Abstract: The predominant religious traditions in the United States have traditionally represented God as a male. While feminist theologians and others have expanded notions of divinity beyond this mold, many Americans are still associated with religious traditions that have traditionally referred to God in paternal terms. Here I use a unique set of data from the General Social Survey to empirically examine beliefs regarding the gender of God in the United States population in 2012. I explore relationships between perception of the gender of God and denominational affiliation, background demographic characteristics, beliefs in the characteristics of God, abstractness of belief in God, and several indicators of self-rated religiosity. I find that people who view God as at least part female associate her with more familial, loving characteristics, while those who view God as exclusively male see him in more authoritative terms. Interestingly, fundamentalists, agnostics and those who see God as personally involved tend to see God as primarily and/or exclusively male.


The divine mother has a long history of various incarnations throughout ancient and modern human history. Ancient fertility goddesses were a virtually universal component of ancient pantheons, and have found some representation in modern Mariology. [1] However, the Christian God that inherited European civilization has a distinctly male character. While there are some glimpses of the maternal divine in the Christian and Jewish theologies and scriptures that form the corpus of the American religious experience [2], traditional patriarchal religious structures overwhelmingly emphasize the paternal aspect of God’s nature.

Attempts by the National Council of Churches and some mainstream and liberal Protestant denominations to incorporate more inclusive, gender-neutral language in ceremony and scriptural translations such as the TNIV represent an effort to downplay the exclusively male characteristics of God without necessarily explicitly affirming the divine feminine, representing a move to a concept of a God that is either equally gendered or that transcends gender. Interestingly, John Paul II’s Catechism of the Catholic Church uses a male-specific pronoun to affirm this latter position: “He is neither man nor woman: He is God” (239). [3] This idea too has a long history, as Plato and other ancient commentators saw the gender distinction as an unnatural separation of two components of the same celestial reality.

Outside of the predominant religious traditions, other denominations making up the American religious mosaic such as Wiccans [4], Mormons [5], Buddhists, and Hindus [6] have analogous variations on the divine feminine. The Mormon cosmology is an especially interesting case because in it the divine feminine is not simply a gendered dimension of some underlying  higher power (as it is in many other traditions), but is a distinct, personal Goddess Herself.

While the emphasis (or lack thereof) on the divine feminine in theology and tradition has been the subject of extensive discussion, here I investigate what the rank and file in the United States actually believes in regard to the gender of God. From 1984-2012, the General Social Survey (hereafter GSS) included in various years of their survey a question asking participants to rank their personal image of God from 1 (completely maternal) to 7 (completely paternal). As could be expected given the United States’ socio-religious background, the largest category chosen by far was 7. However, approximately half the respondents marked that their vision of God incorporated at least some female elements, with 3 percent indicating that they saw God as wholly female. Here I investigate what views, religious beliefs, notions of God, and background characteristics are associated with certain gendered notions of God. While more current data would be preferable, the current sample range is sufficient to explore the nature of the relationships pertaining to gendered beliefs about the divine feminine/masculine.

Specifically, I explore several angles: gender, religious practice, religious belief, religious conservatism/liberalism, perceptions of God, nature of belief in God, and denomination.


As the image-of-God’s-gender variable is quantified on a scale from 1-7, it is tempting to use standard OLS, or at least a nonparametric derivative. However, a quick glance at the structure of the data immediately indicates a much more complicated story, with a skewed bimodal distribution with peaks around 7 (exclusively male), and 4 (equally male and female). Despite most Americans coming from religious traditions where God is represented as entirely male, a surprising number of Americans see God as being at least partially female.


Therefore, I break the results into four conceptually distinct categories: God as predominately maternal (values 1-3), God as equally maternal and paternal (value 4), God as predominately but not completely paternal (values 5-6), and God as completely paternal (value 7). These categories are not constructed at constant intervals along the scale; instead, they follow the contours of the data. I then use an ordered logistic regression, with an exclusively paternal God as the reference case, to logistically measure differences between the reference case and the others on a category-by-category basis. I report my results in odds ratios, these ratios are ratios for the odds that the respondent takes on a certain value. So an odds ratio of 1.03 on a variable indicates that for each increase in 1 for the variable in question, the odds of being in one more category up increases by 1.03. So an odds ratios of 1.12 for, say, male status, means that the odds for a male of being in the next category up on the maternal/paternal God scale is 1.12 times higher.

I use a variety of indicators to measure my concepts of interest. For the GSS years used, race is measured on a three-category scale: self-identified black, white, and other race. Here I use white as the reference category. I use year-fixed effects to control for general secular trends. I measure religious activity using both a measure of private devotion (frequency of prayer), as well as a more public one (self-reported church attendance). In the GSS, frequency of prayer (from “several times a day” to “never”), level of theism (strongly agree that “there is a God who concerns Himself with every human being personally” to strongly disagree), and self-identified level of religiosity (“very religious” to “not religious”) are all ranked 1=highest, so I have inverted those scales to make them more intuitive.

Level of fundamentalism is broken down into mutually exclusive dummy variables of self-identified fundamentalist, moderate, or liberal, with no religious identification as the omitted reference category. Religious denomination is broken down into Protestant, Catholic, and Jewish, with none and other as the omitted reference category. People who, on a measure of whether they believe in God select “no way to find out,” are indicated by a dummy “agnostic” variable, whereas people who state that they don’t believe are measured by an “atheist” dummy variable. Finally, attributes ascribed to God are measured on a scale similar to that used for the mother/father God variable, with two antipodal characteristics placed on a scale: God as a judge or lover, God as a friend or king, and God as a master or spouse.


Table 1: Gendered God image (odds ratios)








Model 1

Model 2

Model 3

Model 4

Model 5











































Years Education
























Other Race




































Religious Person




















































































































































































Exponentiated coefficients
Standard errors in parentheses
* p < .05, ** p < .01, *** p < .001

As noted previously, I report odds ratios. The asterisks indicate how significant the results are. In the social sciences, generally anything that is significant to the .05 level (in this table, anything with one asterisk or more next to it) is considered “statistically significant” enough to make the case that there’s probably a non-chance association. In each regression I control for different sets of variables.

Some general, cross-categorical trends can be identified. Income do not appear to have an effect once other measures of theological predisposition are controlled for, age has a small effect, with older people slightly more likely to identify God as at least part female. Interestingly, females do not actually appear to be any more inclined to believe in a female God than men do, and there is no relationship with education.
Self-identified blacks appear to trend towards seeing God as more female. Private religious activity as measured by prayer frequency and public religious activity as measured by church attendance does not appear to show significant relationships with any category, but self-identified religiosity is negatively associated with seeing God as male.

Unsurprisingly, people who self-identify as fundamentalist are less likely to see God as female or equally gendered (with the reference case being people who do not identify with any religion).

Interestingly, people who see God in theistic terms, as one who is involved in daily lives, tend to see him as exclusively male (the GSS directionality of that measure is reverse from that which is intuitive, so a higher theism score is actually less of a theistic belief, etc.); also interesting is that agnostics are more likely to see God as male. There does not appear to be any relationship between denominational affiliation and perception of God’s gender. People who see God as more female than exclusively male tend to see God more in terms of a lover, friend, and spouse, instead of a judge, king, and master, respectively. 


After the dust settles, what leads people to accepting a less masculine image of deity, either towards an equally gendered, a slightly less patriarchal, or a feminine God? A reflection of the liberal/conservative theological continuum is readily apparent here. There are still some idiosyncratic findings that don’t fit readily into any picture (e.g. agnostics being less likely to see God as part female).

The fact that people who are more likely to see God as personally involved in their lives are more likely to see him as exclusively male perhaps reflects unaccounted for variation in the liberal/conservative continuum. It could be that people who see God as a proximate, personally-involved influence in their lives tend to come from traditional theological backgrounds that characterize God as male, whereas more abstract notions of God stem from a more liberal theological predisposition that is associated with seeing God as female. However, since such notions are in part what characterize a more liberal theological disposition it is not enough to simply ascribe it to theological “liberalism.”

Also, such variation would also probably register on the measures of religious activity, with more formally religious active people tending to be associated with more traditional mindsets, but this does not appear to be the case, as church attendance does not appear to be significantly related to any particular notion of God. Additionally, private religious devotion to God, as measured by prayer, does not appear to be associated with the gender of whomever they are praying to.

The strong relationships between the gender of God and her characteristics suggests that the gender of God is not an incidental, dichotomous category, but is intertwined with fundamental notions of what God is, both in terms of roles and characteristics. However, the lack of relationship with religious affiliation indicates that these differences transcend denominational distinctions. One could speculate about specifics on how the different theologies of Jewish, Protestant, and Catholics (the three major categories examined) could lead to differing ideas on the gender of God, but any such relevant differences that do exist appear to be subsumed by broader frameworks about the fundamental nature of God.


Even though most of the United States population has a religious heritage grounded in a masculine version of deity, a surprisingly large number of Americans are willing to consider the divine feminine as part of their theological worldview, even if traditional Judeo-Christian theology does not.

Mormonism, as a belief system that does include an explicit doctrine of female deity, presents a unique case. While the number of Mormons in the GSS is too small to draw conclusions from for this one year (N=25), their mean response was slightly lower—though not significantly so—than that of the non-Mormons (1.48 versus 1.97, respectively) on my four-point recoded scale, meaning that Mormons were slightly more likely to view God as predominantly or exclusively male.  One might surmise that the possibilities for the divine feminine explicitly opened up in various scriptures might lead Mormons towards having a greater sense of the divine feminine, or alternatively one could just as easily suppose that there’s in general a disconnect between a not-often discussed divine feminine within Mormonism and actual discourse about God within the Church: only a larger sample size will tell that tale.


[1] Benko, S. (1993). Virgin Goddess: Studies in the Pagan and Christian Roots of Mariology . New York: Brill Academic Pub. [Back to manuscript].

[2] Mollenkott, V. R. (1984). The Divine Feminine: The Biblical Imagery of God As Female. New York: The Crossroad Publishing Company. [Back to manuscript].

[3] Catholic Church. (1994). Catechism of the Catholic Church. (L. E. Vaticana, Ed.)Libreria Editrice Vaticana . Doubleday, 239. [Back to manuscript].

[4] Coleman, K. S. (2005). Who’s Afraid of “the Goddess Stuff”? Feminist Theology: The Journal of the Britain & Ireland School of Feminist Theology, 13(2), 217-237. [Back to manuscript].

[5] Paulsen, D. L., & Pulido, M. (2011). “A Mother There:” A Survey of Historical Teachings About Mother in Heaven.“. BYU Studies, 50(1), 78. [Back to manuscript].

[6] Hiltebeitel, A., & Erndl, K. M. (2000). Is the Goddess a Feminist?: The Politics of South Asian Goddesses. New York: NYU Press. [Back to manuscript].


Full Citation for this Article: Cranney, Stephen (2014) "Perception of God as Paternal and/or Maternal in the US Population," SquareTwo, Vol. 7 No. 1 (Spring), http://squaretwo.org/CranneyDivineFeminine.html, accessed <give access date>

Would you like to comment on this article? Thoughtful, faithful comments of at least 200 words are welcome. Please submit to SquareTwo.

COMMENTS: 0 Comments