The variability of acoustical alarm signals among muroid rodents may be understood in terms of a quantitative analysis of their behavioral fitness value. These alarm signals, which are given by a muroid rodent either before or during flight behavior, include teeth-chattering, tail-rattling, hind-feet-thumping, forefeet-pattering, and squeal vocalization. Smith  has considered the fitness value of such alarm calls as deriving from the advantage accruing to closely related animals who hear the alarm signal and freeze or escape before they can be caught by a predator. Paradoxically, the signaler, itself, may be expected to lose fitness in the more limited sense, because it exposes itself more to potential predation. After subjecting the theory to a mathematical analysis, Smith comes to the conclusion that alarm behavior might originate as signals by parents which warned their offspring to flee to safety. Further he considers that they could be maintained in adult populations providing they were made frequently by many members of the population, and providing that these members were closely related. However, the equilibrium conditions for such selection were judged to be inherently unstable. Perhaps it is this theoretical instability that can explain the diversity and variability of motor patterns of alarm. One may assume that alarm signals have appeared and disappeared many times in the course of muroid rodent evolution and development, appearing when the species lives in stable communities of closely related animals, and disappearing when living conditions change.
In contrast to threat and alarm signals, the basic motor patterns of offense and defense remain invariate across species, presumably because they rely on very powerful motivating and releasing stimuli in the opponent. Freezing is practically universal among muroid rodents, presumably because all potential predators use visual movement or noise in their hunting behavior. The bite-and-kick and lunge-and bite attacks are practically universal, presumably because all opponents respond to pain as a powerful motivating stimulus for defense. And flight, unlike other motor patterns, does not depend upon the opponent's response, providing the fleeing animal can reach safety before its opponent catches it.
Much more research is needed to tease out the contributions of genetic and developmental factors to the differences in agonistic behavior among muroid rodent species. Some genetic factors should be expected and analyzed [Adams, 1979c]. However, if developmental factors are ignored, there is the danger that the analysis of behavior will be overly deterministic, as if the animal were passive and its behavior totally determined by immediate stimuli. There is also the danger of presupposing that differences in behavior across species are usually determined by inborn, "species-specific" mechanisms. Unfortunately, as noted earlier, the developmental factors in agonistic behavior have not received adequate attention in research and, as a result, cannot be considered in detail in the present review. Hopefully, they will receive that attention in future research and, consequently, in future reviews.