FRANKLIN D. MCMILLAN

Best Friends Animal Society, Kanab, Utah, USA

The value of perceived control in the lives of humans and nonhuman animals (hereafter animal) has been the subject of investigation for over a half century. What began as a concept of control in life overall (Rotter, 1966) has been subsequently characterized to also include control over specific domains of one’s life and even a specific aversive event. All of these conceptualizations of control have demonstrated links to animal mental health and well-being.

It has long been known that stressors come in a vast array of forms, and while the various factors that make stressors more or less impactful on the individual are not fully understood, the concept of control has emerged as a crucial factor (Steptoe and Poole, 2016). Stressors differ in the degree to which they can be controlled, and the relation between control over an aversive event and the amount of stress or pain the event produces has formed the basis for control becoming a dominant idea in psychological theory and research (Wallston, 2004; Thompson, 2005). Common sense would suggest that the greater the extent that individuals believe that they can prevent, terminate, or lessen the severity of aversive events, the less reason they would have to worry about or fear them (Rodin, 1986). In humans, control is now a major concept in a number of theories of emotional well-being and happiness (Thompson and Kyle, 2000), and Larsen (1989) has written that the ability to yield control for the short term and endure life’s inevitable occasions of discontrol without substantial distress could reasonably be seen as a defining criterion of mental health.

Research has found control to be crucial for a wide range of animal species, including apes, rodents, birds, and fish, indicating that the organism’s desire to influence the impact of aversive and appetitive interactions with the world appears to be central to animal life (Franks and Higgins, 2012). However, one of the defining features distinguishing captive and wild environments is the reduced amount of control captive animals have over what happens to them and the conditions that affect them (Franks and Higgins, 2012; Bassett and Buchanan-Smith, 2007; Jones and McGreevy, 2007) (Fig. 6.1). Noting this, Markowitz and Eckert (2005) have questioned how we can expect animals in our charge to be mentally healthy if they perceive that nothing that they do matters.

6.1 Defining and Characterizing Control ¹

In the presence of an aversive stimulus, ² control may be defined as the perception that one has the ability to produce a response that can influence the aversiveness of an event in a way that maximizes desired outcomes and/or minimizes undesired ones (Thompson, 1981, 2005; Peterson, 1999). Such a response allows the individual to, at will, terminate the event, make it less probable or less intense, or change its duration or timing (Thompson, 1981; Foa et al., 1992) (see Fig. 6.2).

Choice has frequently been considered an important feature of control. For example, Leotti et al. (2010) argue that choice is the means by which organisms exert control over the environment. They maintain that although much of the cognitive processing for behavior involving choice occurs below the state of explicit awareness, all voluntary behavior involves choice, nonetheless. Accordingly, to choose is to express a preference, and that preference is for options that are conducive to achieving favored rewards (Leotti et al., 2010). People report that the acquisition of choice leads to an increased sense of personal control (Rotter, 1966).

Research has further refined the concept of control to whether actual control is required to be present or if the individual only has to think that it is. Data indicate in both animals and humans that the beneficial psychological and physiological effects of control are experienced even in the absence of true control over aversive events, as well as if the individual has the opportunity to exert control but never actually exercises that option (Perlmuter and Monty, 1977). It is the mere perception of control, independent of whether it is exercised, that appears to be important (Bollini et al., 2004).

6.2 Locus of Control

Rotter (1966) presented evidence that people differ in the extent to which they believe that valued outcomes are determined by forces within or outside their own control. He termed this construct the locus of control (LOC), where the locus referred to whether the person perceives that rewards follow from, or are contingent upon, his own behavior or attributes – an internal locus of control – as opposed to the belief that rewards are controlled by forces outside of himself and tend to occur independently of his own actions – an external locus of control (Rotter, 1966).

Rotter (1966) further noted that one’s sense of control applies to both individual aversive experiences and life overall. This has created confusion in the literature, as ‘perceived control’ is often used without clarification as to whether the writer is referring to a specific event, life in general, or a particular domain of life (Thompson, 2005; Steptoe and Poole, 2016). Yet because the scope of control can differ, one could have a high degree of internal LOC in life yet have no perceived control over a specific situation, and, conversely, have a high degree of perceived control over a specific situation but a low external LOC for life overall. In Rotter’s (1966) original conceptualization LOC referred nonspecifically to either control over specific events or control in life events overall; however, the latter is now the generally accepted interpretation of LOC.

The extent to which the construct of the LOC exists in animals is an unanswered question. Little direct evidence exists, and tests designed for determining LOC in children are not useful for animals as they ask the children whether they attribute causes of imaginary events to internal or external forces. Importantly, however, LOC is conceived as a generalized expectancy construct, meaning that it represents individuals’ expectations about receiving desired reinforcements (Rotter, 1966; Wallston, 2004). This in turn raises the consideration of another psychological attribute for which expectancy is a central element: optimism (Scheier and Carver 1992; Peacock and Wong 1996; Gruber-Baldini et al., 2009). Optimism consists of the expectation that good things, rather than bad things, will happen and is considered an aspect of personality (Bollini et al., 2004; Gruber-Baldini et al., 2009). While retaining certain conceptual differences, optimism and LOC have been found to be moderately related and overlapping concepts (Peacock and Wong, 1996). The primary difference is that LOC is the belief that one has control over desirable outcomes, whereas optimism is the expectation of positive outcomes regardless of whether or not they are a result of one’s own actions (Peacock and Wong, 1996). The trait of optimism/pessimism has been tentatively identified in animals over the past decade in the studies of cognitive and judgment bias (Rygula et al., 2015). Still often placed in quotation marks to avoid accusations of anthropomorphism, optimism and pessimism in animals describe a system of expectations analogous to the corresponding cognitive mechanisms in humans.

The optimism connection offers a plausible link between animals and LOC, but an additional connection derives from Rotter’s (1966, pp. 19–20) original proposal that the scale which measures internal versus external LOC ‘appears to measure a psychological equivalent of . . . the sense of powerlessness’. As we will see, powerlessness – or the perception of uncontrollability – has been extensively studied in animals.

However, because LOC as characterized in humans has not been definitively identified in animals, the use of this term in this chapter will be exclusive to humans. Thus, for animals as well as humans, the two types of control – that which is specific to a particular stimulus or event and that which applies to life in general – will be referred to as, respectively, event-specific control and general control (both terms will refer to perceived rather than actual control). General control may be considered in humans to be analogous to LOC.

6.3 The Desirability of Control

Human psychology research has found that the preference for control varies widely among individuals; some individuals appear to have a high desire for control while others prefer to wield less and even no control (Verme, 2009; Steptoe and Poole, 2016). In experimental studies it is not uncommon for some people, when given a choice between controllable and uncontrollable aversive events, to opt for the uncontrollable (Rodin, 1986). More recently, Buchanan-Smith and Badihi (2012) have identified in marmosets comparable individual preferences for making choices.

Notwithstanding the interindividual variance in the desire for control, research indicates that control is generally desired (Catania and Sagvolden, 1980; Overmier et al., 1980; Suzuki, 1999) and likely a psychological necessity in humans and animals (Leotti et al., 2010). In a series of experiments with rodents, Kavanau (1964, 1967) reported that captive rodents exercise control virtually any chance they get, apparently finding it rewarding to exert a high degree of control over their environment. If a light is turned on by the experimenter the animal will turn it off, but if the experimenter turns the light off, the animal, even though nocturnal, will turn it back on. If the experimenter turns on a motorized running wheel, the animal immediately and invariably turns the motor off. However, if the animal is running on such a wheel and the experimenter turns the motor off, the animal immediately turns the motor back on. If mice were awakened from sleep and emerged from their nest boxes, they would soon go back inside on their own. If, however, the experimenters picked the mice up and placed them in the nest box, they would immediately come back out, even if they were placed repeatedly back in the box. In all,

Earlier it was mentioned that one of the basic ways an organism can exert control is through choice. The desire for choice has been studied in several species. When deciding between two options, pigeons (Catania and Sagvolden, 1980) and macaques (Suzuki, 1999) preferred the option that led to a second choice over one that did not, even though both options ultimately led to the same outcome and making a second choice required greater energy expenditure. Rats expressed a preference for a pathway that led to a choice between additional pathways rather than a pathway leading directly to the reward, despite the fact that all pathways eventually led to the same reward (Voss and Homzie, 1970).

The desire for control also appears to at least partially explain the phenomenon of contra-freeloading, in which animals choose to work for an outcome even when the outcome is freely available (Franks and Higgins, 2012). As just one of many examples, Carder and Berkowitz (1970) reported that rats able to acquire free food from a dish would instead push the food dish aside in order to access a pressable lever that would cause a pellet of the same food to fall out for them to eat.

Collectively, these findings imply that the actual nature of the stimulus might not matter as long as the animal can exert some form of control over it and that exercising control can be rewarding in and of itself (Leotti et al., 2010). That the existence of the desire for control is found in many animal species as well as very young human infants led Leotti et al. (2010) to propose that this desire is innate and likely a psychological and biological imperative for survival.

6.4 The Relationship of Control to Mental Health and Well-being

In humans, an extensive theoretical and empirical literature has documented consistent positive correlations between personal control and each of the principal well-being constructs, including general well-being (Larson, 1989; Peterson, 1999), subjective well-being (Spector et al., 2001), emotional well-being (Thompson, 2005; Thompson and Kyle, 2000), and happiness (Larson, 1989; Peterson, 1999; Verme, 2009). Another well-being construct, life satisfaction, is very strongly associated with freedom of choice combined with one’s LOC (Verme, 2009). Most broadly, a general sense of control over one’s life appears to be a requirement for positive mental health (Larson, 1989).

Optimism, mentioned earlier in regard to its overlapping relationship to LOC, has been found in humans to be correlated with happiness (Verme, 2009). Similarly, optimism has been shown to have numerous beneficial effects on psychological well-being, including improved coping with stress and adversity (Scheier and Carver, 1992).

General effects of perceived control on well-being, happiness, and life satisfaction have been well researched in humans. However, due to limitations in animals of self-reporting and accurate assessment of long-term affective well-being states, evidence of a connection between control and well-being in nonhuman species has relied on the demonstration of indirect links between these two concepts, specifically, correlations between control and separate elements, which, taken together, comprise mental health and well-being. For instance, studies may focus on how control impacts the contributors and indicators of well-being, such as reduced stress, but not necessarily identify a direct impact on well-being itself. The evidence for this has emerged from two general areas: the positive effects of possessing control (see Section 6.5), and the negative effects of being denied control (discussed with the concept of learned helplessness, see Section 6.7).

6.5 Effects of Control in Animals

Determining the effects of a cognitive process in animals requires an experimental design that equates all variables between two animals except the cognitive process itself, which allows the conclusion that the observed differences between the animals may be attributed to that specific cognitive process. The classic experimental method for this purpose has been the yoked design, in which a pair of animals are exposed to the same aversive stimulus (e.g., electric shocks or loud noise). One animal, but not the other (yoked), is able to make a behavioral response which terminates the stimulus. The yoked animal thereby receives the same amount of the aversive stimulus as the first animal allows – or disallows – for itself, and any differences in outcomes (physiological or pathological) are presumed to result from the (non)availability of control (Steptoe and Poole, 2016).

Animals with perceived control over aversive events benefit from a diverse array of robust effects, such as a reduced intensity of physiological stress responses (autonomic reactivity and corticosteroid levels) (Joffe et al., 1973; Rodin, 1986), improved coping with a stressful situation (Seligman, 1975), an increase in calm activity (Buchanan-Smith and Badihi, 2012), and a greater ability to relax and function effectively even in potentially dangerous situations (Joffe et al., 1973). The perception of control allays the emotions of fear and anxiety (Solomon and Wynne, 1953; Sambrook and Buchanan-Smith, 1997) and may delay the onset of distress (Russell and Burch, 1959; Wolfensohn et al., 2018). Control appears to exert at least part of its effects through an active fear-inhibitory process such that when an animal processes that it has control over the aversive stimulus neural mechanisms are activated that reduce the generalization of fear, and the fear diminishes (Foa et al., 1992).

Cognitive functioning is also influenced by perceived control. When rats were given control over food or electric shock they subsequently exhibited enhanced learning in a shock avoidance task (Goodkin, 1976). When rats were able to control environmental lighting they performed significantly better over time in a discrimination task compared with rats without such control (Alliger and Moller, 2011).

6.5.1 How does control exert its beneficial effects?

The neuropsychological processes explaining how the perception of control exerts its beneficial effects are still far from understood (Thompson, 2005). Miller (1979) hypothesized that individuals who believe that they have control know that an aversive situation can be allayed at any time and thus can always be kept within the limits of what the individual can endure. The individual thereby feels comforted by the knowledge that he or she is protected against experiences becoming unbearable, allowing the individual to be relaxed and fully functional even in potentially threatening situations. From a different perspective, the benefits of control may derive from the ability to alter the various affective experiences of life. A prominent theoretical model of well-being in humans (Bradburn, 1969; Lyubomirsky et al., 2005) and animals (Broom, 2007; Green and Mellor, 2011; Yeates, 2011) is that well-being results from, or is otherwise associated with, a predominantly positive balance of pleasant feelings over unpleasant. In this view, control would exert its mental benefits by being a major factor in the animal’s (or person’s) ability to tip this balance in the desired direction. Overall, despite the gaps in knowledge of specific mechanisms it is clear that the perception of control in humans and animals provides broad and profound benefits for mental health and well-being.

6.6 Predictability

The question of why control should reduce the stress related to an aversive event has been approached differently by researchers who propose that the predictability of the event plays an integral role (Mineka and Hendersen, 1985; Foa et al., 1992). An event is unpredictable when its probability of occurrence is the same regardless of the events preceding it (Foa et al., 1992). An overlapping and complex interrelationship between control and predictability has been well-recognized, and several authors have questioned whether the two concepts are distinct and separable (Dess-Beech et al., 1983; Mineka and Hendersen, 1985; Foa et al., 1992).

The interrelationship of controllability and predictability can be summarized succinctly. When an animal is given the control to terminate a noxious event then the end of the aversiveness necessarily becomes predictable. However, predictable events may or may not be controllable (Mineka and Hendersen, 1985). In many studies, events that were uncontrollable also tended to be unpredictable, and those that were controllable were also predictable, making it difficult to discern which of the two factors was responsible for the observed effects (Foa et al., 1992; Bassett and Buchanan-Smith, 2007). Even so, some researchers have argued that the two are separable variables with distinguishable effects (Miller, 1979; Overmier et al., 1980) and in experiments that have separated controllability and predictability, controllability has been found to have effects over and above the predictability it provides (Thompson, 1981).

The effects of predictability (in some cases in combination with or indistinguishable from the effects of control) beneficial to well-being include general findings such as predictable aversive events (electric shock) are less physically and psychologically deleterious than unpredictable events (Dess-Beech et al., 1983; Bassett and Buchanan-Smith, 2007). Predictability effectively reduces stress (Seligman, 1968; Rodin, 1986) and unpredictability can be a cause of stress to animals (Taylor and Mills, 2007) as well as making aversive stimuli more stressful (Seligman, 1968; Sapolsky, 1994). For example, animals have more pronounced stress responses when aversive handling is unpredictable than when the same handling is conducted in a predictable way (Anonymous, 2001). Even in the absence of any stressor, the loss of predictability itself elicited a physiologic stress response (Sapolsky, 1994) in laboratory cats (Carlstead et al., 1993b) and dogs (Dess-Beech et al., 1983).

Fear and fear-related disorders are influenced by the degree of predictability. For example, unpredictable shock produces greater generalized fear and arousal than predictable shock, and danger unpredictability can directly produce increased generalized fear regardless of the degree of control (Foa et al., 1992). Studies using animal models of posttraumatic stress disorder (PTSD) have shown that the greater the degree of uncontrollability and unpredictability associated with a given stressor, the more likely the organism will be to develop symptoms of PTSD (Foa et al., 1992).

In contrast to the voluminous evidence of beneficial effects of predictability, some studies have indicated that predictable, but not unpredictable, events can exacerbate stress. Based on behavioral and physiological measures, stressors in some cases have been shown to be more aversive for animals if they are predictable (Dess-Beech et al., 1983; Mineka and Hendersen, 1985; Bassett and Buchanan-Smith, 2007).

The question as to how predictability exerts its beneficial effects parallels the same question for controllability. For some theorists, the benefits of one can be ascribed to the other. For example, Averill (1973) has argued that the positive outcomes of having control can be attributed to the predictability inherent in control, while others reduce the effects of predictability to the effects of added control that comes by allowing the organism to ‘prepare’ for the forthcoming stressful event and thereby modify its impact (Mineka and Hendersen, 1985). This might occur, for instance, when animals receiving predictable shocks would change their posture in order to minimize the experienced intensity of the shocks (Bassett and Buchanan-Smith, 2007). Another explanation for the benefits of predictability is that knowledge of when a stressful event is going to occur also informs the animal when the event is not going to occur, i.e., safe periods (Thompson, 1981; Bassett and Buchanan-Smith, 2007). This ‘Safety Signal’ hypothesis posits that the feedback about safe periods alleviates the chronic state of fear elicited by the uncertainty of when danger might occur. The animal able to predict the event can therefore relax and not feel the need to be on constant alert for the threat (Tsuda et al., 1983; Mineka and Hendersen, 1985). Sapolsky (1994, p. 258) illustrated this using a personal anecdote:

Stay updated, free articles. Join our Telegram channel

Join