Article

Special Issue: Atheism, Secularity, and Science

The God of Nonbelievers: Characteristics of a Hypothetical God

David F. Bradley*, Julie J. Exline, Alex Uzdavines

Department of Psychological Sciences, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106-7123, USA.

*Correspondence | David F. Bradley, Case Western Reserve University, Euclid Avenue, Cleveland, USA; Email: davidfbradley@gmail.com

Citation | Bradley, D. F., J. J. Exline, A. Uzdavines. 2015. The god of nonbelievers: Characteristics of a hypothetical god. Science, Religion and Culture, 2(3): 120-130.

DOI | http://dx.doi.org/10.17582/journal.src/2015/2.3.120.130

Guest Editors | John R. Shook PhD, Ralph W. Hood Jr. PhD, and Thomas J. Coleman III

People who believe in the existence of a god or gods describe their deities as having varying characteristics. In Western monotheistic cultures, people often see their god as an intimate relational partner (Beck and McDonald 2004; Granqvist, Mikulincer, and Shaver 2010) and describe their god in terms of relational characteristics such as being loving, cruel, or distant (Exline, Grubbs, and Homolka 2015). However, not everyone believes in the existence of gods. Can people who do not believe in the actual existence of any gods form an image of a hypothetical god? What relational characteristics do these images of gods have? We asked nonbelievers in the United States if they could generate an image of a hypothetical god, and we asked them to rate the degree to which this hypothetical deity possessed certain relational characteristics (i.e., loving, cruel, and distant). We then examined these views of a hypothetical god in relation to several personality traits and aspects of personal history regarding belief and nonbelief. Broadly speaking, we wanted to examine the hypothetical god images of nonbelievers and to evaluate whether these images would actually carry psychological “weight,” showing meaningful patterns of associations with relevant psychological variables.

Images of Gods

Researchers in the psychology of religion have tried to develop nuanced ways of capturing the characteristics people perceive in a god or gods (for reviews, see Davis, Moriarty, and Mauch 2013; Hall and Fujikawa 2013; Hill and Hood 1999; Rizzuto 1979). Many conceptions of god image focus on specific cognitive beliefs about gods, such as the role gods play in suffering (Hale-Smith, Park, and Edmondson 2012), gods’ moral judgments about suicide (Linehan et al. 1983), the amount of active interference by gods in the physical world (Lawrence 1997), and gods’ level of omniscience and omnipotence (Laurin, Kay, and Fitzsimons 2012; Norenzayan 2013; Shariff and Norenzayan 2011).

Other work has focused on emotions that people experience based on their ideas about gods and/or their perceived relationships with gods. Among religious believers, positive feelings tend to dominate (Exline et al. 2011; Pargament et al. 1998; Wood et al. 2010), such that loving is a very strongly endorsed characteristic of gods by monotheistic believers in Western cultures (Exline, Grubbs, et al. 2015). However, many religious believers do report experiencing negative thoughts and feelings about their gods, termed divine struggles (e.g., Ano and Pargament 2013; for reviews, see Exline and Rose 2013; Exline 2013; Pargament 2007). Along these lines, some religious believers do describe their gods as being cruel (Benson and Spilka 1973; Francis, Gibson, and Robbins 2001; Gorsuch 1968) and distant (Aten et al. 2008; Gorsuch 1968; Kirkpatrick and Shaver 1990; Phillips et al. 2004; Rizzuto 1979), though usually not as strongly as they describe their gods as being loving (Exline, Grubbs, et al. 2015). These three relational characteristics – loving, cruel, and distant – are the three traits we explored in this study.

God images in nonbelievers

On a theoretical level, there is no contradiction between nonbelief in the existence of a god or gods and the ability to form relational images of hypothetical gods. Furthermore, the ability to form relational images of hypothetical gods should not be construed as explicit or implicit acknowledgement by nonbelievers that any god actually exists. Just as one can form an image of Darth Vader from the Star Wars films in one’s mind and describe him as cruel (or, hypothetically, loving) without admitting that Darth Vader truly existed (even a long time ago in a galaxy far, far away), one can form an image of a god in one’s mind and describe the god as loving, cruel, and distant without ceding belief in the existence of that imagined deity, Anselm’s arguments from ontology notwithstanding (see Malcolm 1960).

Nonbelievers have many possible sources for forming their image of a hypothetical god. According to one survey, 68% of current nonbelievers identified as religious believers at age 12 (Kosmin et al. 2009), which means that some nonbelievers will likely be able to form a hypothetical image of god informed by the nature of the god they previously believed existed. Other nonbelievers, especially in relatively religious cultures such as the U. S., will be familiar with popular teachings of what a god is like, that is, how people in their culture tend to think about a god or gods, and they might be able to form an image of god based on these popular beliefs and teachings. Still other nonbelievers could form an image based on what they personally believe a god would be like if a god were to exist.

Extensive prior work has explored and validated the idea that believers in the existence of a god or gods can form images of gods that are relational in nature; however, less work has been done exploring the ways that nonbelievers conceive of gods. One project (Exline et al. 2011) showed that 1) some nonbelievers report experiencing emotions in the past toward a god they used to believe existed and 2) some nonbelievers report that if a hypothetical god were to exist, they would be angry at the god. Another study found that atheists, despite not believing in the existence of a god or gods, experience stress when daring God to harm themselves or close friends or family (Lindeman et al. 2014). Finally, some nonbelievers report that their nonbelief is due, in part, to negative past relational experiences with a god (Bradley, Exline, and Uzdavines 2015; Exline, Bradley, et al. 2015).

We predicted that most nonbelievers in a U.S. sample would be able to generate an image of a hypothetical god. Since most religious believers describe their god as loving more than cruel or distant, we predicted that generating a hypothetical god image based primarily on a past image of god would be associated with describing the hypothetical god as more loving and less cruel and distant. Despite the fact that most people who believe in gods believe in a god that is primarily loving (Exline, Grubbs, et al. 2015), many popular books written from a nonbelief perspective argue that the dominant conception of god in Western culture is truly cruel rather than loving (e.g., Dawkins 2006; Hitchens 2007). We therefore predicted that generating an image of a hypothetical god based on popular teachings about god would be associated with describing the hypothetical god as more cruel.

Personality and god image

Big Five

Very little research has been conducted on the Big Five personality traits of nonbelievers (for a review, see Caldwell-Harris 2012). Research that has been conducted suggests modest but meaningful associations between Big Five personality traits and aspects of nonbeliever attitudes and identity (Silver et al. 2014). This is consistent with results from a meta-analysis finding small but significant associations among religious believers between Big Five personality traits and religious style (Saroglou 2002).

Research regarding religious believers experiencing divine struggles, who are more likely than non-strugglers to see their god as cruel, may be informative. This research suggests that seeing a god as cruel is associated with greater neuroticism (Ano and Pargament 2013; Grubbs, Exline, and Campbell 2013; Wood et al. 2010) and lower agreeableness (Grubbs et al. 2013), whereas more positive views of a god are associated with greater conscientiousness. We thus made the following predictions: 1) Neuroticism will be related to viewing a hypothetical god as more cruel and less loving, 2) agreeableness will be related to viewing a hypothetical god as more loving and less cruel, and 3) conscientiousness will be related to viewing a hypothetical god as more loving. However, we predicted that these associations would be modest in size, especially compared to associations involving more domain-specific constructs.

Attachment

Modern attachment theory proposes that people form cognitive and emotional frameworks for understanding and interacting with others (Ainsworth 1989; Collins and Read 1990; Feeney and Noller 1990). Research suggests that there are four adult attachment styles: secure, preoccupied, dismissive, and fearful (Bartholomew and Horowitz 1991). Of these four styles, two – secure and preoccupied – involve seeing others as essentially good and worthy of relationship, while the other two – dismissive and fearful – involve seeing others as essentially undesirable in terms of relationship. Attachment is likely relevant in how god images are formed because religious believers sometimes report being in a relationship with a god that is qualitatively similar to human attachment relationships (Beck and McDonald 2004; Exline, Homolka, and Grubbs 2013; Granqvist and Kirkpatrick 2013; Kirkpatrick and Shaver 1990; Kirkpatrick 2004). Among nonbelievers, we predicted that secure and preoccupied adult attachment styles would be associated with hypothetical god images that are more loving, less distant, and less cruel. Conversely, we predicted that dismissive and fearful attachment would be associated with hypothetical god images that are less loving, more distant, and more cruel.

Socially desirable responding

Socially desirable responding reflects a tendency to portray oneself in accordance with social expectations (for a review, see Paulhus 2002). As discussed above, several best-selling books about nonbelief argue that the culturally dominant god in the U.S., the Abrahamic god of Judaism, Christianity, and Islam, is a cruel and unloving figure. However, little is known about how widespread that god image is among nonbelievers. We predicted that a tendency toward socially desirable responding would correlate with describing a hypothetical god as more cruel and less loving; however, we expected this relationship to be weak.

Past belief, nonbelief, and god image

History with god and religious participation

As mentioned above, nearly 70% of current nonbelievers reported being religious at age 12 (Kosmin et al. 2009). This finding suggests that many nonbelievers are likely to report emotions (both positive and negative) based on past experiences around a god or gods that they believed to exist at the time. These emotional experiences from earlier in life might, in turn, shape how these nonbelievers think about hypothetical gods in the present. We thus predicted that past positive experiences with a god or gods would be associated with viewing a hypothetical god as more loving, less cruel, and less distant, and that past anger at a god or gods would be associated with viewing a hypothetical god as less loving, more cruel, and more distant. We also predicted that greater levels of earlier participation in religious activities would increase the strength of all attributions toward a hypothetical god, due to increased salience of the concept of god, and therefore would be positively associated with describing a hypothetical god as loving, cruel, and distant. Finding support for these hypotheses would suggest that hypothetical god image among nonbelievers is a construct with psychological impact, rather than a construct with only intellectual or theological impact.

Current attitudes toward god and nonreligious participation

Current conceptions of a hypothetical god are also likely affected by current attitudes regarding hypothetical gods and current participation in explicitly nonreligious activities, such as reading atheist blogs or attending explicitly nonreligious groups. Asking nonbelievers if they would want a god or gods to exist is a simple way of assessing (in general) positive or negative attitudes toward a hypothetical god or gods. We predicted that a greater desire for a god or gods to exist would be associated with a hypothetical god image viewed as more loving, less cruel, and less distant. Finding support for this prediction would help provide evidence that hypothetical god image is a valid, psychologically meaningful construct among nonbelievers.

Engagement with explicitly nonreligious activity might reinforce negative beliefs about religion and hypothetical gods; on the other hand, individuals with negative attitudes toward religion and hypothetical gods might be drawn to participation in nonbelief activities. Through either pathway, we predicted that greater current involvement in explicitly nonreligious activity would be associated with seeing a hypothetical god as less loving, more cruel, and more distant.

Method

Participants and procedures

We recruited participants from Amazon’s Mechanical Turk (MTurk) system (Buhrmester, Kwang, and Gosling 2011) and paid them $2.00 for their participation in a one-hour internet-based survey titled “Emotions, Beliefs, and Attitudes.” This compensation was comparable to the average hourly payment for participation in other MTurk studies (Ross et al. 2010). We collected an initial sample of 458 adults who endorsed atheist or agnostic beliefs about the existence of a god or gods. Of this initial sample, 393 (85.8%) reported that they were able to form a hypothetical image of a god, and 65 (14.2%) reported that they were unable to do so. Participants who were able to form an image of a hypothetical god did not differ from participants who were unable to form an image of god in terms of age, gender, race, country of origin, or sexual orientation (ps > .05). Those indicating that they could not form an image of a hypothetical god were not given questions about characteristics of a hypothetical god. We thus excluded these participants from further analyses.

The final sample consisted of 393 participants (207 women, 174 men, 12 who selected other gender identity labels; Mage = 32.31, SDage = 11.38). Most participants identified as heterosexual (N = 311, 79.1%), followed by bisexual (N = 41, 10.4%), homosexual (N = 19, 4.8%), and other (N = 22, 5.6%). While all participants were U.S. residents, 28 (7.1%) reported being born outside the U.S. Participants predominantly self-identified as White/Caucasian/European-American (N = 343, 87.3%), with fewer participants identifying as Latino/Hispanic (N = 28, 7.1%), African-American/Black (N = 20, 5.1%), East Asian/Pacific Islander (N = 18, 4.6%), American Indian/Native American/Alaska Native (N = 16, 4.1%), and Asian Indian (N = 10, 2.5%).

Measures

Ability to generate a hypothetical image of god

Early in the survey, we asked participants if they were able to “generate an image of a god or gods based on these statements.” We then provided three potential ways to generate an image of a hypothetical god or gods: “An image of a god or gods that you held earlier in your life” (endorsed by 305 participants, 77.6% of sample), “popular religious teachings (for example, how people in your culture tend to think about a god or gods)” (N = 356, 90.6%), and “your own personal ideas about what a god or gods might be like if a god or gods did exist” (N = 229, 58.3%). Most participants (N = 314, 79.9%) reported being able to generate an image of a hypothetical god or gods using more than one source. Those who were able to generate more than one of the above image types were asked to choose the type that best represented the way they generated a hypothetical god image. Combining these responses with the responses from those who only endorsed one type of god image, a plurality of participants (N = 197, 47.0%) used a popular god image in responding to the relevant items, with fewer participants using a past (N = 118, 28.2%) or personal (N = 104, 24.8%) image of god.

Relational characteristics of a hypothetical god

We measured the relational characteristics assigned to a hypothetical image of a god using the God-10, a measure validated among individuals who held some belief in God (Exline et al. 2013). The God-10 contains the prompt, “To what extent does this image of ‘God’ in your mind have these characteristics?” followed by ten traits (e.g., uninvolved, unkind, and forgiving) that participants rate from 0 (not at all) to 10 (extremely). We used responses to these questions to obtain scores on three subscales: cruel (M = 3.20, SD = 2.90, α = .91), loving (M = 6.05, SD = 3.03, α = .95), and distant (M = 5.33, SD = 3.11, α = .90).

Lifetime frequency of positive and angry feelings toward a god or gods

Following an earlier article (Exline et al. 2011), we asked participants, using a scale from 0 (never) to 10 (always), “Looking back over your entire life, how often have you had positive feelings toward a god or gods?” (M = 3.86, SD = 2.66). We included a similar item regarding frequency of feeling angry at a god or gods (M = 3.07, SD = 2.95).

Desire to believe that a god exists

We asked participants, using a scale from 0 (not at all) to 100 (totally), “Regardless of how much you actually believe in a god or gods: how much do you want to believe in the existence of a god or gods?” (M = 37.14, SD = 32.95).

Level of prior religious participation and current nonreligious participation

We asked participants, using a scale from 0 (not at all) to 5 (more than once per day), “When you held some belief in a god or gods, how often did you participate in each of these activities in a typical week?” (Exline, Yali, and Sanderson 2000). Participants who indicated no prior religious belief (N = 63) were given values of 0 on this measure. The six items included praying or meditating, attending religious services, and talking to others about religious/spiritual issues (M = 1.55, SD = 1.10, α = .88).

We also asked participants, “How often have you participated in each of these activities in the past week?” followed by five items intended to parallel the prior set of items, but focused on nonbelief activities. Items included reading books or articles about nonreligious viewpoints or that are critical of religion, attending meetings of a group focused on a nonreligious identity (e.g., Secular Student Alliance, Center for Inquiry), and talking to others about issues involving a nonreligious worldview (M = 0.79, SD = 0.82, α = .80).

Big Five factors of personality

We measured the Big Five factors of personality using the well-validated Big Five Inventory–44 (John, Donahue, and Kentle 1991). This inventory uses responses to 44 items on a 1 (rarely or never) to 5 (very often) scale to measure five personality characteristics: Openness (M = 3.85, SD = 0.60, α = .81), Conscientiousness (M = 3.59, SD = 0.69, α = .86), Extraversion (M = 2.81, SD = 0.87, α = .89), Agreeableness (M = 3.53, SD = 0.64, α = .80), and Neuroticism (M = 2.94, SD = 0.86, α = .89).

Adult attachment

We used the well-validated Relationship Questionnaire (Bartholomew and Horowitz 1991) to measure adult attachment style. We asked participants to rate the extent to which each of four attachment styles described them, using a scale from 1 (does not describe me at all) to 7 (totally describes me). The attachment styles were described to participants in four short paragraphs, each representing one of four prototypical adult attachment patterns: secure (M = 3.88, SD = 1.73), fearful (M = 3.89, SD = 1.85), preoccupied (M = 3.45, SD = 1.71), and dismissing (M = 4.51, SD = 1.66).

Socially desirable responding

We used the 13-item Marlowe-Crowne Scale (Reynolds 1982) to measure a tendency toward responding to items in a socially desirable way. The items are true-false questions designed to detect self-endorsement of unrealistic but socially praiseworthy behaviors, such as “No matter who I’m talking to, I’m always a good listener” and “I’m always willing to admit it when I make a mistake.” Socially desirable responses are summed (M = 5.56, SD = 3.08, α = .76).

Results

Demographic differences

The degree to which a hypothetical god was described as cruel, loving, or distant was not associated with age, gender, race, country of origin, or sexual orientation (ps > .05).

Relational characteristics of a hypothetical god

Participants described their hypothetical image of god as being primarily loving (M = 6.05, SD = 3.03), followed by distant (M = 5.33, SD = 3.11) and, least strongly, cruel (M = 3.20, SD = 2.90; Bonferroni-corrected ps < .01). In a prior study (Exline, Grubbs, et al. 2015), adults who believed in the existence of god also rated their image of god as being primarily loving (M = 8.9, SD = 1.9, α = .94), followed by distant (M = 2.6, SD = 2.2, α = .88) and, least strongly endorsed, cruel (M = 2.0, SD = 1.9, α = .91; Bonferroni-corrected ps < .01). Compared to the earlier sample of believers, our nonbeliever sample described a hypothetical god that was notably less loving, more cruel, and more distant (ps < .001). Thus, although both nonbelievers and believers showed the same ordering of god-related characteristics (loving > distant > cruel), the ratings by nonbelievers suggest a hypothetical god figure who appears less benevolent and engaged than the god figure described by believers.

Table 1: Correlations between aspects of hypothetical god image, attitudes toward god and personality

Personality factors contributing to god image	Loving	Cruel	Distant
	r	r	r
BFI Extraversion	-.08	.01	-.02
BFI Agreeableness	.20**	-.22**	-.13**
BFI Conscientiousness	.04	-.03	-.02
BFI Neuroticism	-.03	.05	.001
BFI Openness	-.01	.07	.13**
Secure attachment	.05	-.11*	-.06
Preoccupied attachment	.10*	.01	.01
Dismissive attachment	-.11*	.09	.10
Fearful attachment	.01	.12*	.06
Social Desirability	.01	-.06	-.10
Nonbelief-related factors contributing to god image
Lifetime positive feelings to god	.28**	-.12*	-.12*
Lifetime anger at god	-.06	.25**	.14**
Prior religious participation	-.03	.13**	.09
Nonreligious participation	-.20**	.26**	.20**
Want god to exist	.22**	-.22**	-.15**
Source of god image
Past	.15**	-.11*	-.12*
Popular	-.05	.14**	.02
Personal	-.11*	-.04	.12*

 

* p < .05; ** p < .01

Correlations with attributes of a hypothetical god

Table 1 shows correlations between variables of interest and the relational attributes assigned to a hypothetical god, namely loving, cruel, and distant. We will discuss personality-level variables first, followed by variables focused on personal religious history.

Personality and attachment style

Personality traits showed only a few modest associations with images of a god that was loving, cruel, or distant (see Table 1). In terms of the Big Five, agreeableness correlated positively with seeing a hypothetical god as loving, and it correlated negatively with describing a hypothetical god as cruel or distant. Openness correlated positively with describing a hypothetical god as distant. As predicted, seeing a hypothetical god as more cruel correlated negatively with secure attachment and positively with fearful attachment. Contrary to predictions, conscientiousness, neuroticism, and socially desirable responding did not show any significant associations with the hypothetical god images.

Religious history, nonbelief, and associated emotions

Variables related to past and present belief and nonbelief were largely consistent with predictions (see Table 1). In terms of emotional factors, lifetime positive feelings toward a god or gods correlated positively with imagining a hypothetical god as loving and negatively with imagining a hypothetical god as cruel and distant. Also, as predicted, lifetime feelings of anger toward a god or gods correlated positively with seeing a hypothetical god as cruel and distant. In short, past emotions centered on god figures showed clear connections with current hypothetical god images. These connections with emotion extended to current motives as well: One’s current desire for a god to exist correlated positively with describing a hypothetical god as loving, and it correlated negatively with describing a hypothetical god as cruel and distant.

As predicted, recent nonreligious participation correlated positively with describing a hypothetical god as cruel and distant, and it correlated negatively with describing a hypothetical god as loving. Contrary to our predictions, level of prior participation in religious activities correlated positively with seeing a hypothetical god as cruel, but was not correlated with seeing a hypothetical god as distant or loving.

Source of god image

As predicted, forming an image of a hypothetical god based on an image of god held earlier in life was associated with describing a hypothetical god as more loving, less cruel, and less distant (see Table 1). Also as predicted, using popular conceptions of a god as a blueprint for a hypothetical god was associated with describing the hypothetical god as more cruel. Finally, as predicted, participants who used a personal idea of what god would be like if a god existed described a hypothetical god as more distant.

Table 2: Simultaneous multiple regression predicting relational characteristics of a hypothetical god

Personality factors contributing to god image	Loving	Cruel	Distant
	R2 =.23**	R2 =.26**	R2 =.14**
	β	β	β
BFI Extraversion	-.08	.05	-.05
BFI Agreeableness	.21**	-.16**	-.07
BFI Conscientiousness	.03	-.01	.01
BFI Neuroticism	-.05	.01	-.06
BFI Openness	-.03	.01	.09
Secure attachment	-.01	.01	.04
Preoccupied attachment	.06	.05	.04
Dismissive attachment	-.03	.05	.08
Fearful attachment	.08	.02	-.01
Social Desirability	-.10	.07	-.09
Nonbelief-related factors contributing to god image
Lifetime positive feelings to god	.26**	-.15**	-.13*
Lifetime anger at god	-.13**	.28**	.15**
Prior religious participation	-.07	.10	.08
Nonreligious participation	-.15**	.20**	.14**
Want god to exist	.16**	-.19**	-.13*
Source of god image (dummy coded with reference group = personal)
Past	.16**	-.04	-.17**
Popular	.07	.14**	-.07

 

* p < .05; ** p < .01

Multiple regression predicting attributes of a hypothetical god

We used simultaneous multiple regression to consolidate the correlates of hypothetical god images described above (see Table 2). As predicted, broad personality traits did not strongly predict descriptions of a hypothetical god. With the exception of agreeableness (which predicted seeing a hypothetical god as more loving and less cruel) and social desirability (which predicted seeing a hypothetical god as less loving), the personality variables were not significant predictors of the characteristics of hypothetical gods imagined by nonbelievers.

The variables associated directly with nonbelief and religious history were more informative. As shown in the middle section of Table 2, higher levels of lifetime positive feelings toward a god or gods predicted views of a hypothetical god as more loving and less cruel and distant, whereas lifetime negative feelings toward a god or gods showed the opposite pattern. Higher self-reported desire for a god to exist predicted describing a hypothetical god as more loving and less cruel and distant. Level of current participation in explicitly nonreligious activities predicted describing a hypothetical god as less loving and more cruel and distant.

Lastly, the foundation of a participant’s image of a hypothetical god was associated with describing that hypothetical god as loving, cruel, and/or distant. As shown in Table 2, we compared participants who generated an image of a hypothetical god based on past or popular conceptions a god to participants who used current personal conceptions of a hypothetical god. Generating an image of a hypothetical god based on a previously held personal image was associated with seeing the hypothetical god as more loving and less distant. Conversely, using popular teachings about a god as the basis of for one’s image was associated with seeing the hypothetical god as more cruel.

Discussion

People who believe in the existence of a god or gods often describe these deities in diverse ways. One dimension along which believers have been shown to differ is the degree to which their gods hold different relational qualities, such as being loving, cruel, or distant (Exline, Grubbs, et al. 2015). It stands to reason that nonbelievers, too, differ from one another in terms of what comes to their minds if or when they imagine a god. The ability to imagine what a god might be like if a god were to exist does not imply a subconscious knowledge or belief in the existence of a god, no more than the continuing debates over the nature of Severus Snape six years after the publication of the last Harry Potter book (e.g., Asher-Perrin 2013) are indicative of widespread deficits in reality testing among Harry Potter fans. Little empirical work has been done to examine the nature of nonbelievers’ image of a hypothetical god. We used a measure of relational qualities, previously used in believer samples, to examine the nature of nonbelievers’ images of hypothetical gods and identify potential predictors of this hypothetical image of god.

Nonbelievers’ images of god

Contrary to our expectations, the most strongly endorsed relational quality of a hypothetical god was loving rather than distant or cruel. It is unlikely that this result was due to perceived social pressure or demand characteristics of the study, as socially desirable responding was not correlated with describing a hypothetical god as loving, cruel, or distant. We compared the nonbelievers in the present study to religious believers in a previous study. As expected, believers described their image of God as being more loving, less distant, and less cruel than nonbelievers’ image of a hypothetical god; nonetheless, both believers and nonbelievers showed a similar pattern of endorsement of relational characteristics: loving was the most strongly endorsed, followed by distant, and, least, cruel. Further research, such as a study that simultaneously collects god image data from both believers and nonbelievers, is necessary before making firm conclusions regarding differences and similarities in god images between these two groups.

Predictors of relational qualities of a hypothetical god

As we predicted, relative to past and present experiences with belief and nonbelief, personality traits were not strong predictors of seeing a hypothetical god as loving, cruel, or distant. While some adult attachment styles correlated weakly with relational characteristics of a hypothetical god, these associations did not appear as statistically significant predictors in multiple regression. The lack of a robust relationship between adult attachment style and relational characteristics of a hypothetical god among nonbelievers may be due to nonbelievers not imagining the hypothetical god as a deity that would be the subject of continued relationship, whereas religious believers often perceive themselves as being in an actual relationship with a god. We might expect that the more frequent interactions between individual and deity would lead to greater transference of adult attachment style onto the deity.

As predicted, greater lifetime history of positive feelings toward a god or gods predicted seeing a hypothetical god as more loving, less cruel, and less distant. Similarly, greater lifetime history of anger toward a god or gods predicted seeing a hypothetical god as more cruel and distant and less loving. This suggests that past relational experiences with a god or gods specifically, as opposed to more broad and general interpersonal relationship styles (e.g., adult attachment styles), could have more power to predict nonbelievers’ current ideas about a hypothetical god.

As predicted, increased participation in explicitly nonreligious activities, such as attending meetings of nonbelief-focused groups and spending time thinking about issues related to nonbelief, predicted describing a hypothetical god as less loving, more cruel, and more distant. However, the direction of causality is unclear here. This relationship could exist because people who view a hypothetical god more negatively see nonbelief as more important than other nonbelievers do, which leads them to further engage in nonbelief. On the other hand, it could be that nonbelief-espousing media and nonbelief-centered groups portray deities in a more negative light, such that spending time engaged with nonbelief leads to the adoption of negative views about a god or gods.

As predicted, the basis for generating an image of a hypothetical god differs among nonbelievers, and these differences predict variance in hypothetical god image. Generating an image of a hypothetical god based on a previously held god image was associated with imagining a more loving and less distant god, while generating a hypothetical god based on popular teachings about what a god is like was associated with imagining a more cruel god. This suggests that nonbelievers who think about god in terms of prior personal belief think of a more loving god, rather than being filled with anger, hatred, or other negative emotions toward god. On the other hand, nonbelievers who think about god in terms of popular teachings about god see god as a cruel being. This relationship remains significant even in a statistical model that includes past emotional experiences with gods and level of prior religious participation, suggesting that this is a distinct source of difference between nonbelievers.

Limitations and future directions

In this study, we recruited a sample of adults who lived in the U. S. Nonbelievers from other cultures may form a different hypothetical image of god, due to different cultural teachings about a god or gods or different experiences in relation to religion and religious belief. Some cultures, such as those more dominated by polytheistic belief, might tend to think of some gods as more cruel or loving than other gods (Exline, Kamble, and Homolka 2014). Other cultures, where deities are not seen as relational but rather as impersonal energy forces, may have nonbelievers who view a hypothetical god as more distant than loving or cruel. Additionally, while there was some ethnic diversity in our sample, a more ethnically diverse sample may have allowed for greater power in identifying the effect of ethnicity on god image among nonbelievers. Finally, in sampling only adults, we leave open the question of how nonbelieving children and adolescents conceive of a hypothetical god.

We used exclusively self-report measures in this study, which introduces difficulties related to impression management and level of insight. We tried to account for this possibility by informing participants of the confidential nature of the study and controlling for socially desirable responding. However, the possibility of misreporting remains.

In future work, longitudinal or experimental designs would facilitate development of causal models of hypothetical god image. Our study suggests that lifelong patterns of relating to a god or gods might impact how nonbelievers construct a hypothetical image of god. This seems to suggest some causal connection; however, the reports we collected of life history with regard to a deity were retrospective, introducing the possibility of reporting biases. Longitudinal or experimental designs would help to eliminate such biases.

Relational characteristics of a hypothetical god represent only one of many aspects of a god’s identity. Nonbelievers may also differ on other dimensions of deific identity, such as the god’s attitudes toward human curiosity, role in answering prayers, or plan for admission to the afterlife (if any). Further exploration of nonbelievers’ attitudes and assumptions about potential deities is warranted. However, the data presented here do provide a starting point, suggesting that many nonbelievers are able to conceptualize a hypothetical god, and that this hypothetical image of god has some psychological weight as evidenced by the observed relationships with emotional and personal-historical variables.

Acknowledgments

We are grateful for funding support from the John Templeton Foundation, Grant # 36094.

References

• Ainsworth, Mary S. 1989. Attachments beyond infancy. The American Psychologist 44: 709–716. http://dx.doi.org/10.1037/0003-066X.44.4.709
• Ano, Gene G., Kenneth I. Pargament. 2013. Predictors of spiritual struggles: An exploratory study. Mental Health, Religion & Culture 16: 419–434. http://dx.doi.org/10.1080/13674676.2012.680434
• Asher-Perrin, Emily. 2013. Severus Snape does not deserve your pity. Tor.com. Available at: http://www.tor.com/blogs/2013/04/severus-snape-does-not-deserve-your-pity (accessed 31 December 2014).
• Aten, Jamie D., Michael Moore, Ryan M. Denney, Tania Bayne, Amy Stagg, Stacy Owens, Samantha Daniels, Stefanie Boswell, Jane Schenck, Jason Adams, Charissa Jones. 2008. God images following Hurricane Katrina in south Mississippi: An exploratory study. Journal of Psychology and Theology 36: 249–257.
• Bartholomew, Kim, Leonard M. Horowitz. 1991. Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology 61: 226–244. http://dx.doi.org/10.1037/0022-3514.61.2.226
• Beck, Richard, Angie McDonald. 2004. Attachment to God: The Attachment to God Inventory, tests of working model correspondence, and an exploration of faith group differences. Journal of Psychology and Theology 32: 92–103.
• Benson, Peter, Bernard Spilka. 1973. God image as a function of self-esteem and locus of control. Journal for the Scientific Study of Religion 12: 297–310. http://dx.doi.org/10.2307/1384430
• Bradley, David F., Julie J. Exline, Alex Uzdavines. 2015. Relational reasons for nonbelief in the existence of gods: An important adjunct to intellectual nonbelief. Manuscript submitted for publication.
• Buhrmester, Michael, Tracy Kwang, Samuel D. Gosling. 2011. Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science 6: 3–5. http://dx.doi.org/10.1177/1745691610393980
• Caldwell-Harris, Catherine L. 2012. Understanding atheism/non-belief as an expected individual-differences variable. Religion, Brain & Behavior 2: 4–23. http://dx.doi.org/10.1080/2153599X.2012.668395
• Collins, Nancy L., Stephen J. Read. 1990. Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology 58: 644–663. http://dx.doi.org/10.1037/0022-3514.58.4.644
• Davis, Edward B., Glendon L. Moriarty, Joseph C. Mauch. 2013. God images and god concepts: Definitions, development, and dynamics. Psychology of Religion and Spirituality 5: 51–60. http://dx.doi.org/10.1037/a0029289
• Dawkins, Richard. 2006. The God Delusion. New York, NY: Houghton Mifflin Harcourt.
• Exline, Julie J. 2013. Religious and spiritual struggles. In Kenneth I. Pargament, Julie J. Exline, James W. Jones (Eds.), APA handbook of psychology, religion, and spirituality (Vol 1): Context, theory, and research (pp. 459–475). Washington, DC: American Psychological Association. http://dx.doi.org/10.1037/14045-025
• Exline, Julie J., David F. Bradley, Alex Uzdavines, Joshua B. Grubbs. 2015. Reasons for belief and nonbelief in God. Manuscript in preparation.
• Exline, Julie J., Joshua B. Grubbs, Steffany J. Homolka. 2015. Seeing God as cruel or distant: Links with divine struggles involving anger, doubt, and fear of God’s disapproval. The International Journal for the Psychology of Religion, 25: 29–41. http://dx.doi.org/10.1080/10508619.2013.857255
• Exline, Julie J., Steffany J. Homolka, Joshua B. Grubbs. 2013. Negative views of parents and struggles with God: An exploration of two mediators. Journal of Psychology & Theology 41: 200–212.
• Exline, Julie J., Shanmukh V. Kamble, Steffany J. Homolka. 2014. God’s role in suffering and anger toward God(s): Perspectives of Hindu undergraduates in India. Poster presented at the annual meeting of the American Psychological Association, Washington, DC.
• Exline, Julie J., Crystal L. Park, Joshua M. Smyth, Michael P. Carey. 2011. Anger toward God: Social-cognitive predictors, prevalence, and links with adjustment to bereavement and cancer. Journal of Personality and Social Psychology 100: 129–148. http://dx.doi.org/10.1037/a0021716
• Exline, Julie J., Eric D. Rose. 2013. Religious and spiritual struggles. In Raymond F. Paloutzian and Crystal L. Park (Eds.), Handbook of the psychology of religion and spirituality (2nd ed., pp. 380–398). New York, NY: Guilford Press. http://dx.doi.org/10.1037/14045-025
• Exline, Julie J., Ann M. Yali, William C. Sanderson. 2000. Guilt, discord, and alienation: The role of religious strain in depression and suicidality. Journal of Clinical Psychology 56: 1481–1496. http://dx.doi.org/10.1002/1097-4679(200012)56:12<1481::AID-1>3.0.CO;2-A
• Feeney, Judith A., Patricia Noller. 1990. Attachment style as a predictor of adult romantic relationships. Journal of Personality and Social Psychology 58: 281–291. http://dx.doi.org/10.1037/0022-3514.58.2.281
• Francis, Leslie J., Harry M. Gibson, Mandy Robbins. 2001. God images and self-worth among adolescents in Scotland. Mental Health, Religion & Culture 4: 103–108. http://dx.doi.org/10.1080/13674670126955
• Gorsuch, Richard L. 1968. The conceptualization of God as seen in adjective ratings. Journal for the Scientific Study of Religion 7: 56–64. http://dx.doi.org/10.2307/1385110
• Granqvist, Pehr, Lee A. Kirkpatrick. 2013. Religion, spirituality, and attachment. In Kenneth I. Pargament, Julie J. Exline, James W. Jones (Eds.), APA handbook of psychology, religion, and spirituality (Vol 1): Context, theory, and research (pp. 139-155). Washington, DC: American Psychological Association. http://dx.doi.org/10.1037/14045-007
• Granqvist, Pehr, Mario Mikulincer, Phillip R. Shaver. 2010. Religion as attachment: Normative processes and individual differences. Personality and Social Psychology Review 14: 49–59. http://dx.doi.org/10.1177/1088868309348618
• Grubbs, Joshua B., Julie J. Exline, W. Keith Campbell. 2013. I deserve better and God knows it! Psychological entitlement as a robust predictor of anger at God. Psychology of Religion and Spirituality 5: 192–200. http://dx.doi.org/10.1037/a0032119
• Hale-Smith, Amy, Crystal L. Park, Donald Edmondson. 2012. Measuring beliefs about suffering: Development of the Views of Suffering Scale. Psychological Assessment 24: 855–866. http://dx.doi.org/10.1037/a0027399
• Hall, Todd W., Annie Fujikawa. 2013. God images and the sacred. In Kenneth I. Pargament, Julie J. Exline, & James W. Jones (Eds.), APA handbook of psychology, religion, and spirituality (Vol 1): Context, theory, and research (pp. 277-292). Washington, DC: American Psychological Association. http://dx.doi.org/10.1037/14045-015
• Hill, Peter C., Ralph W. Hood. (Eds.). 1999. Measures of religiosity. Birmingham, AL: Religious Education Press.
• Hitchens, Christopher. 2007. God is not great: How religion poisons everything. New York, NY: Twelve Books.
• John, Oliver P., Edward Donahue, Richard Kentle. 1991. The Big Five Inventory—Versions 4a and 54. Berkeley, CA: Institute of Personality and Social Research, University of California, Berkeley.
• Kirkpatrick, Lee A. 2004. Attachment, evolution, and the psychology of religion. New York, NY: Guilford Press.
• Kirkpatrick, Lee A., Phillip R. Shaver. 1990. Attachment theory and religion: Childhood attachments, religious beliefs, and conversion. Journal for the Scientific Study of Religion 29: 315–334. http://dx.doi.org/10.2307/1386461
• Kosmin, Barry A., Ariela Keysar, Ryan Cragun, Juhem Navarro-Rivera. 2009. American nones: The profile of the no religion population: A report based on the American Religious Identification Survey 2008. Hartford, CT.
• Laurin, Kristin, Aaron C. Kay, Grainne M. Fitzsimons. 2012. Divergent effects of activating thoughts of God on self-regulation. Journal of Personality and Social Psychology 102: 4–21. http://dx.doi.org/10.1037/a0025971
• Lawrence, Richard T. 1997. Measuring the image of God: The God Image Inventory and the God Image Scales. Journal of Psychology and Theology 25: 214–226.
• Lindeman, Marjaana, Bethany Heywood, Tapani Riekki, Tommi Makkonen. 2014. Atheists become emotionally aroused when daring God to do terrible things. International Journal for the Psychology of Religion 24: 124–132. http://dx.doi.org/10.1080/10508619.2013.771991
• Linehan, Marsha M., Judith L. Goodstein, Stevan L. Nielsen, John A. Chiles. 1983. Reasons for staying alive when you are thinking of killing yourself: The Reasons for Living Inventory. Journal of Consulting and Clinical Psychology 51: 276–286. http://dx.doi.org/10.1037/0022-006X.51.2.276
• Malcolm, Norman. 1960. Anselm’s ontological arguments. The Philosophical Review 69: 41–62. http://dx.doi.org/10.2307/2182266
• Norenzayan, Ara. 2013. Big gods: How religion transformed cooperation and conflict. Princeton, NJ: Princeton University Press.
• Pargament, Kenneth I. 2007. Spiritually integrated psychotherapy: Understanding and addressing the sacred. New York, NY: Guilford Press.
• Pargament, Kenneth I., Bruce W. Smith, Harold G. Koenig, Lisa Perez. 1998. Patterns of positive and negative religious coping with major life stressors. Journal for the Scientific Study of Religion 37: 710–724. http://dx.doi.org/10.2307/1388152
• Paulhus, Delroy L. 2002. Socially desirable responding: The evolution of a construct. In Henry I. Braun, Douglas N. Jackson, David E. Wiley (Eds.), The role of constructs in psychological and educational measurement (pp. 49–69). Mahwah, NJ: Lawrence Erlbaum.
• Phillips III, Russell E., Kenneth I. Pargament, Quinten K. Lynn, Craig D. Crossley. 2004. Self-directing religious coping: A deistic God, abandoning God, or no God at all? Journal for the Scientific Study of Religion 43: 409–418. http://dx.doi.org/10.1111/j.1468-5906.2004.00243.x
• Reynolds, William M. 1982. Development of reliable and valid short forms of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical Psychology 38: 119–125. http://dx.doi.org/10.1002/1097-4679(198201)38:1<119::AID-JCLP2270380118>3.0.CO;2-I
• Rizzuto, Ann-Marie. The birth of the living God: A psychoanalytic study. Chicago, IL: University of Chicago Press.
• Ross, Joel, Lilly Irani, M. Six Silberman, Andrew Zaldivar, Bill Tomlinson. 2010. Who are the crowdworkers? Shifting demographics in Amazon Mechanical Turk. Presented at alt.CHI 2010.
• Saroglou, Vassilis. 2002. Religion and the five factors of personality: A meta-analytic review. Personality and Individual Differences 32: 15-25. http://dx.doi.org/10.1016/S0191-8869(00)00233-6
• Shariff, Azim F., Ara Norenzayan. 2011. Mean gods make good people: Different views of God predict cheating behavior. International Journal for the Psychology of Religion 21: 85–96. http://dx.doi.org/10.1080/10508619.2011.556990
• Silver, Christopher F., Thomas J. Coleman III, Ralph W. Hood Jr., Jenny M. Holcombe. 2014. The six types of nonbelief: A qualitative and quantitative study of type and narrative. Mental Health, Religion & Culture 17: 990-1001. http://dx.doi.org/10.1080/13674676.2014.987743
• Wood, Benjamin T., Everett L. Worthington, Julie J. Exline, Ann M. Yali, Jamie D. Aten, Mark R. McMinn. 2010. Development, refinement, and psychometric properties of the Attitudes Toward God Scale (ATGS-9). Psychology of Religion and Spirituality 2: 148–167. http://dx.doi.org/10.1037/a0018753