Do I contradict myself?
Very well then I contradict myself,
(I am large, I contain multitudes.)
—Walt Whitman, Song of Myself (from Leaves of Grass, Reference Whitman1855/1960)1. Introduction
Why do people behave in such contradictory ways? Traditional psychological approaches have struggled to account for the problem of cross-situational inconsistency (e.g., Mischel & Shoda Reference Mischel and Shoda1995). Indeed, although the field has identified numerous factors that potentially influence a person's behavior – situational factors and individual differences, physiological need states and philosophies of life, conscious and unconscious mental processes, group norms and transcendent values – these influences are bound to conflict with one another, producing apparent inconsistencies in one's overt, observable behavior. These inconsistencies must be managed – rationalized and accounted for – in order to maintain the appearance of stability and predictability to others. The trust that others have in us is the most important determinant of their impressions of and liking for us (Fiske et al. Reference Fiske, Cuddy and Glick2007), and inconsistencies lead to the impression of being a “phony,” the most negatively viewed personality trait of all, out of 555 traits in Anderson's (Reference Anderson1968) normative ratings.
Here, we present the Selfish Goal model, which holds that these inconsistencies in judgment and behavior can be meaningfully understood as the output of multiple, and in some cases, competing goal influences. Whether conscious or unconscious, every goal essentially programs particular sets of behaviors to be enacted by the person pursuing that goal; however, a single person cannot physically express all of these goal influences simultaneously. Instead, he or she expresses many of these influences one at a time, thereby generating behaviors that appear inconsistent across extended time periods. In other words, contradictions in individual behavior may occur because a person, as Whitman acutely observed, contains multitudes of influences.
The inspiration for this target article is interdisciplinary, as we hope will be its appeal. Our core insight is derived from a prominent theory in evolutionary biology. In The Selfish Gene (1976), Richard Dawkins describes how every organism is composed of multiple genes, each of which can be seen as using that organism as a survival machine. Through the blind processes of natural selection, genes exert an influence on their host organism's behavior that maximizes their chances of propagation into future generations, sometimes to the detriment of the host organism's life. We believe it is useful to describe a person's psychological structure in an analogous fashion. Thus, we conceptualize each individual as comprised of multiple, oftentimes conflicting goals, each of which influences that person in a systematic, “selfish” manner.
As we will argue, human goal pursuit – whether operating consciously or unconsciously – constrains a person's information processing and behaviors in order to increase the likelihood that he or she will successfully attain that goal's end-state. These multiple, sometimes conflicting goals can produce different behaviors, judgments, and even self-representations in the same person that may appear inconsistent or contradictory across time, because they will vary as a function of which goal happens to be most active and motivating in the particular situation. Put another way, observed incoherencies in a person's actions may result because behavior is being selected (and is coherent) at a lower, less apparent goal level.
First, we trace the evolution of adaptive human behavior back to its unconscious and goal-driven roots. We survey evidence that an unconscious, less integrated system coordinated organism-level behavior prior to the evolution of conscious processing capabilities. When consciousness did emerge, we suggest that it was built upon preexisting unconscious processes, thereby producing the striking structural and phenomenal similarities observed between conscious and unconscious goal pursuits in recent psychological research.
Second, we generate predictions from this framework regarding the qualities and outcomes of human goal-driven behavior in the present day. If, as we assume, individual behavior arose from a system of unconscious and selfish processes that predated the evolution of consciousness, we should find contemporary signatures of the unconscious processes operating to guide behavior and judgment independently of conscious guidance (automaticity). To the extent that these unconscious systems evolved to facilitate adaptive responses to the environment, and were the proximate control system over those responses, they should be found to produce a “full-system” orientation toward specific end-states – adjusting perceptual, evaluative, cognitive, as well as motivational, parameters (reconfiguration) to optimize attainment of the desired end-state.
The present model also predicts that the evolution of consciousness from unconscious processes should be revealed in the observation of highly similar properties between the two denominations of goal pursuit (similarity). In addition to the straightforward prediction that unconscious goal pursuit should be characterized by similar processes and outcomes as revealed by a century of research on conscious goal pursuit, The Selfish Goal model makes the nonobvious prediction that conscious goal pursuit also should be found to operate in ways only recently discovered to be true of unconscious goal pursuit. That is, even conscious goals should be found to operate selfishly, even if doing so produces behavioral inconsistencies for the individual (inconsistency). Recent findings from social and developmental psychology, as well as neuroscience, will be marshaled in support of each of these principles.
A model such as this, with interdisciplinary theories and implications, is likely to encounter challenges such as establishing shared definitions and drawing from areas of relative inexpertise. Although we acknowledge these challenges, we believe that our theoretical contribution brings important recent developments from social psychology to bear in discussions with other disciplines that share the goal of understanding individual human behavior.
2. Who – or what – is in control of an individual's behavior?
We all share the intuition that the form and content of our behaviors are produced through conscious, intentional choice and internal processes of which we are aware and able to report on reasonably accurately to others (Wegner Reference Wegner2002). In the past, traditional psychological approaches to human motivation similarly assumed an agentic, conscious self at the helm, deliberately forming judgments, making decisions about which courses of action to take, and then guiding one's behavior along those intentional lines (e.g., Ajzen & Fishbein Reference Ajzen and Fishbein1980; Bandura Reference Bandura1986; Baumeister et al. Reference Baumeister, Bratslavsky, Muraven and Tice1998; Locke & Latham Reference Locke and Latham1990; Mischel Reference Mischel1973). In some prominent models, conscious choice of behaviors or goals to pursue was conceptualized as a bottleneck – nothing happened without one's awareness and conscious consent (e.g., Bandura Reference Bandura1986; Locke & Latham Reference Locke and Latham2002, p. 705).
Since then, three modern lines of research have cast doubt on the basic assumptions of conscious-centric control models. First, studies increasingly highlight the power of situational variables in determining behavior, including external influences that override internal sources of control such as self-values and personality (e.g., Darley & Latane Reference Darley and Latane1968; Milgram Reference Milgram1963; Mischel Reference Mischel1973; Ross & Nisbett Reference Ross and Nisbett1991). Second, research on the limits of introspective access demonstrates that people are often unaware of the reasons behind their actions and the actual sources of their evaluations and subjective feelings about the external world (Bar-Anan et al. Reference Bar-Anan, Wilson and Hassin2010; Nisbett & Wilson Reference Nisbett and Wilson1977; Wilson & Brekke Reference Wilson and Brekke1994) – access they would be expected to have if they were consciously aware of making those choices and deciding what to do. Third, “dual-process” models (e.g., Chaiken & Trope Reference Chaiken and Trope1999; Morewedge & Kahneman Reference Morewedge and Kahneman2010; Posner & Snyder Reference Posner, Snyder and Solso1975; Strack & Deutsch, in press) hold that external situational influences often operate in an automatic and implicit fashion to directly instigate the higher mental processes involved in information processing and behavior, thereby bypassing the consciousness bottleneck and eliminating the need for an agentic “self” in the selection of all behavioral and judgmental responses (Bargh Reference Bargh2007; Bargh et al. Reference Bargh, Schwader, Hailey, Dyer and Boothby2012).
Research involving the passive activation or priming of higher-order concepts by contextual features combines the situational and automatic emphases of contemporary research. Priming studies have consistently demonstrated that the mere exposure to environmental events is sufficient to directly trigger higher mental processes, in the absence of any conscious intentions or awareness that they operate (see Dijksterhuis et al. Reference Dijksterhuis, Chartrand, Aarts and Bargh2007; Higgins Reference Higgins, Higgins and Kruglanski1996). Unconscious processes have been shown to produce evaluation and social judgment (Fazio Reference Fazio, Sorrentino and Higgins1986; Ferguson Reference Ferguson2008), stereotyping and prejudice (Devine Reference Devine1989), social behavior (Bargh et al. Reference Bargh, Chen and Burrows1996; Dijksterhuis & van Knippenberg Reference Dijksterhuis and van Knippenberg1998; Loersch & Payne Reference Loersch and Payne2012), and goal pursuit (Chartrand & Bargh Reference Chartrand and Bargh1996; Custers & Aarts Reference Custers and Aarts2005; Marien et al. Reference Marien, Custers, Hassin and Aarts2012), with research participants unaware of the role played by these external stimuli in their decisions and behaviors.
Although important strides have been made, key issues remain. The environment is a rich source of multiple, oftentimes simultaneous cues – many of which are linked to different, often competing behavioral impulses. How are these parallel environmental influences funneled into the necessarily serial behavior of a single individual (the “reduction problem”: Bargh Reference Bargh2006; Merker Reference Merker2007)? An important aim of the present model is to address this question – namely, to argue that the currently active goal is the proximate controller of present behavior and judgment, overriding other conflicting influences in a “single-minded” fashion if necessary to facilitate goal completion.
2.1. The Selfish Goal
To illustrate how behavior selection might occur at a less-centralized level than that of the individual person (where it is overtly expressed), we describe how goals direct and constrain a person's behavioral possibilities as if the goals themselves were selfish.
The Selfish Goal model offers general predictions regarding individual behavior in the current day. The automaticity principle proposes the existence of unconscious processes tied to individual judgment and behavior. As we will review, research has indeed discovered psychological processes that are dissociated from individual awareness or guidance, yet can change how that person sees the world and behaves in response (Bargh et al. Reference Bargh, Schwader, Hailey, Dyer and Boothby2012). For example, priming (passively and temporarily activating) an individual's internal goal representation affects subsequent judgments and behaviors in a manner consistent with him or her being in a motivated state (Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001; Bargh et al. Reference Bargh, Gollwitzer, Oettingen, Fiske, Gilbert and Lindzey2010; Dijksterhuis & Aarts Reference Dijksterhuis and Aarts2011).
The second principle, reconfiguration, holds that changes in a person's judgments and behaviors during goal operation occur in order to optimize that person's chances of completing the goal. In particular, we focus on experiments featuring unconscious forms of pursuit, because by definition participants are unaware of the goal and hence cannot guide their own perceptions and behavior in a goal-congruent way. Nevertheless, research suggests that people who are unaware that they are pursuing a goal respond to the world in a way that maximizes the likelihood of goal completion, such as by paying more attention to objects in the environment that would assist with goal pursuit and becoming predisposed to like and physically approach those objects. Goals operate autonomously (i.e., independent of guidance from the conscious individual) through these mechanisms to encourage achievement of their associated end-states.
Because all goals are selfish, the similarity principle holds that conscious goal pursuit should resemble its unconscious counterpart in these regards. Recent research reveals a number of striking similarities between conscious and unconscious goal operation, from shared neural correlates and subprocesses to similar autonomous effects upon individual judgment and behavior. Even when a person consciously engages in goal pursuit, one is not necessarily controlling (or even aware of) how that goal has transformed one's experience of the world. As an analogy, when conversing with another person, we are not aware of how our ideas and thoughts are being transformed into grammatical and coherent sequences of words.
The inconsistency principle proposes that as multiple goals within a single individual become active, operate, and turn off, the person pursuing those goals may appear to be acting inconsistently, or in a manner that seems contrary to his or her interests. This is particularly evident at the stage of goal completion, where the mental representation of a goal becomes less accessible but, ironically, the pursuer becomes more likely than usual to exhibit behaviors that contradict the recently completed goal. We will cast these lines of research in the light of evolutionary theories to argue that the proximate control of individual behavior within a situation rests with the currently active goal, and if more than one goal is currently active, then with the most incentivized (important) of these.
2.2. Terminology
Before proceeding, it is important to make clear what we mean by some commonly used terms and concepts. A goal is a construct of central importance in psychology, yet little consensus exists regarding its precise definition (Austin & Vancouver Reference Austin and Vancouver1996; Elliot & Fryer Reference Elliot, Fryer, Shah and Gardner2008). For example, goals are sometimes equated with related concepts such as needs, motives, biological goals, and drives (e.g., Austin & Vancouver Reference Austin and Vancouver1996; Pervin Reference Pervin1989;); others draw distinctions (e.g., Gollwitzer & Moskowitz Reference Gollwitzer, Moskowitz, Higgins and Kruglanski1996). Some perspectives conceptualize goals as explicit standards for behavior set and regulated by the individual (e.g., Bandura Reference Bandura1986; Locke & Latham Reference Locke and Latham1990); others suggest that they can be triggered by environmental contexts and operate independently of individual awareness through a combination of cognitive and affective mechanisms (e.g., Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001; Custers & Aarts Reference Custers and Aarts2010; Kruglanski et al. Reference Kruglanski, Shah, Fishbach, Friedman, Chun, Sleeth-Keppler and Zanna2002).
Here, we follow the standard social-cognitive definition of goals as mental representations of desired end-states (Aarts & Dijksterhuis Reference Aarts and Dijksterhuis2000; Bargh Reference Bargh, Higgins and Sorrentino1990; Fishbach & Ferguson Reference Fishbach, Ferguson, Kruglanski and Higgins2007; Kruglanski et al. Reference Kruglanski, Shah, Fishbach, Friedman, Chun, Sleeth-Keppler and Zanna2002). We are less interested in long-term, chronic goals (e.g., raising children; sometimes referred to as “life tasks”; Cantor & Blanton Reference Cantor, Blanton, Bargh and Gollwitzer1996; Pervin Reference Pervin1989) and more interested in time-limited instrumental behaviors enacted in the current situation (e.g., being helpful to others; satisfying one's hunger). Our focus is less on the very concrete actions that can be described in fully objective terms (e.g., pressing a button; opening a door) and more interested in higher-level end-states that provide those actions with meaning and that come in both conscious and unconscious denominations.Footnote 1 A person can reach the same end-state by “operating” on the environment in different ways (e.g., one can be helpful by taking out the trash or by uttering supportive statements); the notion that Skinner (Reference Skinner1953, p. 65) captured with the concept of an operant, or a “set of acts” defined by their effect, Kruglanski et al. (Reference Kruglanski, Shah, Fishbach, Friedman, Chun, Sleeth-Keppler and Zanna2002; also Austin & Vancouver Reference Austin and Vancouver1996) termed equifinality and Lashley (Reference Lashley and Kluver1942) referred to as motor equivalence, in which a motoric action goal can be realized in a variety of ways.
Historically, there has been high consensus regarding the features of conscious information processing: a person is aware of those processes and intends them to occur; they operate in a serial fashion, are somewhat resource limited at any given time, and are controllable (e.g., Bargh Reference Bargh, Wyer and Srull1984; Reference Bargh, Wyer and Srull1994; Posner & Snyder Reference Posner, Snyder and Solso1975). In contrast, views of the unconscious have arisen out of the study of processes that do not possess all of these features (see Bargh & Chartrand, Reference Bargh, Chartrand, Reis and Judd2000). For example, “preconscious” or “preattentive” processes operate automatically upon encountering stimuli in the environment and feed subsequent conscious processes (Neisser Reference Neisser1967; Posner & Snyder Reference Posner, Snyder and Solso1975), whereas other unconscious processes become automatic through a process of skill acquisition (e.g., Shiffrin & Schneider Reference Shiffrin and Schneider1977) given the intent to engage those processes in the first place, such as when typing or driving.
By unconscious, we refer to information-processing events in the human nervous system that, although capable of influencing a person's behaviors, emotions, cognitions, and motivations, “do not influence subjective experience in a way that [he or she] can directly detect, understand, or report the occurrence or nature of these events” (Godwin et al. Reference Godwin, Gazzaley, Morsella, Pereira and Lehmann2013; Morsella & Bargh Reference Morsella, Bargh, Decety and Cacioppo2011; Nisbett & Wilson Reference Nisbett and Wilson1977). Note that by this definition, one can be consciously aware of a stimulus event but unconscious of its influence on oneself (see Bargh Reference Bargh, Bornstein and Pittman1992).Footnote 2 A process is described here as unconscious even if a person is able to report the presence of a stimulus and respond to it (e.g., in an experiment, one is judging stockings presented in an array, deems the right-most choice of highest quality, and provides reasons why) but lacks awareness regarding the causes and processes underlying the effect (e.g., that one may have been biased by a position effect on evaluation that favored the right-most stocking independent of quality-related factors; Nisbett & Wilson Reference Nisbett and Wilson1977).
This use of the term “unconscious” highlights the unintentional nature of the process, along with a lack of awareness of its underlying causes and processes (see Bargh Reference Bargh, Bornstein and Pittman1992; Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001). This definition is consonant with uses of the term “unconscious” in both evolutionary theory and historically by psychologists. For example, Darwin (Reference Darwin1859), one of the earliest scientists to refer to unconscious processes, employed the term when describing how the farmers and stock breeders of his time made implicit use, without explicit awareness, of the laws of natural selection in order to produce larger ears of corn and fatter sheep. Similarly, Dawkins (Reference Dawkins1976) wrote of nature as the “blind watchmaker, the unconscious watchmaker” to signify that there was no agentic, intentional guiding hand in producing these adaptive designs (see also Bargh & Morsella Reference Bargh and Morsella2008; Buss et al. Reference Buss, Haselton, Shackelford, Bleske and Wakefield1998; Dennett Reference Dennett1991; Reference Dennett1995). According to Brill (Reference Brill and Brill1938), Sigmund Freud credited Mesmer and the early hypnotists with the original discovery of the unconscious when they caused their subjects, through post-hypnotic suggestion, to behave in ways that they did not intend.
3. The primacy of the unconscious
Why individuals behave as they do within given situations is a fundamental question of social psychology (Aronson Reference Aronson1995), and social-psychological answers to this question should, ideally, be in harmony with relevant knowledge from the natural sciences (Pinker Reference Pinker1997). Accordingly, we describe literature on consciousness and evolutionary biology that provide the theoretical backbone of the Selfish Goal. The following section is not intended as a comprehensive review of either topic, which would be outside of the scope of the present article; rather, areas of consensus are highlighted because they provide central guiding insights.
3.1. Prior to the evolution of consciousness
Understanding how consciousness evolved remains one of the greatest scientific mysteries (Kennedy Reference Kennedy2005). For certain portions of the puzzle, however, consensus is beginning to emerge (for review, see Godwin et al. Reference Godwin, Gazzaley, Morsella, Pereira and Lehmann2013), and these areas of agreement offer insight regarding present-day individual behavior. For example, few experts would object to the claim that, at some point in the evolutionary past, something unconscious became something that was conscious (e.g., Corballis Reference Corballis, Zelazo, Moscovitch and Thompson2007; Deacon Reference Deacon1997; Dennett Reference Dennett1991; Donald Reference Donald1991; Godwin et al. Reference Godwin, Gazzaley, Morsella, Pereira and Lehmann2013; Macphail Reference Macphail1998). Much of the activity in the brain and in the rest of the nervous system operates unconsciously, reinforcing the argument that consciousness is not an inherent property of human cognition and likely emerged from an evolutionary landscape that had until then been the province of “not-conscious,” or unconscious processes.
Moreover, the phenomena linked to consciousness are themselves associated with but a subset of neuroanatomical regions in the central nervous system (e.g., Dehaene & Naccache Reference Dehaene and Naccache2001; Morsella et al. Reference Morsella, Krieger and Bargh2009). Identifying which particular regions are responsible for consciousness remains an issue of active debate, as once-prevalent views of the cerebral cortex as the “organ of consciousness” face direct challenge by theories and evidence linking subcortical structures such as the upper brainstem to the instantiation of conscious states (e.g., Godwin et al. Reference Godwin, Gazzaley, Morsella, Pereira and Lehmann2013; Merker Reference Merker2007; Penfield & Jasper Reference Penfield and Jasper1954).
The evidence that only a subset of processes and regions of the brain are associated with consciousness, plus the fact that humans share much of this unconsciously operating nervous system with earlier-evolving members of the animal kingdom (some of whom arguably lack consciousness), leads to the conclusion that conscious processes are a phylogenetically later adaptation of the brain. As Dennett (Reference Dennett1991, p. 171) pointed out, “Since there hasn't always been human consciousness, it has to have arisen from prior phenomena that weren't instances of consciousness.”
One can infer additional characteristics of the unconscious system guiding behavior. The existing system may have been unconscious and perhaps less centralized in the absence of consciousness, but it was nevertheless capable of guiding an organism's behavior toward adaptive outcomes. Borrowing an example from Merker (Reference Merker2007), there is little reason to believe that the cubomedusa jellyfish, with its noncephalized nervous system, experiences consciousness (i.e., that there is anything it is “like” to be this jellyfish). Yet, despite the fact that it lacks a brain even remotely like a human's, the jellyfish still exhibits adaptive behaviors, including flexible, goal-driven movements in response to environmental stimuli such as prey.
Put another way, although consciousness is traditionally associated with behavior selection and coordination, this does not mean the older, unconscious, and less-centralized system(s) did not also give rise to adaptive, streamlined behavior at the level of the individual organism. (For examples of sophisticated unconscious control and conflict management, see Morsella Reference Morsella2005; Suhler & Churchland Reference Suhler and Churchland2009.) Indeed, evidence of behavior directed toward adaptive ends is omnipresent in the animal kingdom – even in creatures not assumed to possess consciousness – so much so that it is, in the view of one evolutionary biologist, “perhaps the most characteristic feature of the world of living organisms” (Mayr Reference Mayr1976, p. 389).
Many behaviors that serve the ultimate “end” goal of all adaptations (i.e., differential reproduction) are also goal-driven on a more proximate level. For example, when Dawkins (Reference Dawkins1976) identified “apparent purposiveness” as an important property of adaptive behavior, he clarified, “By this I do not just mean that it seems to be well calculated to help the animal's genes to survive, although of course it is. I am talking about a closer analogy to human purposeful behavior … [organisms] that behave as if motivated by a purpose” (p. 50). Indeed, evolutionary biologists and psychologists alike link motivations or goals to adaptive outcomes (Campbell Reference Campbell and Schilpp1974; Cosmides & Tooby Reference Cosmides and Tooby2013; Kenrick et al. Reference Kenrick, Griskevicius, Neuberg and Schaller2010; Pinker & Bloom Reference Pinker and Bloom1990; Popper Reference Popper1972; Symons Reference Symons, Barkow, Cosmides and Tooby1992; Tetlock Reference Tetlock2002; Tomasello et al. Reference Tomasello, Carpenter, Call, Behne and Moll2005; Tooby & Cosmides Reference Tooby, Cosmides, Barkow, Cosmides and Tooby1992). A bird starting on its migration, a male displaying to a female, a predator stalking its prey, and the prey's subsequent flight – all are acting toward a specific end-state in the local environment (e.g., reaching safety and shelter; obtaining food to eat).
Goal-related constructs are prominently featured in models of adaptive human behavior, further rooting these behavioral patterns deep in the evolutionary past (Bugental Reference Bugental2000; Cosmides & Tooby Reference Cosmides and Tooby2013; Kenrick et al. Reference Kenrick, Griskevicius, Neuberg and Schaller2010; Neuberg et al. Reference Neuberg, Kenrick, Maner, Schaller, Forgas and Williams2004). Similar to its use in social cognitive literature, these goal constructs involve “if-then” contingencies in which environmental cues trigger psychological and behavioral phenomena that are designed by natural selection for a purpose – to facilitate adaptive outcomes. Accordingly, different goals have their own specific evolutionary function (e.g., attaining safety; securing a mating partner) and are associated with different decision rules, neuroanatomical regions, and behavioral responses. An organism that suddenly encounters predators should be most likely to fight, flee, or freeze; that same creature, in the company of potential mates, should be compelled to act differently by approaching and displaying instead.
Although research suggests that adaptive behavior across the animal kingdom can be organized into goal categories, it is less clear as to how multiple behavioral impulses – each designed for specific functions – are funneled into streamlined behavior. Evolutionarily adaptive goals require different, often incompatible organism-level behaviors for its pursuit (e.g., fleeing vs. eating), which must be expressed in serial fashion at critical moments. Kenrick et al. (Reference Kenrick, Griskevicius, Neuberg and Schaller2010, p. 303) emphasized this tension as a “continual interplay between motivational systems and the perception of affordances (fitness-relevant threats and opportunities) in the immediate environment.” What, then, are the characteristics of an evolutionarily ancient system that could integrate multiple, sometimes competing influences into overt behavior expressed at the level of the individual organism – particularly in the absence of overarching conscious processing to integrate and prioritize these goals?
3.2. The selfish gene and the Selfish Goal
Understanding the component pieces of the puzzle offers a preliminary step toward an answer, and evolutionary biology provides a useful metaphor with which to describe their operation. As previously stated, in The Selfish Gene, Dawkins (Reference Dawkins1976) described how genes, through the blind process of natural selection, influence the design of their host organism in order to maximize their chances of propagation into future generations. This theory characterizes genes as essentially “selfish” in that their only concern is their own replication, not the welfare of their host organism, except as it might impact on replication. (Note that this selfish influence on one level can also be seen as cooperative at another level, as Dawkins [Reference Dawkins1976] pointed out, insofar as multiple genes must coexist and be expressed within a single individual.)
We argue that the dynamic between genes and their host organism is analogous to the relationship between goals and the individual pursuing them. Every organism is comprised of multiple genes, all using that organism as their survival machine into the next generation. Similarly, a person comprises many different goals, each of which operates on that individual to produce successful pursuit of a specific end-state.
Supporting the notion that unconsciously operating goal processes are capable of influencing individual-level outcomes, experiments from both evolutionary psychology and social cognition highlight early-stage orienting mechanisms (e.g., selective attention and perception) that serve as “building blocks” for subsequent human behavior (Balcetis & Dunning Reference Balcetis and Dunning2006; Maner et al. Reference Maner, DeWall and Gailliot2008; Neuberg et al. Reference Neuberg, Kenrick, Maner, Schaller, Forgas and Williams2004). Given that a person is effectively steered toward particular sets of actions (and simultaneously away from others), these processes can be seen as operating in self-interested ways, as a gene encodes organism-level behavior that reliably promotes replication of that gene, and not necessarily the health or well-being of the host organism itself.
3.3. After the evolution of consciousness
Although little consensus exists regarding how consciousness evolved, even the basic claim that consciousness evolved in a nervous system comprosed of unconscious, less centralized processes can be informative. Evolutionary theorists note that given the constraints of building costs and materials, evolution behaves as a “tinkerer,” modifying existing structures in a gradual, incremental fashion, instead of creating entirely new ones each time from scratch (Allman Reference Allman2000; Dawkins Reference Dawkins1976; Jacob Reference Jacob1977). Solutions created from this process may not be perfect, but nevertheless satisfice (Dennett Reference Dennett1995; Simon Reference Simon1956).
As the process of natural selection uses old elements and alters them for new functions, traces of the original structure and function often remain in the end-product. Evidence of this co-option process can also be found across the animal kingdom, as when the vibration-sensitive reptilian jaw evolved into mammalian middle-ear bones, fish fins morphed into forelimbs for land-dwelling creatures, and the esophagi of freshwater fish expanded to serve terrestrial lung functions (for more examples, see Jacob Reference Jacob1977; Pinker Reference Pinker1997).
The human nervous system also bears marks of evolutionary tinkering (e.g., Anderson Reference Anderson2010; Buss et al. Reference Buss, Haselton, Shackelford, Bleske and Wakefield1998; Jacob Reference Jacob1977; Morsella Reference Morsella2005), which suggests that conscious processes were likely shaped by the preexisting unconscious behavioral control system (see Bargh & Morsella Reference Bargh, Morsella, Agnew, Carlston, Graziano and Kelly2010). One would expect to see, as one does, that conscious states are selectively integrated into the existing unconscious system, while many behavioral processes remain unaffected. Conscious processes should also retain features of the preexisting elements integrated into its design. Consequently, one might also expect conscious processes to function in a similarly “selfish” manner as do their unconscious counterparts, because both operate on the nervous system in the service of adaptive, individual-level behavior.
3.4. Summary
Inherent in any attempt to integrate fields is the risk of drawing too many assumptions. For this reason, the Selfish Goal relies on a limited number of claims about the evolution and function of consciousness: (1) the relatively late evolutionary emergence of consciousness, leading to the further assumption that (2) some other, unconscious, less centralized system(s) must have been guiding serially expressed, adaptive behavior prior to that point in evolutionary history. Finally, we assume a basic principle of evolution through natural selection, that (3) it occurs through gradual changes to the existing machinery (e.g., Allman Reference Allman2000), and not via dramatic, complete overhauls of that machinery. In other words, less integrated unconscious goal processes lay down the landscape upon which their conscious counterparts operate today.
Given the theoretical evidence to support these assumptions, we surmise that at a hypothetical Time 1, a less centralized system of unconscious, adaptive processes, including goals, drove behavior within the same biological entity. At Time 2, consciousness evolved within the nervous system and was likely shaped in structure and function by its integration with the preexisting unconscious landscape. By theoretical Time 3 (the present day), we see events indicating further development of conscious processes (e.g., expansion of cortical matter; metacognitive abilities; representation of self; and tighter executive control over the initial impulses to action). Nonetheless, some original unconscious processes from Time 1 remain, operating in concert with and independent of later arriving conscious processes.
In the next section, we consider the implications of this model for behavior in current-day contexts. Evidence for four principles will be offered in support of the claim that a person's judgments and behaviors can be meaningfully examined from the perspective of which goals he or she is currently pursuing. In so doing, we hope to clarify how inconsistencies in a person's behavior might result from a behavioral guidance system comprised of selfishly acting goals.
4. The active goal as proximal cause of human behavior
My thinking is first and last and always
for the sake of my doing,
and I can only do one thing at a time.
—William James (Reference James1890/1981)Figuratively put, the Selfish Goal model posits a system wherein multiple goals engage in the selection of behavior while the individual person is actually “expressing” it. As multiple conscious and unconscious goals operate, each steering the individual toward specific (and oftentimes conflicting) end-states, inconsistencies in individual behaviors will arise. Note that goals often can and do encourage behaviors that are consistent with (or at least not opposed to) other goals' end-states, or one's consciously preferred outcomes (e.g., exercising self-control by declining dessert is consonant with a dieter's self-reported preferences). We focus on telling instances when the two interests are inconsistent or at odds with each other.
In extreme cases, the tension between the behavioral imperatives issued by the currently active goal and the other priorities of the person pursuing that goal (over time and across situations) can produce trade-offs between what is “good” for the goal being pursued versus what is “good” for the individual. This dynamic is most evident in addictions (e.g., Baker et al. Reference Baker, Piper, McCarthy, Majeskie and Fiore2004) where the addict reports feeling helpless to resist the urge to consume the drug and may engage in self-destructive behavior (as well as behavior that typically contradicts his or her important self-values, such as lying and stealing) in order to acquire the drug. Decades of research on drug abuse substantiate the similarity between addictions and more standard goal pursuits (e.g., Loewenstein Reference Loewenstein1996), as expressed colloquially by the phrase, “addiction hijacks the motivational system.”
The pursuit of everyday goals has “selfish” effects as well and may cause an individual to desire things that he or she might not have wanted if not actively pursuing the goal. For example, young women primed with the mating goal express more positive attitudes toward and stronger intentions to engage in attractiveness-enhancing yet dangerously unhealthy behaviors such as spending time in tanning booths and taking diet pills (Hill & Durante Reference Hill and Durante2011). Those behaviors may facilitate the currently active goal of mating (by increasing one's sex appeal) but operate to the long-term detriment of the individual. Indeed, participants' attitudes toward these behaviors when the mating goal was not currently active were considerably more negative. In a similar vein, males who are unconsciously pursuing status goals express greater hostility to other males (Griskevicius et al. Reference Griskevicius, Tybur, Gangestad, Perea, Shapiro and Kenrick2009). In this case, increased willingness to aggress furthers the active goal (by physically intimidating one's competition) but ultimately jeopardizes the health of the individual pursuing that goal, both during pursuit and afterwards.
The Selfish Goal perspective is in harmony with other motivational models in social psychology, particularly the Evolved Hierarchy of Needs (Kenrick et al. Reference Kenrick, Griskevicius, Neuberg and Schaller2010) and Functionalist Social-mindset theory (Tetlock Reference Tetlock2002). We take a complementary perspective to Kenrick et al. (Reference Kenrick, Griskevicius, Neuberg and Schaller2010) by focusing on the general structure of goal influence, as opposed to domain-specific content, and offering more proximate explanations for behavior rather than the ultimate causes of it (see Killeen Reference Killeen2001). Tetlock's (Reference Tetlock2002) theory posits the existence of evolved complex mindsets associated with different social motives depending on the individual's current situation and contextual role. When perceived as misbehaving according to group norms, an individual adopts a “defense attorney” mindset that focuses on external reasons for the socially negative action in order to maintain or restore one's good standing in the group; however, the same individual faced with another person's misbehavior adopts a very different “prosecutorial” mindset, with the opposing goal of holding the miscreant's feet to the fire in order to uphold those important group norms.
This model is also inspired by extensive research on Goal Systems Theory (GST; Kruglanski et al. Reference Kruglanski, Shah, Fishbach, Friedman, Chun, Sleeth-Keppler and Zanna2002). GST holds that goals are mental representations with cognitive and motivational properties and exist within systems of interconnected goals. Within this goal system, limited cognitive resources are distributed in a constant-sum fashion, which can lead to mental resources being pulled toward and away from particular goals, depending on which is active. For example, the activation of a goal to seek entertainment or to eat a yummy calorie-rich food (i.e., temptations) can trigger self-regulatory goals to get some work done or to eat something healthy instead (Fishbach et al. Reference Fishbach, Friedman and Kruglanski2003). We similarly highlight conflict among goals. Whereas GST attributes this tension to limited cognitive resources, our evolutionarily inspired model focuses on conflict within an evolved system that must funnel multiple impulses operating in parallel into streamlined behaviors within a single physical body.
4.1. Predictions from the present model
In section 3 we reviewed theories that hold that unconscious, selfish goals lay the evolutionary foundation for the operation of their conscious counterparts. Specifically, the Selfish Goal model holds that multiple, selfish goals indirectly guide the streamlined behaviors expressed by a single individual. This framework predicts the following about individual judgment and behavior in current-day contexts:
-
1. The automaticity principle. Unconscious processes can influence behavior in the absence of individual awareness or guidance.
-
2. The reconfiguration principle. The most motivating or “active” goal should constrain the individual's information processing and behavioral possibilities in a way that encourages achievement of the goal's end-state.
-
3. The similarity principle. Conscious goal pursuit should recruit similar processes and produce similar outcomes as unconscious goals.
-
4. The inconsistency principle. Temporarily active goals can produce outcomes for the individual that appear inconsistent over time or contrary to the individual's interests.
4.2. The automaticity principle
We will first review evidence for the existence and qualities of unconscious processes that are tied to overt human behavior yet dissociated from individual awareness. Dawkins (Reference Dawkins1976) notes that adaptive behavior in the animal kingdom “works on a time-scale not of months but of seconds and fractions of seconds. Something happens in the world … in milliseconds nervous systems crackle into action, muscles leap, and someone's life is saved – or lost” (p. 55). The past quarter century of psychological research documents similar, lightning-fast psychological mechanisms in humans (e.g., automatic evaluations and preferences, automatic influences of the perceived environment on behavioral responses, and an automatic mode of goal pursuit; Bargh et al. Reference Bargh, Schwader, Hailey, Dyer and Boothby2012; Bargh & Ferguson Reference Bargh and Ferguson2000; Greenwald & Banaji Reference Greenwald and Banaji1995), which were observed through methodological advances such as sequential priming and the Implicit Association Test. For example, the affective sequential priming task allows researchers to measure the implicit value of a target (and its associated goal) in terms of the extent to which it automatically facilitates response time to positive or negative concepts. Because less time is needed to categorize a target stimulus as positive or negative when it is preceded by a concept of the same valence (but not otherwise semantically related; e.g., Fazio Reference Fazio, Sorrentino and Higgins1986), researchers using this task are able to examine attitude activation and influence occurring outside of the individual's awareness and control.
Automatic evaluations play an important role in guiding both cognition (Ferguson et al. Reference Ferguson, Bargh and Nayak2005; Niedenthal Reference Niedenthal1990; Cantor & Blanton Reference Cantor, Blanton, Bargh and Gollwitzer1996) and overt behavior (the latter supporting the idea that these mechanisms may have served adaptive roles in the past; Roe & Simpson Reference Roe and Simpson1958). For example, implicit positive evaluations automatically create approach behavioral dispositions (muscular readiness) toward stimuli and negative automatic evaluations to avoidance (withdrawal) behaviors (Chen & Bargh Reference Chen and Bargh1999). Indeed, Ferguson (Reference Ferguson2008) revealed that positive automatic evaluations toward cooperation-related primes predicted participants' actual helping behaviors on a subsequent task (e.g., volunteering time to assist another person). Moreover, this link between automatic evaluation and muscular readiness has recently been successfully exploited in therapeutic techniques for the treatment of addictions, with patients making incidental avoidance arm movements in response to addiction-related stimuli across hundreds of trials, which has the consequence of significantly reducing their cravings and use of the substance (Wiers et al. Reference Wiers, Houben, Roefs, de Jong, Hofmann, Stacy, Gawronski and Payne2010).
4.2.1. Conscious awareness or intention is not required for goal pursuit
In one of the initial experimental demonstrations of unconscious goal pursuit, Bargh and colleagues (Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001; see also Chartrand & Bargh Reference Chartrand and Bargh1996) found that merely exposing participants to words related to cooperation (e.g., help, assist) in the guise of an ostensible language test increased cooperative behavior compared to another condition in which participants were explicitly instructed to cooperate. Following the experimental task, participants were explicitly asked the extent to which they had tried to cooperate during the task. In the explicit cooperation condition, these estimates correlated positively with actual cooperative behaviors (indicating some degree of conscious awareness of the reason for their behavior in the task). The estimates provided by participants who had been unconsciously primed, however, were uncorrelated with their actual cooperative behaviors, suggesting that they had been pursuing the cooperation goal without knowing they were doing so at the time.
Other research has consistently linked goal priming with downstream effects on judgment and behavior that are characteristic of a person in a motivated state. Phenomenal qualities once considered exclusive to conscious, deliberate goal pursuit, such as persistence in the face of obstacles, increase in goal strength over time, and changes in mood and goal strength depending on the fate of the goal attempt (Atkinson & Birch Reference Atkinson and Birch1970; Bandura Reference Bandura1986; Gollwitzer & Moskowitz Reference Gollwitzer, Moskowitz, Higgins and Kruglanski1996; Goschke & Kuhl Reference Goschke and Kuhl1993; Heckhausen Reference Heckhausen1991; Lewin Reference Lewin1926) have been demonstrated as well for unconsciously operating goals (e.g., Aarts et al. Reference Aarts, Custwers and Holland2007; Bargh et al. Reference Bargh, Schwader, Hailey, Dyer and Boothby2012; Bargh et al. Reference Bargh, Gollwitzer, Oettingen, Fiske, Gilbert and Lindzey2010; Bongers et al. Reference Bongers, Dijksterhuis and Spears2009; Custers & Aarts Reference Custers and Aarts2005; Custers et al. Reference Custers, Maas, Wildenbeest and Aarts2008; Dijksterhuis & Aarts Reference Dijksterhuis and Aarts2011; Ferguson Reference Ferguson2008; Förster et al. Reference Förster, Liberman and Friedman2007).
In their review of goal-priming studies, Custers and Aarts (Reference Custers and Aarts2010) noted that the traditional determinants of whether a conscious goal is pursued, attainability and desirability (e.g., Austin & Vancouver Reference Austin and Vancouver1996), also increase the chances that an unconscious goal will be pursued. First, priming the mental representation of a goal activates the skeletomotor impulses connected to execution of that particular action; the researchers note that this essentially “readies” a person for pursuit, thus increasing the attainability of that goal (i.e., facilitating action on time-limited opportunities).
A wide variety of situational features have been shown experimentally to prime or unconsciously activate relevant goals, from social contexts such as having power (Chen et al. Reference Chen, Lee-Chai and Bargh2001; Custers et al. Reference Custers, Maas, Wildenbeest and Aarts2008), to material objects such as dollar bills or briefcases (Kay et al. Reference Kay, Wheeler, Bargh and Ross2004; Vohs et al. Reference Vohs, Mead and Goode2006), scents (Holland et al. Reference Holland, Hendriks and Aarts2005), and even the names of significant others in one's life (Fitzsimons & Bargh Reference Fitzsimons and Bargh2003; Shah Reference Shah2003). In the everyday world, the presence of a goal-relevant object usually signals the presence of an opportunity for pursuit (e.g., when a person encounters a piece of cake, usually, he or she has an opportunity to eat it). The context-sensitivity of goal activation highlights how goal processes can unconsciously prepare a person for pursuit the instant that potential opportunities arise.
Second, Custers and Aarts (Reference Custers and Aarts2010) note the mental representation of a goal can become associated with positive affect; this “tagging” signals goal desirability, which subsequently guides individual behavior toward the desired end-state. In one experiment, participants were exposed to either goal (puzzle-related) words subliminally paired with either positive words (e.g., friend; beach), negative words (e.g., garbage; disease), or neutral words (Custers & Aarts Reference Custers and Aarts2005). Participants were most interested in completing puzzles if they had been seen the puzzle-related words subliminally paired with positive words. Conversely, subliminally pairing goal representations with negative affective stimuli decreased the effort people expended toward associated end-states (Aarts et al. Reference Aarts, Custwers and Holland2007; for similar results with subliminal monetary reward cues, see Schmidt et al. Reference Schmidt, Palminteri, Lafargue and Pessiglione2010).
Theoretically, a self-perpetuating cycle of goal pursuit could occur under certain circumstances. Similar to how a successful selfish gene replicates itself into future generations (Dawkins Reference Dawkins1976), goal representations become more likely to be pursued in the future, through mechanisms that operate independent of individual awareness or guidance. Once a goal representation becomes associatively linked to the situations in which it is frequently and consistently pursued (Bargh Reference Bargh, Higgins and Sorrentino1990), this link increases the likelihood that the goal will be pursued through those specific means instead of others (e.g., Bargh Reference Bargh, Higgins and Sorrentino1990; Gollwitzer Reference Gollwitzer1999; Veltkamp et al. Reference Veltkamp, Aarts and Custers2008). Similarly, Strack and Deutsch (Reference Strack and Deutsch2004) concluded that need states become strongly linked in memory with the behaviors and situations in which the need was satisfied, so that subsequent experiences of that need activate a bias in attention, or perceptual readiness (Bruner Reference Bruner1957) for those same situational cues. Hence, unconscious goals may become active at precisely those situations in which the individual has historically attained the goal, such that with each successful attempt, the original cues become more reliable triggers of pursuit in the future.
Affective processes may also reinforce the pursuit of successful unconscious goals. Research suggests that when people pursue conscious goals successfully (versus unsuccessfully), naturally occurring positive (versus negative) affect produces a strengthening versus weakening of that goal-pursuit tendency (selecting that goal over other possibilities) in future situations (e.g., Bandura Reference Bandura1977; Reference Bandura1986). A similar positive conditioning effect occurred during the selection of unconscious goals, as shown by Aarts et al. (Reference Aarts, Custwers and Holland2007). Consequently, it is possible that a snowballing or virtuous-cycle effect is possible wherein goals activated unconsciously by environmental cues that are successful in their pursuit will naturally become stronger (i.e., more likely to be pursued in the future), whereas those not as successful became weaker based on performance feedback (success vs. failure) from the environments frequented by the person.
4.2.2. Dissociation between active goal operation and individual awareness
Neuroscientific evidence also supports the dissociation of action systems from awareness. That executive control structures can operate without the person's awareness of their operation would require the existence of dissociable component processes within executive control or working memory structures (Baddeley Reference Baddeley2003; Baddeley & Hitch Reference Baddeley, Hitch and Bower1974; Buchsbaum & D'Esposito Reference Buchsbaum and D'Esposito2008). Evidence of such dissociations has been reported in stroke patients with “environmental dependency syndrome” caused by lesions in the frontal cortical lobes (Bogen Reference Bogen1995; Lhermitte Reference Lhermitte1986). The behavior of these patients was almost entirely driven by situational cues – for example, gardening in a public park (for hours) upon sight of a rake, drinking a glass of water every time the glass was filled despite complaining about being painfully full – with the patients entirely unaware of the oddity and irrationality of their behavior.
Moreover, patients with lesions in the ventral-visual system are unable to consciously report the visual properties of an object, yet are able to incorporate the inexplicable information into movements to grasp the object; whereas those with lesions in the parietal lobe can report on the object but are unable to reach for it successfully (Goodale et al. Reference Goodale, Milner, Jakobson and Carey1991). Brain damage associated with anarchic hand syndrome results in autonomous, goal-directed movements of a limb that cannot be consciously inhibited (e.g., one patient, M.P., reportedly would use one hand to select a channel on the television remote while the other hand – against his will – would immediately press another button; Marchetti & Della Sala Reference Marchetti and Della Sala1998).
This evidence has led some to conclude that conscious intentions are represented in the prefrontal and premotor cortex, whereas the parietal cortex houses the representation used to guide action (Frith et al. Reference Frith, Blakemore and Wolpert2000). (For additional evidence of the operation of action systems dissociated from conscious awareness, see Dijksterhuis & Aarts [Reference Dijksterhuis and Aarts2011], Milner and Goodale [Reference Milner and Goodale1995], Morsella & Bargh [Reference Morsella, Bargh, Decety and Cacioppo2011], and Wegner [Reference Wegner2002].) We take such findings as additional support for the notion that the mechanisms guiding individual behavior evolved separately from the mechanisms furnishing conscious awareness of their operation.
4.2.3. Unconscious goal conflict
Moreover, to the extent that individual behavior was originally driven by multiple, unconscious goals, research should reveal the presence of goal conflict that occurs in the absence of awareness. Kleiman and Hassin (Reference Kleiman and Hassin2011) replicated procedures and findings from the earlier goal-priming studies (Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001) to study whether cooperation and wealth accumulation goals could come into conflict with each other, while the person remains unaware of this tension. Participants were exposed to either cooperation-related or goal-neutral words in an ostensible language test before completing a Commons Dilemma exercise in which participants' usual responses are to compete and thus to accumulate wealth (although they could cooperate instead). The experimenters reasoned that unconsciously priming participants with the goal to cooperate would give rise to goal conflict. Consistent with this prediction, goal-primed participants manifested signs of goal conflict, displaying increased arousal through higher skin conductance levels, requiring more time to reach their decisions, and vacillating more between cooperative and selfish behaviors, as compared to control participants. Moreover, these implicit markers of goal conflict persisted over a five-minute delay and were uncorrelated with participants' self-reported experiences of conflict, thus supporting the notion that conflict can be goal-driven but can occur outside of awareness.
Yet another prediction that follows is that goal conflicts can be resolved independently from conscious guidance. Although social psychological research has yet to directly address this question, there is reason to believe that unconscious conflict resolution is possible. As described above, psychological research has revealed unconscious mechanisms that function to preemptively minimize goal conflict. An active, unconscious goal may dominate other goals by superseding conflict altogether, similar to the manner with which people physically remove temptations from their surroundings to bolster their chances of successful self-regulation (e.g., Kuhl Reference Kuhl and Maher1984; Kuhl & Weiss Reference Kuhl, Weiss, Kuhl and Beckmann1994). Indeed, a phenomenon called goal shielding demonstrates how an active goal can unconsciously interfere with how opposing goals come to mind (and thus subsequently influence perceptions and behavior; Shah et al. Reference Shah, Friedman and Kruglanski2002). In experiments on this topic, participants are slower to recognize words related to one goal (e.g., diet) if they are first subliminally primed with a word related to a competing, yet desirable goal (e.g., cake, which is eaten for enjoyment; Shah et al. Reference Shah, Friedman and Kruglanski2002; see also Fishbach & Shah Reference Fishbach and Shah2006; Veling & van Knippenberg Reference Veling and van Knippenberg2006).
Research also suggests that people who are good at self-regulating are particularly effective at inhibiting temptation-related stimuli: these participants are actually faster to recognize goal-related words (e.g., work) if those words are subliminally preceded by temptation-related words (e.g., basketball; Fishbach et al. Reference Fishbach, Friedman and Kruglanski2003). According to Counteractive Control Theory (Fishbach & Trope Reference Fishbach and Trope2005), temptations preemptively bolster people's higher-order goals and diminish the appeal of the temptation, thereby driving effective self-regulation. Such findings and theories are in harmony with the premise that unconscious goals are associated with mechanisms that negotiate conflict in the absence of awareness. Stated another way, self-control (or control of the self in situations of potential goal conflict) can be exerted unconsciously.
In the study of complicated goal systems, social cognitive experiments may best contribute by documenting phenomena observed in the activity of a single goal (or a few goals), whereas other areas of science (e.g., computer science) may be better suited to examine how goals interact within dynamic, larger systems. We offer here a preliminary conjecture regarding the resolution of goal conflict that, following a period of goal conflict (signified by processes similar to those documented in Kleiman and Hassin [Reference Kleiman and Hassin2011]) and resolution, the output might involve phenomena described by Counteractive Control Theory, wherein the strongest or most motivating goal eventually “wins out” in the self-control conflict and eventually steers the individual's behavior within that particular situation or in others like it.
4.3. The reconfiguration principle
The reconfiguration principle predicts that the downstream constellation of goal-priming effects observed in social cognitive research can be understood as constraining the person's cognitive and affective machinery for the purposes of facilitating goal pursuit. A stronger version of this hypothesis holds that the active goal is powerful enough to reconfigure that mental machinery, to the point of making typically effortful processes efficient and automatic if this is necessary for successful goal pursuit, and by inhibiting the normally chronic, automatic effects if they would serve to interfere with successful goal completion.
4.3.1. Attention and perception
Once active, the goal directs one's attention toward some (i.e., goal-instrumental) stimuli and away from others; in effect, the world is seen through goal-colored glasses. Salient, unusual events can be missed entirely when a person is pursuing a goal, as in the “inattentional blindness” research (Mack Reference Mack2003; Simons & Chabris Reference Simons and Chabris1999). In one well-known study, many participants who were busy with the explicit, conscious task of counting the number of ball tosses between actors on a computer display actually failed to notice a gorilla walking right through the ball-tossing game.
Indeed, similar to its conscious counterpart (Anderson & Pichert Reference Anderson and Pichert1978; Hastie & Park Reference Hastie and Park1986), an unconsciously operating impression formation goal causes greater selective attention to behavioral information inconsistent with the target's general and emerging pattern of behavior (Chartrand & Bargh Reference Chartrand and Bargh1996; McCulloch et al. Reference McCulloch, Ferguson, Kawada and Bargh2008). Highly accessible goal constructs provide “orienting value,” automatically guiding the individual's attention to relevant stimuli in the environment (Bruner Reference Bruner1957; Roskos-Ewoldsen & Fazio Reference Roskos-Ewoldsen and Fazio1992), which increases the probability that these objects will be used to achieve that goal.
Goal-facilitating objects can also appear more accessible along different dimensions, for example, by appearing closer in proximity to the pursuer or even larger in size. Balcetis and Dunning (Reference Balcetis and Dunning2010) demonstrated that people perceive the spatial orientation of desirable objects (which are the objects that help people achieve their goals; Ferguson & Bargh Reference Ferguson and Bargh2004) as being closer, compared to undesirable objects that are the same actual distance away. Similarly, Veltkamp and colleagues (Reference Veltkamp, Aarts and Custers2008) established that participants who were subliminally primed with a gardening goal overestimated the size (height) of goal-instrumental objects (e.g., a shovel), but not of goal-irrelevant objects (e.g., a pen).
These recent findings are consonant with a long-standing research tradition in social psychology on the motivated perceptual interpretation of events (e.g., Kunda Reference Kunda1990). For example, at sporting events, fans of both teams involved are certain that the referees are against them (Hastorf & Cantril Reference Hastorf and Cantril1954), and people consider studies that produce evidence consistent with their existing beliefs (e.g., concerning the death penalty) as being objective and properly conducted, whereas those studies producing findings contradicting their beliefs are believed to be flawed and biased (Lord et al. Reference Lord, Ross and Lepper1979). Self-protective motivations transform ego-threatening outcomes (e.g., failure on a test) into more palatable versions that place the blame on external factors (e.g., blaming others; Crocker & Park Reference Crocker and Park2004). One might consider one's teacher to be fair and competent until she gives one a bad grade, and thereafter one might consider her to be biased and incompetent (Sinclair & Kunda Reference Sinclair and Kunda2000). In each of these examples, information processing occurs in the service of the goal, regardless of whether the individual is aware of it.
4.3.2. Evaluation
People's everyday judgments of other people, objects, and events are strongly influenced by how those stimuli relate to the goals they are pursuing. This principle has been a staple of social and motivational psychology since the seminal writings of Kurt Lewin (Reference Lewin1935, p. 78) who defined the valence of an environmental object or event in terms of whether it helps or hinders the attainment of one's current goals and the satisfaction of one's current needs. Experimental findings continue to substantiate Lewin's dictum (Ferguson & Bargh Reference Ferguson and Bargh2004). In one such experiment involving the sequential priming paradigm (Ferguson Reference Ferguson2008), participants were subliminally exposed to either cooperation goal primes (giving, nice) or control primes. Afterwards, their implicit positivity toward either goal-relevant (help and generous) or control words was measured. Participants who were subliminally primed to cooperate displayed increased positive attitudes toward goal-instrumental words but not control words; and in a subsequent experiment, participants' implicit positivity toward goal-related primes predicted the degree to which they actually helped another person.
Active goal influence is so powerful that it can change evaluations of friends, enemies, and even significant others – the very people about whom one's opinions presumably remain stable over time. Fitzsimons and Shah (Reference Fitzsimons and Shah2008) found that participants who were unconsciously primed with an achievement goal evaluated friends who had helped them with their academic pursuits more positively compared to friends who had not helped them academically. This momentary favoritism toward goal-instrumental friends was not observed for unprimed control participants (see also Fitzsimons & Fishbach Reference Fitzsimons and Fishbach2010).
The goal-driven nature of these interpersonal evaluations is reminiscent of the successful intervention by Sherif and colleagues (Reference Sherif, Harvey, White, Hood and Sherif1961) in the classic “Robbers' Cave” study. The Rattlers and the Eagles, two warring groups of boys at a summer camp, antagonized each other with increasing violence until they were given an important shared goal. In a situation where everyone's cooperation was needed (e.g., freeing a truck that was stuck in the mud to get food for the entire camp), a Rattler's help became instrumental for an Eagle's goals (and vice versa). Changing the campers' goals dramatically changed how Rattlers and Eagles perceived one another and transformed summer-long rivals into close friends.
4.3.3. Overriding automatic processes that conflict with the active goal pursuit
The transformational power of the active goal over cognitive and affective processes is further indicated by its ability to override otherwise chronic, automatic encoding tendencies. For example, there is much evidence of the automatic manner in which other people are automatically encoded or categorized in terms of their race, age, and gender (e.g., Bargh Reference Bargh, Chaiken and Trope1999; Brewer Reference Brewer, Srull and Wyer1988). Recent research, however, suggests that chronic goals assumed to be egalitarian inhibit the same prejudicial biases previously assumed to be automatic and uncontrollable (e.g., Kunda & Spencer Reference Kunda and Spencer2003; Maddux et al. Reference Maddux, Barden, Brewer and Petty2005; Moskowitz et al. Reference Moskowitz, Gollwitzer, Wasel and Schaal1999).
A similar overriding effect of automatic, prejudicial processes occurs with temporarily active goals as well. Macrae and colleagues (Reference Macrae, Bodenhausen, Milne, Thorn and Castelli1997) found that giving participants a task goal to detect the presence or absence of dots on facial photographs eliminated any automatic stereotype activation effects when minority faces were shown (since stereotype activation was irrelevant to participants' processing objective). Research also suggests that default negative racial responses to African-American faces on the Implicit Association Test (IAT) can be flipped into positive evaluations when participants are informed that those same faces belong to their online teammates (Van Bavel & Cunningham Reference Van Bavel and Cunningham2009). These findings are consistent with the notion that joint goals (which are introduced by new alliances) can override automatic processes, causing the recategorization of out-group members into in-group members.
4.3.4. Creation of temporary automatic processes
Goal operation can give rise to novel and temporarily automatic effects as well. For example, implementation intentions, in which one commits oneself to a goal-furthering action in advance by mentally associating a specific concrete goal-pursuit action with an expected future event (“when, where, and how” the action will take place), have been shown to be highly effective means to attain otherwise difficult ends (diet, exercise, difficult health regimens; Gollwitzer Reference Gollwitzer1999; Webb & Sheeran Reference Webb and Sheeran2006). Implementation intentions effectively delegate control over one's future behavior to the environment, so that a specified, reliably occurring (e.g., routine) future event becomes the automatic trigger of that desired behavior (Gollwitzer Reference Gollwitzer1999). In this way a temporary or strategic automatic effect is created in the service of conscious goal pursuit.
Many so-called automatic effects are in fact goal-dependent (Bargh Reference Bargh, Uleman and Bargh1989), in that they only occur in the context of a specific active goal, such as in many automated acquired skills (e.g. Shiffrin & Schneider Reference Shiffrin and Schneider1977). Driving a car is commonly understood as an automatic skill for the experienced driver, but this depends on the individual having the goal of driving somewhere in the first place. Stimuli that produce an immediate and unintended reaction under one active goal do not do so under a different active goal, such as when driving a car and kicking one's right foot out to hit the brakes upon suddenly seeing a red light, but not kicking out that foot when seeing the same red light while walking on the sidewalk (Asch Reference Asch1952, p. 105; Bargh Reference Bargh, Bornstein and Pittman1992).
Spencer and colleagues (Reference Spencer, Fein, Wolfe, Fong and Dunn1998) provide perhaps the most dramatic example of a nonautomatic process becoming automatic when it facilitates the current goal pursuit. Research suggests that conditions such as attentional load can prevent people from engaging in negative stereotyping processes. Spencer and colleagues (Reference Spencer, Fein, Wolfe, Fong and Dunn1998) reasoned, however, that negative stereotyping is a means through which one can enhance one's own self-esteem (at the expense of others), and therefore should persist even in conditions that normally impede stereotyping effects given participants' active needs to restore their self-esteems. Indeed, by providing (bogus) feedback that participants had done very poorly on a task, the experimenters were able to elicit automatic stereotyping effects under conditions where such processes normally do not occur, thereby providing a particularly powerful demonstration of the active goal's ability to “selfishly” reconfigure a person's cognitive machinery in the service of its own pursuit.
4.4. The similarity principle
If conscious and unconscious goal pursuits make use of the same underlying motivational circuits and system, one would expect a high degree of similarity between conscious and unconscious goals in terms of the various component processes recruited, as well as the outcomes produced. An even stronger hypothesis is that conscious goal operation will share the autonomy of operation that is necessarily observed with unconscious goals (because the latter operate outside of conscious self-control). Thus, even when an individual is consciously pursuing a goal, that person is not necessarily in control of or even aware how engaging in goal pursuit has transformed his or her behavior and experience of the world. Recent experiments directly bear on this prediction.
4.4.1. Similarity in processes and neural structures
As previously mentioned, in the original goal-priming studies, unconsciously operating goals produced similar outcomes as when those same goals are pursued consciously, as well as with the same phenomenal qualities (Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001; Chartrand & Bargh Reference Chartrand and Bargh1996). Since then, research continues to demonstrate similarities between both varieties of pursuit, including the recruitment and use of similar affective processes (Custers & Aarts Reference Custers and Aarts2010).
Lending further credence to the notion that conscious and unconscious goals operate using similar processes, McCulloch and colleagues (Reference McCulloch, Ferguson, Kawada and Bargh2008) demonstrated that a primed, unconsciously operating goal to form an impression of a target person follows the same processing stages as was long known for the case of conscious impression formation (see Hamilton et al. Reference Hamilton, Katz and Leirer1980; Srull & Wyer Reference Srull and Wyer1979; also Chartrand & Bargh, Reference Chartrand and Bargh1996). Compared with a nonprimed control group, in separate experiments, priming the impression formation goal caused participants to (a) be faster to encode behaviors in trait-categorical terms, (b) be more likely to form associations between behaviors, and (c) notice and remember impression-inconsistent behaviors, all well-established subprocesses of conscious impression formation.
Indeed, cognitive neuroscience studies of the brain regions involved in motivated behavior support a model wherein the same underlying mechanisms govern both unconscious and conscious forms of goal pursuit. Pessiglione and colleagues (Reference Pessiglione, Schmidt, Draganski, Kalisch, Lau, Dolan and Frith2007) showed that people automatically increased effort on a hand-grip task when the reward cue (amount of money to be won on that trial) was presented subliminally, the same as what occurred when the reward cue was presented to conscious awareness. They also found that the same region of the basal forebrain moderated task effort level in response to both the consciously perceived and the subliminally presented reward signals. The authors concluded that “the motivational processes involved in boosting behavior are qualitatively similar, whether subjects are conscious or not of the reward at stake” (Pessiglione et al. Reference Pessiglione, Schmidt, Draganski, Kalisch, Lau, Dolan and Frith2007, p. 906).
Mainstream accounts of executive control or working memory within cognitive science long held that all of the contents of working memory were accessible to conscious awareness – indeed, until recently, “working memory” and “conscious awareness” were used as synonymous terms (e.g., Smith & Jonides Reference Smith and Jonides1998). Yet for goal pursuits to operate unconsciously, in real-time interaction with the fluid and dynamic external environment, active goals must make use of flexible working memory structures that operate on and often transform incoming informational input to serve the goal's agenda (Cohen et al. Reference Cohen, Dunbar and McClelland1990).
Participants in studies in which goals are primed and activated unbeknown to them cannot know in advance which goal-relevant stimuli might be presented; in fact, they are not even aware of which stimuli are goal-relevant and which are not. Nevertheless, in each experimental demonstration of unconscious goal pursuit, the primed goal produced the goal-appropriate outcomes, just as with conscious goal pursuit. For the obtained results to have occurred, the active goal had to be ready for whatever goal-relevant environmental input might arise and then operate on it when it did occur; unconscious goal pursuit therefore must involve the use of executive control and working memory functions as used in conscious goal pursuit (Frith et al. Reference Frith, Blakemore and Wolpert2000; Hassin Reference Hassin, Hassin, Uleman and Bargh2005).
Direct evidence on this point has been provided recently by Marien et al (Reference Marien, Custers, Hassin and Aarts2012). In six experiments, the researchers subliminally primed a variety of different goals (e.g., socializing, academic performance) and demonstrated that the unconsciously activated goal “hijacked” the executive control functions of the mind, as revealed by their taking attentional capacity away from an ongoing conscious task such as proofreading.
In this way, unconsciously operating goals can produce the same flexibility of responding to a given set of stimuli that traditionally has been considered the exclusive province of conscious processes (cf. Morewedge & Kahneman Reference Morewedge and Kahneman2010).
4.4.2. Autonomy of conscious goal operation
If conscious and unconscious goals are similar, then the autonomy with which unconscious goals operate (which is clearly demonstrated when unconscious goals reconfigure judgment and behavior, with the individual unaware of the goal and thus unable to guide its progress) should also characterize the pursuit of conscious goals. That is, even the goals one intends to pursue, and of which one is aware, are capable of producing information processing and behavioral effects consistent with the goal's agenda but not necessarily with the individual's self-related values and/or overall interests, thus potentially leading to unintended consequences of intended (conscious) pursuits.
Accordingly, Bargh et al. (Reference Bargh, Green and Fitzsimons2008) hypothesized that conscious goal pursuit would operate on any relevant or applicable (Higgins Reference Higgins, Higgins and Kruglanski1996) information in the environment regardless of whether the individual intends or is aware of this operation. Participants watched a videotaped interaction with the goal of evaluating the interviewee's suitability for a specified job – either as a restaurant waiter or a newspaper crime reporter. These jobs were included because the desired personality characteristics of a waiter – deferential and polite – and a crime reporter – tough and aggressive – are opposites of each other. In a control condition, participants were told merely that they were watching two people getting reacquainted.
During the taped interview, the two conversation partners were interrupted by a person named Mike, who behaved either politely or rudely. After viewing the tape, participants were given a surprise impression task in which they rated Mike, and not the job candidate on whom they had consciously focused. Not surprisingly, participants in the waiter and control conditions liked the interrupter significantly more if he were polite than if he were rude, but participants in the reporter-goal condition actually showed the reverse preference: They liked rude Mike more than polite Mike. Intuitively, these results may appear surprising. As shown by the ratings of participants in the control condition, people do not normally find rude, aggressive people likable. This research suggests, however, that people will actually like such offensive people if such traits are valued within their currently active goal pursuits – even if the consciously intended focus of that goal pursuit had been another person entirely.
In the next section, we explore how conscious goals not only operate, but also turn off in ways that create inconsistencies in a person's behavior over time. The following evidence for the inconsistency principle is thus further evidence for similarities in operation and outcomes between conscious and unconscious goals.
4.5. The inconsistency principle
As multiple goals within a single individual operate autonomously, becoming activated, operating, and turning off with achievement of their associated end-states, a person's behaviors and judgments will continue to vary as a function of which goals are most motivating. To an outsider, that person's behavior may appear inconsistent over time and, at extremes, even contrary to his or her general self-interests.
As previously mentioned, multiple goals operate by changing a person's perceptions and behaviors in ways that encourage the attainment of their own end-states. A person may pursue a goal even if doing so jeopardizes his or her physical health (Griskevicius et al. Reference Griskevicius, Tybur, Gangestad, Perea, Shapiro and Kenrick2009; Hill & Durante Reference Hill and Durante2011), and even to the point of feeling helpless to prevent the goal pursuit (Loewenstein Reference Loewenstein1996). Indeed, many of the ways through which goal operation influences information processing and behavior (i.e., operating outside of a person's awareness and guidance; changing how he or she perceives and behaves in the world) help explain why that person's actions may map inconsistently, weakly, or at times not at all, onto what is clearly beneficial for that person.
4.5.1. Effects of goal turn-off
Attainment of the goal's end-state can also produce inconsistent effects for the individual holding that goal, as when goal completion leads to behaviors that are contrary to expressed important values. In the “goal turn-off effect,” once a goal pursuit attempt is completed, the goal deactivates (e.g., Atkinson & Birch Reference Atkinson and Birch1970; Lewin Reference Lewin1926) and then for a time inhibits the mental representations used to attain that goal (Förster et al. Reference Förster, Liberman and Higgins2005; Marsh et al. Reference Marsh, Hicks and Bink1998), theoretically in order to allow other important goals to be pursued. When a goal is achieved, its downstream influence on cognition and behavior disappears for a time, which can ironically produce behaviors contrary to those originally encouraged by that goal.
For example, research on “moral licensing effects” (Monin & Miller Reference Monin and Miller2001) demonstrates how the operation and completion of conscious goals can produce judgments that can appear inconsistent with that individual's recent behaviors or self-professed values (also Bargh et al. Reference Bargh, Green and Fitzsimons2008). In one study, participants who were given the opportunity to disagree with blatantly sexist comments were ironically more likely than a control group to recommend a man than a woman for a stereotypically male job (Monin & Miller Reference Monin and Miller2001). Similarly, in another study, supporters of then-U.S. presidential candidate Barack Obama were first given the opportunity to publicly endorse him (or not, in the control condition; Effron et al. Reference Effron, Cameron and Monin2009). Afterwards, all participants judged the suitability of a job for white versus black people and allocated funds to organizations serving white or black people. Compared to the control group, endorsing Obama caused participants to rate the job as more suitable for whites than blacks and to allocate funding to white causes at the expense of black causes. This latter effect held only for those participants who had scored high on a measure of racial prejudice.
4.5.2. Relation to self-deception phenomena
The capability of active goals to operate and become completed independently from the individual's conscious desires can be seen as “self-deceptive' insofar as the individual remains unaware – or inconsistently aware – of his or her motives. If participants do not know the actual reasons for their behavior when it is influenced by unconscious means (such as through priming manipulations), they should be prone to misattribute the reasons for their behavior to plausible (oftentimes desirable) reasons that are accessible to their conscious awareness. Just such an effect was reliably demonstrated in a series of studies by Bar-Anan et al. (Reference Bar-Anan, Wilson and Hassin2010). Compared to a control condition, male participants primed with the mating goal were more likely to choose to work with a female tutor on Topic A than a male tutor on Topic B; however, they later explained their choice in terms of greater interest in that topic (which had been randomly paired with either the male or female tutor).
Recent theories identify occurrences of self-deception as an evolutionarily adaptive strategy (e.g., McKay & Dennett Reference McKay and Dennett2009; von Hippel & Trivers Reference von Hippel and Trivers2011), because if one is not aware of pursuing a particular goal, one will be less likely to display subtle but telling cues of lying (such as looking nervous or being cognitively overloaded). The Selfish Goal model suggests an alternative, or at least an additional, perspective: To the extent that human judgment and behavior were driven by goal processes before a central “self” even evolved, many instances of “self-deception” can be seen as a result of the autonomous nature of all goal pursuits. Both conscious and unconscious goals encourage single-minded pursuit of the end-state and are capable of producing effects that appear on the surface to be in the service of “deceiving” the individual, such as a “study buddy” being a best friend before, but not following, a big exam.
5. Conclusions
When viewing a person's judgments across time and situations, observers tend to see that person as a coherent whole – as a single agent selecting behaviors expressed in a complementary single body. In this article we argued for a change in perspective. Our central claim here is that an individual comprises multiple goals, each of which exerts a “selfish” influence on how that person sees the world and behaves in it, guiding judgments and behavior in the service of the current goal but not necessarily in the service of the individual's actual, overall best interests.
The Selfish Goal model predicts widespread automaticity of higher mental processes and reconfiguration of a person's perceptual and behavioral processes according to which goals are most motivating in the current situation. Consistent with the notion that they guided behavior in the evolutionary past prior to consciousness, goals operate by steering an individual's behavior toward goal-specified end-states, even in the absence of that individual's awareness or consent. Hypothesized similarities between conscious and unconscious goals led to the observation that, indeed, both forms of pursuit recruit similar processes and are capable of operating and turning off without guidance or even awareness of the individual. As a person temporarily acts in alignment with one goal and then suddenly does not when that goal is completed, inconsistencies to behavior as well as unintended effects arise.
5.1. Future directions
Important future directions include clarifying the relationship between the present model and contemporary motivational models in psychology, particularly as they apply to conscious and unconscious goals. Although we have highlighted the similarities between conscious and unconscious goals, recent work is focusing on the key differences between varieties of thought. Notably, however, one recent review has concluded that the difference between conscious and unconscious processes is not in the relative role played in the guidance and production of behavior (Baumeister & Masicampo Reference Baumeister and Masicampo2010), consistent with our present argument that much of behavior is governed by unconsciously operating mechanisms. The growing consensus appears to be that conscious processes are superior in serving integrative functions (see Morsella Reference Morsella2005 regarding phenomenal awareness; Baumeister & Masicampo Reference Baumeister and Masicampo2010 regarding conscious thought), especially regarding certain types of information such as simulating future scenarios and the perspectives of other people (Baumeister & Masicampo Reference Baumeister and Masicampo2010; see also Dijksterhuis & Aarts Reference Dijksterhuis and Aarts2011). This is clearly a fruitful avenue for further comparative research in order to determine even more precisely the uniquely important functions we gain from conscious processing capabilities.
Given the generativeness of the analogy professed herein between genes and goals, another direction for future research involves understanding the precise nature of that link. Indeed, Ernst Mayr (Reference Mayr1976) argued that this link was critical for the expression in present time of genetic influences from the very distant past: how genes both pre- and post-natally influence the content, emergence, and structure of programs that guide adaptive attentional, judgmental, motivational, and behavioral responses to one's environment.
5.2. Implications
Today, just as did Freud (Reference Freud, Brill and Unwin1901) more than a century ago in The Psychopathology of Everyday Life, contemporary psychological theorists are invoking the concept of motivation (unconscious or conscious) in their explanations for why people behave in ways that seem to run against their self-interest and values. The Selfish Goal model offers potential theoretical insight and implications for motivationally based research at multiple levels of analysis.
For example, people often experience self-control failures in health-related domains (e.g., smoking, drug abuse). In these cases, people are aware that certain actions are detrimental to their physical health, yet they somehow fail to act on that available information. Goal-related interventions might highlight the importance of changing the environment to activate intervention-consistent goals and understanding the dynamic relationship between opposing goals in situations of self-conflict (for example, a dieter is more likely to eat calorie-rich foods after exercising; Fishbach & Dhar Reference Fishbach and Dhar2005). Indeed, models of goal conflict have been used to predict people's success at managing the competing goals of eating for enjoyment and dieting (Stroebe et al. Reference Stroebe, Mensink, Aarts, Schut and Kruglanski2008). Strengthening the automatic activation of motoric avoidance goals in the presence of addiction-related stimuli is proving an effective method for reducing intensity of cravings for the addictive substance (Wiers et al. Reference Wiers, Houben, Roefs, de Jong, Hofmann, Stacy, Gawronski and Payne2010). Using strategically automatic implementation intentions, in which goal pursuits are automatically activated at a future point in time by concretely specified (in advance) future situational conditions, is proving a boon to overcoming difficulties in engaging in healthy behaviors across a variety of domains (Sheeran et al. Reference Sheeran, Gollwitzer and Bargh2013).
Future research might view self or relationally destructive thoughts and behaviors through the lens of autonomously operating goals. Understanding how a goal may become associated with particular behavioral patterns can inform clinical topics as from self-injury to dysfunctional relationships. Indeed, the pervasive and seemingly paramount needs to protect one's self-esteem (Crocker & Park Reference Crocker and Park2004) and to negotiate self-protection and connectedness goals in one's interpersonal relationships (e.g., Murray et al. Reference Murray, Derrick, Leder and Holmes2008) ironically lead people to avoid potential sources of social support. Identifying which goals are most motivating for these self-sabotaging individuals and understanding how goal operation can affect self-concepts and interpersonal judgments may provide new directions for therapeutic interventions.
Another intriguing research possibility involves integrating systems-based approaches to the study of goals. For example, goals may encourage behaviors toward their own end-state that can be seen as contrary to the individual's best interests (e.g., self-injurious behavior); but what of the goals that also promote specific social structures? In political psychology, Jost et al. (Reference Jost, Pietrzak, Liviatan, Mandisodza, Napier, Shah and Gardner2008) have focused on system justification effects, in which people perceive the current status quo regarding political power and division of resources as legitimate and fair – even those who are low status and for whom the system actually operates against their self-interests. The researchers explicitly appealed to the operation of an unconscious system-justification motive in order to account for these “relatively puzzling cases of conservatism, right-wing allegiance, and out-group favoritism among members of low-status groups,” which can only be understood if they are “not even aware of the extent to which they are privileging the status quo and resisting change” (Jost et al. Reference Jost, Pietrzak, Liviatan, Mandisodza, Napier, Shah and Gardner2008, p. 596). Thus, viewing human behavior from the perspective of the selfish goal may deepen understanding of social structures wherein individuals behave in ways that seem clearly contrary to their own best interests.
Just as had Dawkins (Reference Dawkins1976) with the “Selfish Gene” theory, we should emphasize that The Selfish Goal model does not necessitate selfishness at the level of the individual person. The individual or “self” is composed of many goals – self-interested ones to be sure, but prosocial and morally principled ones as well (e.g., Mansbridge Reference Mansbridge and Mansbridge1990; Miller Reference Miller1999). Studies of children as young as 18 months old suggest that they can represent the intention of their caretakers and use the knowledge to help themselves (Tomasello et al. Reference Tomasello, Carpenter, Call, Behne and Moll2005); that this is an early emerging human capacity supports the idea that it is the active goal that is selfish, not necessarily the person.
Additionally, more communal goals such as cooperation, egalitarianism, helping, and putting the welfare of others over one's own have been shown to operate entirely automatically and unconsciously (Bargh et al. Reference Bargh, Gollwitzer, Lee-Chai, Barndollar and Trotschel2001; Moskowitz et al. Reference Moskowitz, Gollwitzer, Wasel and Schaal1999; Over & Carpenter Reference Over and Carpenter2009), demonstrating a moral implication of the Selfish Goal: that selfishly operating goals can produce unselfish outcomes at the level of the individual person.
As selfishness, in common parlance, means putting one's own welfare and needs above those of other people (Elster Reference Elster and Mansbridge1990; Jencks Reference Jencks and Mansbridge1990), one telling demonstration of a “selfless” (at the level of the individual) selfish goal comes from the Chen and colleagues' (Reference Chen, Lee-Chai and Bargh2001) study in which participants unconsciously primed with power were given a choice of experimental tasks to complete, with full knowledge that another participant would have to do the remaining tasks. For participants with a communal relationship orientation (who tend to care more about the welfare of those over which they have power; see Clark & Mills Reference Clark and Mills1993), being primed with power actually activated their communal or altruistic goals, causing them, under the influence of having power, to selflessly shoulder more, not less, of the task burden for the other participant. (For other participants without this communal orientation, power indeed had the effect of producing more selfish, self-interested behavior at the expense of others.) Furthermore, these participants subsequently also reported a greater concern with social approval and expressed fewer racist attitudes.
When examining behavior from the perspective of the individual, people can appear inconsistent by thinking, feeling, and acting in contradictory fashion over time. The Selfish Goal model addresses this puzzle, examining behavior from a goal-pursuit perspective. Specifically, behavior selection at the goal level reconfigures the individual's information processing and behaviors, produces person-level behavior that can be unconscious and autonomous, and results in tension between what is “good” for the individual versus what is “good” for the goal. Together, these outcomes explain how contradictions – such as being capable of both selfish and altruistic judgments and behaviors – can arise in a single individual across time and situations.
Yet, these inconsistencies do pose a problem for individuals in a social world in which trust and predictability of behavior are at a premium and are essential for positive, cooperative relations with one's peers (e.g., Fiske et al. Reference Fiske, Cuddy and Glick2007; Tetlock Reference Tetlock2002). Thus we note in closing that several recent accounts of the purpose of conscious thought have argued that it evolved (was selected for as an adaptive advantage) in order to manage these same public inconsistencies that are produced by selfish-goal operations (Baumeister & Masicampo Reference Baumeister and Masicampo2010; Mercier & Sperber Reference Mercier and Sperber2011; see also Gazzaniga Reference Gazzaniga1984). The conscious self, in this view, is not so much involved in the guidance of our purposive behavior so much as it is in the business of producing rationalizations and socially acceptable accounts for the actions produced at the goal level. Tetlock (Reference Tetlock2002) has argued that our accountability to others was so important over evolutionary time that we evolved the “politician” (or “defense attorney”) social mindset in order to maintain good relations within our group.
In a zinger often credited to Woody Allen, a character in the classic movie The Big Chill (1983) remarks that rationalizations were more important than sex, because he had gone months without sex but hardly a day without a good rationalization. That's no small potatoes: not being able to explain or justify any negative outcomes one was involved in to one's peers could come at the cost of ostracism or worse; being able to give a plausible positive account would thus have strong survival value (e.g., Panksepp, Reference Panksepp1998). In assigning this valuable politician role to conscious thought and the conscious self, room is thus made for autonomous goal processes as the proximal determinant of human judgment and behavior.
ACKNOWLEDGMENTS
Preparation of this manuscript was supported in part by Grant R01-MH60767 from the U.S. Public Health Service. We are indebted to the editor for patient guidance of the manuscript through the review process, and to the reviewers for thoughtful feedback that substantially improved the paper. We are grateful to past and present members of the Automaticity in Cognition, Motivation, and Evaluation (ACME) lab at Yale for their constant support. A long-term intellectual debt is owed to the pioneering motivation research and theorizing of Peter Gollwitzer and Gabriele Oettingen; in many ways it constitutes the shoulders on which the Selfish Goal model stands, though the arguments and conclusions herein are our responsibility alone. Special thanks for thoughtful feedback provided on previous versions of this paper go to Joshua Ackerman, Lindsey Beck, Margaret S. Clark, Adam Hahn, Spike Lee, Edward Lemay, Ezequiel Morsella, Anna Kaisa-Newheiser, Andy Poehlman, Alexandra Sedlovskaya, Hyunjin Song, Lawrence Williams, Nicholas Wilson, and Eric Uhlmann.
Target article
Selfish goals must compete for the common currency of reward1
Related commentaries (1)
The Selfish Goal: Autonomously operating motivational structures as the proximate cause of human judgment and behavior