Ch. 2— Introduction • 39 
WHAT IS AN ALTERNATIVE? 
Defining the word "alternative" is in a sense al- 
ways doomed to failure: Regardless of how accom- 
modating or strict the definition, many will fault 
it. The term evolved in the political arena, coined 
by animal welfare activists and for the most part 
nonscientists, and yet it has direct implications for 
scientists using laboratory animals. Its meaning 
varies greatly, depending on who uses it and the 
context in which it is used. 
The definition of "alternatives” employed by OTA 
obviously affects this entire assessment: It defines 
the scope of the study. Too narrow a definition 
would dispose of the need for this report, while 
too broad a definition would render it unmanage- 
able. Defining alternatives as the nonuse of ani- 
mals, as some would have it, would restrict the 
bounds of the study to the consideration of inver- 
tebrate organisms, chemicals, plants, and comput- 
ers. On the other hand, stretching the definition 
to include humans, for example, would create a 
whole new series of issues that would be virtually 
impossible to address within one assessment. With 
these concerns in mind, OTA chose to define 
"alternatives" as encompassing any subjects, 
protocols, or technologies that “ replace the 
use of laboratory animals altogether, reduce 
the number of animals required, or refine 
existing procedures or techniques so as to 
minimize the level of stress endured by the 
animal" (4; adapted from 5). 
Some examples of alternatives under this defi- 
nition include computer simulations to demon- 
strate principles of physiology to medical students, 
the use of the approximate lethal dose methodol- 
ogy in acute toxicity studies, and the increased use 
of anesthetics with pain research subjects. The "re- 
duction" part of the definition indicates that the 
increased use of cultured cells, tissues, and organs 
instead of whole animals is also an alternative. A 
very broad interpretation of alternatives might also 
include the substitution of cold-blooded for warm- 
blooded vertebrates. 
BIOLOGICAL MODELS 
When animals— or alternatives— are used in re- 
search, testing, and education, it is because they 
possess a simpler or more accessible structure or 
mechanism in comparison with the object of pri- 
mary interest (which is often the human) or are 
themselves the object of primary interest, or be- 
cause certain procedures cannot be carried out 
on humans. Viewed from this perspective, both 
animals and alternatives stand as models. In the 
broadest sense, a biological model is a surrogate, 
or substitute, for any process or organism of ulti- 
mate interest to the investigator. It is a represen- 
tation of or analog to some living structure, orga- 
nism, or process. 
In addition to analogy, biology has another ana- 
lytical tool at its disposal— homology , which is cor- 
respondence in structure and function derived 
from a common evolutionary origin (i.e., a com- 
mon gene sequence). The most closely related spe- 
cies are generally presumed to offer the best homo- 
logs. Relationships between species are not always 
known in detail, however, and unresolved ques- 
tions about evolutionary events and pathways are 
numerous. Care must therefore be used in evalu- 
ating the degree of homology and the extent to 
which it relates to analogy (3). 
Some biological mechanisms, such as the cod- 
ing of genetic information and the pathways of 
metabolism, arose early in evolution. These mech- 
anisms have been highly conserved and are widely 
shared by organisms, including humans, at the cel- 
lular and molecular levels. Thus, good models for 
these fundamental molecular mechanisms in hu- 
mans can be found in a wide array of organisms, 
some of which, such as bacteria, have structures 
and functions far less complex than those of mam- 
mals (3). 
Several characteristics are important in choos- 
ing a model for research, testing, or educational 
purposes. The most important is the model’s dis- 
crimination— the extent to which it reproduces the 
particular property in which the investigator is 
interested. With greater discrimination, the pre- 
