Use does not wear ragged the fabric of friendship: Thinking of objects as alive makes people less willing to replace them

Anthropomorphic beliefs about objects lead people to treat them as if they were alive. Two experiments test how anthropomorphic thought affects consumers’ product replacement intentions. Consumers induced to think about their car in anthropomorphic terms (i) were less willing to replace it and (ii) gave less weight to its quality when making replacement decisions. Instead, they (iii) attended to (experimentally induced connotations of) the car’s “warmth”, a feature usually considered relevant in the interpersonal domain. While anthropomorphic beliefs about brands are often seen as advantageous by marketers because they increase brand loyalty, similar beliefs about products may be less desirable. Although most people know that cars and computers are inanimate, at times they treat them as if they were alive. Most of us at some point have argued (or pled) with a computer, felt attached to a favorite sweater or expressed love for a new car. Anthropomorphic objects are also commonly featured in movies and books intended for both children (e.g., The Brave Little Toaster, Disch,1987) and adults (e.g., 2001: A Space Odyssey, Kubrick, 1968). People are experts in representing, predicting and interacting with both the physical and social worlds. When thinking about objects and people, there are important differences in the processing of information, both at a neural level (Mitchell, Heatherton & Macrae, 2002; Yoon, Gutchess, Feinberg & Polk, 2006) and in terms of self-report and behavioral measures (e.g. Cosmides, 1989). Thus whether a target is categorized as living or not has downstream consequences on what information comes to mind, what information is deemed relevant and what predictions seem plausible when thinking about the target. We examine the consequences of anthropomorphism on the kinds of information consumers’ attend to when forming product replacement intentions. Whereas objects are primarily evaluated based on how well they function, functionality is far less important when evaluating people – and we assume that the same may hold for anthropomorphized objects. We predict, and find, that priming anthropomorphic beliefs about a product shifts attention away from pragmatic considerations (such as the product’s functionality) and makes consumers sensitive to information that figures prominently in the interpersonal realm (such as the product’s perception as “warm” or “cold”). Specifically, consumers (i) reported lower intentions to replace their car when they were induced to think about it in anthropomorphic terms. More important, (ii) thinking about one’s car in anthropomorphic terms decoupled replacement intentions from the car’s perceived quality and functionality (Studies 1 and 2). Instead, (iii) consumers who were primed to Thinking of Objects as Alive 3 think about their car in anthropomorphic terms turned to information that is valued in the interpersonal domain and were less likely to replace their car when its color was associated with (experimentally manipulated) “warm” rather than “cold” connotations (Study 2). Next, we review relevant research and develop our rationale in more detail. Anthropomorphic Thought In principle almost any object of thought can become anthropomorphized, even invisible entities (e.g. ghosts; Bering, 1999) or purely symbolic concepts (e.g. brands; Aaker, 1997; for a review see Guthrie, 1993). People are particularly likely to spontaneously anthropomorphize objects that have human-like physical features, such as eyes (Haley & Fessler, 2005; Jipson & Gelman, 2007), hands (Woodward, 1999) or a human-like form (Aggarwal & McGill, 2007), or that behave in an apparently complex or intentional manner (Heider & Simmel 1944; Epley, Waytz, Akalis & Cacioppo, 2008). In addition, individual difference variables, like loneliness and desire for social contact, are associated with a higher propensity to anthropomorphize objects (Epley, Akalis, Waytz, & Cacioppo, 2008). Throughout, the findings are compatible with general models of knowledge accessibility (Higgins, 1996) and suggest that any variable that increases the temporary or chronic accessibility of agentic beliefs is likely to foster anthropomorphic thought about objects (for a review of the causes of anthropomorphic beliefs see Epley, Waytz & Cacioppo, 2007). Once a product is anthropomorphized, it becomes possible for consumers to enter into a relationship with it, changing the emotional quality of consumers’ experience with their possessions. Social interaction is pleasurable, and imbuing a product with anthropomorphic cues leads consumers to experience more positive affect when interacting with it (Wang, Baker, Wagner & Wakefield, 2007). Indeed, even in the absence of anthropomorphic cues, some consumers spontaneously experience uniquely social emotions when interacting with objects, including love (Schultz, Kleine & Kerman, 1991) and trust (Aaker, Fournier & Brasel 2004). Over time connections to special products and objects can grow, leading consumers to form a sense of attachment that parallels interpersonal attachment patterns (Wallendorf & Arnould, 1988; Ball & Tasaki, 1992; Bowlby, 1980). As a result, much like in interpersonal relationships, consumers can come to depend on objects, feeling a sense of security when they are close and distress when they are absent (Thomson, MacInnis & Park, 2005). Some companies have experimented with capitalizing on customers’ feelings of loyalty to reduce business costs. For Thinking of Objects as Alive 4 example, Zipcar decided to name all of its rental vehicles and reported that this led customers to be more diligent about cleaning and maintaining them (Levine, 2009). In combination, these observations suggest that consumers should be less willing to replace a product they own when they think about it in anthropomorphic terms. In addition, anthropomorphizing a product should reduce the usually observed link between perceived product quality and replacement intentions. This prediction follows from the observation that thought about the inanimate world is usually dominated by instrumental considerations, whereas thought about the social word is not. For example, perceived product quality and performance are central to product evaluation and these instrumental attributes usually trump more symbolic attributes, such as brand image and brand personality (Zentes, Morschett & Schramm-Klein, 2008). In a study of retail store brands, Zentes and colleagues found that the effect of store performance on attitudinal and behavioral measures of customer loyalty far exceeded the effect of customers’ perceptions of brand personality. In contrast, people maintain interpersonal bonds with relatively little regard for the instrumental contributions offered by others (Leach, Ellemers, & Barreto, 2007; Jones & Vaughan, 1990). Instead, considerations of the intentions behind actions (Falk, Fehr & Fischbacher, 2008), communality and relational status often trump considerations of instrumentality, like the actual contributions of the other’s skills and competence to one’s own outcomes (Wojciszke, Bazinska & Jaworski, 1998; for a recent review see Ybarra et al. 2008). In a relational context, social features of identity are particularly salient and people are more likely than not to behave in a manner consistent with accessible features of identity (Oyserman, 2009). As a result, social relationships are maintained even when they require costly sacrifices (Simmons, Marine & Simmons, 1977) and people routinely invest resources in caring for the old, the sick and the weak, even when they can no longer serve a useful function (Branscombe, Wann, Noel, & Coleman, 1993; Levine & Moreland, 2002). By changing the perceptions of a product, anthropomorphic cues may also make relational aspects of the self salient, allowing the product to fulfill identity based motives (e.g., Aaker & Akutsu, 2009; Shavitt, Torelli & Wong, 2009). When people think about objects in anthropomorphic terms, they apply knowledge about the social world to the inanimate world. This is reflected in categorization decisions (e.g., Jones et al., 2001) and the use of social categories (like gender or age) in the description of objects (e.g., Heider & Simmel, 1944; Berry & Springer, 1993). Moreover, people apply social schemas to Thinking of Objects as Alive 5 objects and prefer objects that fit the schema. This leads consumers to prefer product lines that match social structures, such as “family”, and cars that smile rather than frown (Aggarwal & McGill, 2007). Thus, “ownership” may instead become a “relationship” and the features relevant to relational partners may become applicable to anthropomorphized objects. In combination, these considerations lead to the following three predictions. When consumers think about a product in anthropomorphic terms they are (i) less willing to replace it; (ii) less likely to base replacement decisions on instrumental attributes of the product, thus decoupling perceived product quality and replacement intentions; and (iii) more likely to consider attributes that are usually deemed important in the interpersonal domain. The Present Research We test these predictions in two studies, using an experimental strategy that differs from the two dominant strategies employed in anthropomorphism research. One common strategy is essentially correlational and relies on observed parallels between social cognition and object cognition as evidence of anthropomorphic thought (e.g., Aggarwal, 2004; Berry & Springer, 1993; Heider & Simmel, 1944; Schultz, Kleine & Kerman, 1991; Thomson, MacInnis & Park, 2005; Wallendorf & Arnould, 1988; Ball & Tasaki, 1992). A second common strategy endows objects with human features to elicit anthropomorphic beliefs, e.g., by having the object talk, smile, frown, and so on (e.g., Aggarwal & McGill, 2007; Epley et al., 2008; Jones, Smith & Landau, 1991). Although both approaches have demonstrated anthropomorphic thoughts about objects, skeptics may wonder to which extent the results reflect the spontaneous application of social knowledge to objects as opposed to the influence of pragmatic linguistic or situational considerations. From this perspective, parallels between how individuals describe objects and people may reflect mere linguistic convenience. Moreover, the presentation of novel objects with human features may convey that the object is to be thought about in human terms – or why else would it be endowed with these features? Hence, the observation of apparent agentic beliefs about an anthropomorphized object may, at least in part, be the result of pragmatic inferences about the intentions of the communicator (Schwarz, 1996). The present studies attenuate these ambiguities in two ways. First, our dependent variable does not consist of object descriptions that may invite the use “human” terms for reasons of linguistic convenience. Instead, we assess product replacement intentions as a decision to which social norms are applicable. Second, we hold the object constant and manipulate, through a Thinking of Objects as Alive 6 preceding question, whether technical or “psychological” attributes of the object are likely to come to mind. Following this manipulation, participants describe the product in their own words and report how likely they are to replace it in the near future.