Display Diversity 
281 
nomic endeavor. However, the function of a 
behavior must be understood to ultimately 
classify the displays of a species’ repertoire. 
Unfortunately, this understanding is ob- 
tained by inference. While the descriptive 
approach to display analysis can be quanti- 
fied to give a detailed statistical description 
of the physical signals, one assigns function 
to a display type by correlating the display 
type with the social contexts in which it 
appears. This is still a relatively unrefined 
means of evaluating the information content 
of the displays. Nevertheless, this technique 
must be employed to verify the classification 
of descriptively derived display types. 
Context can be a criterion for determining 
display types. Display variability can make 
it difficult to distinguish between what might 
be two distinct but similarly appearing dis- 
play types and what might be a single display 
type with much variation. Specifically, the 
signature display (A display) and the B dis- 
play of A. Umifrons are differentiated by the 
appearance of the dewlap in relation to the 
head-bob pattern (Fig. 8). During the signa- 
ture display, the dewlap appears before the 
first head bob, and in the B display the dew- 
lap appears during the first bob or as late 
as the fifth bob (Hover and Jenssen, 1976). 
Such a division of display types might seem 
arbitrary were it not that the signature dis- 
play and the B displays appear in different 
contexts. 
To quantify the difference of display usage 
by A. Umifrons, a cage 3 meters long was 
divided in half by a sliding partition (Fig. 
1), and a male and female were paired in 
each end. After a day, observations were 
made for an hour with the partition in place. 
Dewlapping and signature displays composed 
98 percent of the displays during male-fe- 
male association; B displays were rare. The 
partition was then removed, and displays 
were recorded for an equal length of time. 
The frequency of B displays rose seven-fold 
during male-male interactions. This observa- 
tion and the fact that in almost 700 observed 
displays the signature display (A display) 
appeared eight times more frequently than 
the most common B display variant indicates 
to me that there is no functional continuum. 
Dividing the behavior into two display types 
was suggested by the animals’ own use of 
these display patterns. 
Distinctions between different display 
types within a species’ repertoire and actual 
socially significant variations of a particular 
display type may be complex and open to 
interpretation. This interpretation will re- 
flect the investigator’s conceptual orienta- 
tion. For this reason, a framework is needed 
in which to categorize not only the stereo- 
typed portions of display behavior, but also 
the seemingly less predictable elements. By 
being able to analyze and evaluate the entire 
behavior, we will make greater progress in 
unraveling the enigmas of display function 
and evolution. See Table 2 for selected ex- 
amples. 
SPECULATION: FACTORS AFFEC'HNG 
THE EVOLUTION OF DISPLAY 
CHARACTERISTICS 
Because anoline lizards are easily studied 
in the field and have undergone extensive 
ecological radiation, AnoUs investigations 
have generated notable advances in ecological 
theory (Lister, 1976a, h; Rand, 1969; Rough- 
garden, 1972; Schoener, 1967, 1968, 1969, 
1970a, 19706, 1971; Williams, 1969, 1973). 
This same genus is a potential fountainhead 
for understanding factors influencing be- 
havioral evolution. 
Some immediate questions and specula- 
tions which come to mind regarding the 
evolution of anoline visual signals are the 
relationship between display characteristics 
and the following: 
Types of Social Organization 
Large-bodied species with relatively ex- 
pansive territories might be expected to have 
simple repertoires containing few display 
types, with most ritualized disputes being 
settled while the conspecifics are still widely 
separated. The display type for declaring 
