 
distances at that--could be the 
source of the simultaneous 
impression we typically seem to 
have of the distances from us of 
all things in the scene. But 
convergence in interaction 
with stereopsis or with 
pictorial cues could yield such 
an impression. The logic is 
this: If, in the figure to the 
left, the cylinder appears to be 
X distance from us because of 
the cue of convergence, and if 
the pyramid appears to be Y 
distance behind the cylinder 
because of the cue of 
stereopsis, it follows that the 
pyramid is X + Y distance from 
us. Similar interactions may 
occur between convergence 
and pictorial information. 
Thus, if pictorial information 
leads to a perceptual scene that 
is vividly three-dimensional 
and convergence anchors any 
single point in it to a definite 
distance from us, then, ipso 
facto, the whole scene takes on 
the appropriate set of absolute 
distances. But without such 
absolute distance, the display 
lacks realism.