||Releasing and Directing Stimuli of Defense||Page 16|
The releasing and directing stimuli of the defensive upright posture have been studied in detail for R norvegicus. Vibrissal or facial tactile stimuli are necessary to release the behavior in naive animals [Thor and Ghiselli, 1975], although visual stimuli may suffice for releasing the behavior in experienced animals [Kanki and Adams, 1978]. Vibrissal stimuli are necessary for directing the behavior; this is shown by the fact that an experienced animal will continue to show upright posture in response to visual releasing stimuli following removal of the vibrissae, but the posture is not properly oriented with respect to the opponent [Kanki and Adams, 1978].
Presumably, the lunge-and-bite attack is released and directed in part by visual stimuli, since it may be released and oriented accurately from a distance which precludes tactile stimuli [cf Arvola et al, 1962].
Although appropriate experimental studies have not been done, it would appear that the motor pattern of flight is released and directed by a complex synthesis of stimuli and memory systems which include learned "knowledge" of a "flight path" or "escape route." Animals released in a familiar environment are more likely to flee immediately than animals released in a strange environment; this has been noted in L lemmus [Myllymaki et al, 1962], P polionotus [Blair, 1951], M musculus [Crowcroft, 1966] , and P leucopus [Metzgar, 1967]. The sensory modalities involved may include olfactory, tactile, and visual stimuli, as well as some kind of kinesthetic memory [Myllymaki et al, 1962; Crowcroft, 1966]. The importance of learning and memory [Adams, 1979b] can explain why arboreal species escape into bushes or shrubs while terrestrial species run on trails in the grass and underbrush [Eisenberg, 1968; Smith and Speller, 1970]. The learning can include information regarding lack of an escape route; thus, the likelihood of blind, undirected flight is reduced as laboratory rats learn that there is no exit from a chamber [Blanchard et al, 1976].
Stimuli from the predator or feared object may also function as releasing and directing stimuli of flight. Thus, Blanchard and Blanchard  have shown that a laboratory rat will flee to avoid a discriminable shock "probe," but that it will simply freeze if the shock is delivered through the floor without any discriminable source.
There is no reason to believe that the other motor patterns of defense, such as sound production, piloerection, tail-raising, pheromone secretion, urination, defecation, and hormone secretion require specific releasing and directing stimuli.