2. The PCT label also distinguishes this theory fromother sociologicaltheories with similar names (Gibbs 1989; Collins 1988:451ff.; Hagan, Gillis,and Simpson 1985, 1990; Hagan, Simpson, and Gillis 1987).
3. This simplicity of calculation is one of the majordifferences between PCTmodels and the robotics models usually proposed by Artificial-Intelligenceresearchers or other cognitive psychologists.
4. Recently Powers and other PCT researchers havebegun to refer to thisaspect of control theory as HPCT, for "hierarchical perceptual control theory."The word "hierarchy" in this context refers entirely to levels of perceptionwithin the human organism, not to any social hierarchy comprising people inunequal relationships. To avoid confusion among sociologists, who usuallythink in terms of social hierarchies, I will in this paper use only the acronymPCT to describe the theory as a whole.
5. The PCT-human is thus someone who tends to"satisfice" rather than"optimize" (see March and Simon 1958).
6. The connection between emotion and system gain wassuggested to me byRichard Marken.
7. Powers, however, has presented his own views on theimplications of PCT for social organization in several places (e.g.,1992:91-127). My presentation here is intended to be generally compatible withhis perspective.
8. Several demonstrations of "the test" have beencarried out in thelaboratory (Marken 1982, 1983, 1989; Powers 1989b; Wiggins and Heise 1987).
9. Hugh Gibbons, writing about the legal system from aPCT perspective, hassuggested that the "deep structure" of American law includes axiom that "thewill of each person is entitled to respect" (1984:175). By this axiom, acts offorce, by which one person deliberately disrupts the perceptual control (whichGibbons takes as synonymous with will [p. 171]) of another can only bejustified if is "part of a general scheme of respect for the will of the other"(p. 175).
10. My discussion in this section on interpersonalcontrol depends heavily onWartenberg (1990), whose perspective, though not based on PCT, is highlycompatible with it.
11. Just as in the case of force and coercion, theuse of incentives enablesone person to discourage certain actions of a second person but depends on thesecond to choose a reference level for any alternative action. Theoperant-conditioning experimenter, for instance, must wait for the experimentalanimal to perform the desired action, or some approximation of it, before anyreinforcement can take place (see Skinner 1974). Subsequent reinforcements canthen be used to "shape" the action more closely to the experimenter'sspecifications by, in effect, starving out the alternatives. However, noamount of reinforcement can cause an animal to perform an action for which itlacks the appropriate pre-existing control systems (see Hershberger 1986).
12. This is not to deny the similarities betweenpeople in their perceptualorganization. Powers posits the existence of a "Boss Reality" (1992:204)which, although it cannot be directly known, places severe constraints on howmuch the control systems relating to a given aspect of the environment can varyand still be capable of maintaining stable perceptual control. This isreminiscent of Berger and Luckman's arguments about the "massive facticity" ofsocial reality (1966).
13. Frye (1983) makes use of a similar metaphor, thebars of a birdcage, indescribing structures of oppression.
14. A few over-eager control theorists haveconjectured that the hierarchiesthat make up business or family organizations function in the same way ascontrol systems or hierarchies of control systems within the human body (e.g.,Lord and Kernan 1987; White 1986). My discussion of the problems inherent ininterpersonal control suggests some of the difficulties of this position (seealso Powers 1986, 1992:83,91-127).
15. Stryker (1987) contends that researchersinfluenced by Powers, includingMcPhail and Heise, have contributed to the "revitalization" of the symbolicinteractionist tradition in recent years.
16. While Fararo sees some continuity between Parsonsand PCT, the readershould not conclude that PCT is simply an updated version of Parsons or ofsystems thinking in general as it has been practiced in sociology (e.g., Batesand Harvey, 1975; Buckley, 1967), nor that it is a recycling of the generalsystems proposals of Miller (1978) and von Bertalanffy (1967). Richardson(1991), in his comprehensive history of cybernetic thought and systems modelsin social sciences, delineates clearly the differences between Powers's viewsand other more conventional system thinking. Richardson, incidentally, sees inPowers's thought promising material for reintegrating divergent socialscientific ideas about feedback.