206 
Greenberg 
cept resists attempts at rigorous definition 
and yet is of obvious use. Atz (1970) ex- 
amined the criteria used to recognize hom- 
ologous behavior and argued that only 
morphologically defined behavioral patterns 
could be properly homologized. 
Hodos (1977) extended Atz’s ideas and 
concluded that behavior patterns may be re- 
garded as homologous only to the extent that 
their morphological substrates can be traced 
back to a common ancestral precursor. He 
proposes a morphocentric definition of be- 
havioral homology : 
Behaviors are considered homologous to 
the extent that they can be related to 
specific structures that could, in princi- 
ple, be traced back through a genealogi- 
cal series to a stipulated ancestral 
precursor irrespective of morphological 
similarity. 
According to this definition, two behavioral 
patterns would be homologous if they are 
associated with homologous structures, no 
matter how different the functions of the 
behavioral patterns might be. Such morpho- 
centric definitions of behavioral homology 
rightly regard behavioral patterns as evan- 
escent extensions of anatomical structures 
into evolutionary time. But, it is the very 
disjunction of behavior and morphology that 
Atz regards as the crucial problem in con- 
sidering behavioral homology. 
Strictly applied, a morphocentric defi- 
nition creates a difficulty if two different 
behavioral patterns associated with hom- 
ologous structures have arisen independently. 
To avoid using the term in a misleading 
sense, perhaps the definition should be 
amended to include a stipulated ancestral 
behavioral pattern. Such a condition would 
also satisfy the definitions of behavioral 
homology ventured by Baerends (1958) 
among many others (Atz, 1970). Of course, 
“stipulated ancestral” behavioral patterns 
are derived from the comparative study of 
closely related extant species, and thus all 
proposed homologies must be regarded as 
hypotheses that ultimately are untestable 
(but see Evans, 1959a, 19596, 1959c). The 
heuristic value of such considerations, how- 
ever, is uncontested. 
Hodos also suggests that any statement about 
homology must specify the basis on which 
it is adduced, that is, precisely what the an- 
cestral precursor structure might be. The 
level of analysis depends on the detail of 
anatomical structures considered, and their 
specification is a necessary part of a state- 
ment of a homology or proposed homology. 
To further clarify the issue, Hodos recom- 
mends the use of the terms “homoplasy,” 
referring to similar behavioral patterns, as- 
sociated with nonhomologous structures, and 
“analogy” to refer to behavioral patterns 
which are functionally similar. 
Analogy is important to many researchers, 
particularly comparative psychologists who 
are searching for simplified models for hu- 
man behavior. This anthropocentric selection 
of research problems has been criticized by 
Beach (1960), but, nevertheless, functionally 
analagous behavioral patterns in different 
species are useful for suggesting causal 
mechanisms (Zeigler, 1973). Wickler’s 
(1973) analyses of convergent behavioral 
adaptation provide excellent examples of 
particular behavioral functions which are 
the outcome of different internal mechanisms 
in distantly related animals. Under such cir- 
cumstances we may be more confident that 
similarities in the behavioral patterns of 
species that are not evolutionarily related 
are necessary to the solving of a particular 
ecological problem. Indeed, Lorenz (1971, 
1974) regards the study of analogy crucial to 
an understanding of the survival function of 
a behavioral pattern. 
Regardless of the more useful perspective 
for approaching a specific problem, an ex- 
cellent case can be made for the presentation 
of behavioral data in two complementary 
forms: a behavior inventory, in which be- 
havioral patterns are defined in morphologi- 
cal terms with minimal, if any, regard for 
function, and an ethogram, in which the con- 
text and apparent functions of behavioral 
patterns are described. 
