Keywords

As we reach the end of this volume, it is time to formulate some conclusive remarks. In fact, all of the approaches presented in the various chapters of this book have sufficiently demonstrated that they do not lead to conclusions; on the contrary, they open a wide range of stimulating questions and hypotheses that await testing. This does not mean that the efforts invested by so many researchers from different countries to better understand how children represent godsFootnote 1 have been merely exploratory. The primary purpose of summarizing the overall project here was precisely to overcome the mere juxtaposition of a variety of approaches in the representation of supernatural agents in children. This is why we introduced the construction of an integrative model at the beginning of this book (see Chap. 2, this volume). The main perspective of this integrative model is developmental: it focuses on the child’s development of representations of gods. The need to write such a chapter, as to lead the whole project on children’s representations of gods elsewhere, is twofold. First, in the field of psychology of religion, research on children is underrepresented. Second, talking, in general, about children’s representations of gods means nothing if it is not conceived in a developmental perspective, simply because children’s concepts change considerably between the age of four and the age of fourteen.

After briefly recalling some of the main results presented in the book, this concluding chapter discusses successively the contributions and limitations of using drawings as a means of studying children’s representations of god, then the benefits and limitations of interdisciplinary approaches. Finally, it presents four future directions of research.

Main Results for the Study of Children’s Representations of God

We do not pretend to summarize exhaustively in a few lines all the results presented in this book. We will only mention some salient points in connection with the project Children’s Drawings of Gods.Footnote 2

Let’s start by recalling that the integrative model presented at the beginning of the book is largely based on the research described in subsequent chapters. We placed the integrative model at the beginning of the book as an invitation to the reader: a means of engaging the reader’s interest in the other chapters by introducing their topics and providing additional interpretation of their results. We provide substantial documentation and information to developmental psychologists interested in religious development, and to teachers involved in the religious education of children at schools and in various religious contexts.

A theme that has accompanied us throughout this research is that children who draw god do not simply produce a spontaneous pictorial representation of a supernatural being, but rather, they compose their drawings with reference to representations already available in their social environment. In itself, this observation is not new. Harms (1944) had already shown that from the age of twelve, when asked to draw god, a substantial number of children and adolescents will draw representations typical of the religious traditions with which they are familiar, or to which they belong.

However, a main finding of these studies shows that it is not necessary to belong to a religious tradition and to receive an education from it in order to create religious representations graphically. The ability to conceive supernatural beings and to produce iconographic representations is universal. Children who are not in contact with religious traditions easily find in the media (films, cartoons, comics, video games, etc.) sources of inspiration for representing god.

Another finding worthy of note is the impact of religious education. In the Children’s Drawings of Gods project, we take care to collect drawings in the public school setting (in order to reach an all-round population), as well as in the setting of religious education (e.g., in parishes or in religious education classes at school). However, we found that these settings did not serve as effective predictors of religious socialization. We could not, based solely on the setting of the data collection, predict which children would have strong religious socialization and which would not. On one hand, children from very religious families may attend public schools; and, on the other hand, children taking part once a week in religious education classes at school may not be distinguishable from other children in the same school who do not participate in such classes. More in-depth analyses have yet to be conducted on the data that has been collected. Let us note for the moment that, analyses of the Swiss sub-collection seem to indicate that receiving religious education increases the tendency to produce an emotionality intense drawing, which is probably linked to familiarity with the concept of god. This does not mean that the child who receives a religious education will lose all critical capacity and then reproduce only stereotypical religious representations. On the contrary, thinking about a concept can help to distance oneself from naïve and spontaneous representations. De-anthopomorphization and ambivalence are two strategies, among others, that children can use in their drawings in order to allow anthropomorphic features to express that gods are intentional agents, and at the same time not simply draw figure that is merely human (see Chaps. 3 and 4, this volume).

A number of findings highlight the impact of culture. Cross-cultural analyses indicate some differences in the hierarchy of colour preferences for drawing god, and at the same time they expose the dominant role of the colours yellow, (also orange and red in Japan). It seems that the idea of light is universally associated to the concept of god, and that yellow, sometimes orange, and red are the preferred colours used to represent light (see Chap. 8, this volume). The impact of the religious background also becomes evident when analysing the gendered characteristics of the god figure (see Chap. 5, this volume). In Western countries, where the concept of god is clearly associated with a masculine figure, feminine representations of gods are rare. However, in Japan or in Buryatia, feminine representations of gods are quite frequent, and they are mainly produced by girls. Other interesting aspects concern the impact of culture on the position of the god figure in the drawing (see Chaps. 6 and 7, this volume). Intercultural analyses have yet to be extended to all sub-collections to better understand the impact of the cultural background not only on the spatial location of god, but also on the emotions associated with god figures, and on the possibility that an attachment bond is, or has been, established with god (see Chaps. 10 and 11, this volume). Indeed, the same depth of analysis and coding has not be carried out on all of the sub-collections that are mentioned in this volume. There is still work on the board to be done. Nevertheless, at this moment, it is time to present some reflections on the effectiveness of using drawings to study children’s representations of god, and to appreciate the contributions of an interdisciplinary work.

Contributions and Limitations of the Methodology of Drawings for Studying Children’s Representations of God

One of the main strengths of using the technique of drawing as a means of collecting data is its great flexibility and its potential to be used in all types of research (quantitative, qualitative, or their combination). Drawings can also be used with a wide variety of topics, either as a main method of research or as a supplementary procedure. The present volume is a first attempt to demonstrate the diversity of empirical studies that rely on drawing as the main method used to study children’s representations of god. Although the majority of studies presented in this volume used quantitative designs, drawing technique also fits very well, if not better, with qualitative research designs. Whatever the type of methodological design, the present research project, we expect, could demonstrate the value and singularity of this method for exploring children’s imagination of the divine.

Another great advantage of drawing as a tool for research is the richness and diversity of information that can be found in such material. In contrast to language-based methods, drawings are able to reveal the information that is not accessible using techniques such as questionnaires and interviews. All elements of the drawings may be important sources of knowledge: the participant’s choice of character, object, symbol, or process to represent the topic (god, in our case), the choice of colour, the size of objects, their location within the drawing space of a sheet of paper, the composition, the omission of some parts of a figure, and the degree of originality (that is, whether or not a drawing represents a conventional image of the divine that exists in the child’s cultural or religious environment). Moreover, drawings, when combined with narratives, become an even richer source of information. At the same time, the richness of information that is available in drawings also represents an inherent disadvantage of this method. Undoubtedly, a complete and proper analysis of drawings is a very elaborate and time-consuming process. It requires a good knowledge of both the general particularities of children’s drawings and the specific cultural and religious realities in which drawings were collected. In terms of methodological concerns, this can present a challenge for researchers (although the degree of challenge depends on the research questions and the chosen approach). In this regard, we had to make choices. We therefore gave up filming the children while they were drawing. This technique was used for example by Wiedmaier (2008, 2010) and allows very fine analyses of the drawing elaboration process. However, it is too time-consuming when it comes to processing a large amount of drawings. Nonetheless, drawing, as a research tool, serves as an extremely valuable instrument to go beyond, as some researchers said, the “intellectualized, theological, cognitive understanding” of god, and to reveal an emotional and experiential image of god (Rizzuto, 1970; Hoffman, 2005; Gibson, 2008).

Psychological inquiry into religious phenomena would be not complete if the fields of psychology and cognitive science did not address religious images and religious imagery. As Moore (1977) argued, there were images, visual symbols, and pictographs before written texts in the history of ancient religions. Even in religions with a large corpus of scriptures, such as Buddhism, the use of images developed as a parallel tradition, only partly related to the texts. The implementation of drawing as a research tool offers great benefits to the advance of our knowledge in this domain. For instance, through drawings we can better understand the vitality and stability of some religious images, the origins and development of religious symbolic thinking, and the functionality of religious images on both the individual level and the group level. Another important question to which drawing as a method could offer a valuable insight is the question of embodiment in the human’s imagination of the divine. Hodge and Sousa (2018) correctly pointed out that it is commonplace in the cognitive science of religion (CSR) literature to read that supernatural agents are believed to be disembodied beings while the majority of gods across the vast majority of religions are represented as embodied. Unfortunately, this issue is largely ignored in the psychological and cognitive investigation of religion. What is the relationship between the divine and the form in which it is represented? This is an intriguing question and if we do not engage it, we miss an important element in a human’s imaginings of the divine. One critical point of view concerning the use of drawings in study of god representation should be mentioned here. According some authors, the task of drawing god forces children to look for a pictorial form of god and pushes the results in an anthropomorphic direction (Tamminen, 1991; Tamm, 1996). Actually, the tendency to anthropomorphize supernatural agents has been confirmed in numerous studies using different types of measurement such as verbal, or experimental (see for instance, Deconchy, 1967; Barrett & Keil, 1996; Barrett, 1998; Demoulin et al., 2008). Accordingly, the problem is not in the method. Anthropomorphism of supernatural agents is an aspect of a much broader phenomena, that of attributing human characteristics to nonhuman phenomena. We find evidence of this attribution in art and in technological gadgets (Guthrie, 2015; Waytz et al., 2013). Therefore, it makes no sense to exclude this method from the field of the psychological study of religion. On the contrary, it is necessary to expand the research by examining not only anthropomorphism in our (necessarily human) conception of the mind of god (as cognitive science does) but also anthropomorphism in physical, bodily representations of supernatural agents. Along with that, we should remind ourselves that a child’s drawing of god cannot be taken as an adequate and exact reproduction of child’s inner idea of god. Drawing often results from a creative and imaginative process in which a child is involved. The graphic representation also depends on children’s manual skills, his or her aesthetic preferences, the materials used, and the research context and setting. Consequently, children’s pictorial representation should be viewed rather as a metaphor, or as a symbolic representation, that makes visible the essence of child’s idea of god that she or he has at the time of drawing. In this sense, the child’s choice of how s/he represents god can also provide rich insight into individual and group creation and the use of symbol systems. The choices made by the child also reflects the transmission of religious beliefs and practices within the culture.

Another important outcome of using the drawing method is the facilitation of cross-cultural research. In actuality, the field of the psychological study of religion suffers from the same weakness that the whole field of psychological science suffers from: the understanding of human religiosity has largely been constructed on an empirical foundation that was gathered from WEIRDFootnote 3 people (Henrich et al., 2010). According to Hoffman et al. (2007), the majority of god image theory and research assumes the Judeo-Christian worldview, therefore, religious and spiritual diversity need to be addressed. Understanding what is universal or variable about human’s perception of supernatural agency necessarily requires the study of religious data from a diversity of populations and religions. Drawing is particularly useful for collecting data from a great diversity of populations because there is no need to test and validate psychological constructs and instruments designed to measure them. Many of the existing measures of god concept and/or image are limited to a Christian population. The development of new measures demands a considerable effort from researchers and it is not a straightforward task for those who are not experienced in such tasks or who have very limited resources (or none) for the correct execution of all necessary procedural work. This is especially true for researchers working outside of economically developed countries. Additionally, drawing technique involves the minimal use of language and, consequently, eliminates many problems associated with the use of language in data collection and any subsequent comparative analyses. The religious and cultural diversity of data we have been able to collect in the present project clearly confirms this advantage.

Finally, drawing is regarded as particularly well suited for data collection that involves children. It is even considered to be “child-centred” in the sense that it may be familiar to, and even enjoyable for the child. Drawing may be more sensitive to children’s (especially younger children’s) particular competencies or interests (Punch, 2002; Mitchell, 2006). According to Punch (2002), younger children may have a limited and different use of vocabulary and understanding of words, they could have difficulty expressing their views freely while being interviewed by an unknown adult researcher in a one-to-one situation, and they may have a shorter attention span than adult participants. In contrast to interviews or written surveys, drawing can encourage children to be more actively involved in the research, and enable them to feel more at ease with an adult researcher. The drawing method gives to children more control over their experience of the research process as well as over their form of expression. Children have more time to think about what they want depict and they can modify their drawings to better express their ideas (Punch, 2002).

In summary, the drawing method certainly has much to offer to the psychological inquiry of children’s concepts of the divine by proposing new research questions and strategies. Now it is important to continue to explore the potential of this method, to elaborate new types of analysis, and to continue the critical discussion regarding theoretical and methodological approaches to using drawings not only within psychological science but also in dialogue with other scientific disciplines. In many ways, the field of psychology has remained isolated from other social sciences studying religion. Children’s drawings are a particularly relevant subject for interdisciplinary or multidisciplinary research in this domain. Actually, research on how people imagine and conceptualize supernatural agency could benefit greatly from joint efforts of different disciplines, leading to richer outcomes through interdisciplinary collaborative work. This aspect of our project will be discussed in more detail in next sections.

Benefits and Limitations of Interdisciplinary Approaches That Combine Computer Vision, Database Management, and Developmental Psychology

Throughout the project, numerous epistemological, and methodological issues have necessitated strong interdisciplinary collaborations. This is common in research within the digital humanities that also involves computer science, information technology, and design (as demonstrated by F. Darbellay, see Chap. 19, this volume). The present research project gathered specialists in computer vision, database construction, data analysis, psychology, sociology, etc. The collaboration was crucial, first for the database construction. Infrastructuring a huge set of data and associated tools for the multifactorial understanding of what (and how) influences the representation of supernatural agents led us to mobilize specialists of digitization and database experts in order to tackle the specificities of the sources, the materiality of the data, the social conditions of its gathering, and the relevant metadata (see Chap. 18, this volume). It was also important for the digitization and the data treatment of the drawings, themselves. Thus, computer vision specialists cooperated with psychologists to negotiate definitions of colours, their equivalences in various contexts (e.g., is the orange or red used for the sun in Japan equivalent to the yellow used in other cultures?) and their composition (e.g., is green a mix of some yellow and blue which can be quantified or is it a specific colour unto itself?). The interdisciplinary discussion included questions such as: How can we transcribe the colours in automatized recognition of light in drawings? Does the figure of god tend to be located in the upper part of the drawing? Is the figure of god larger in drawings made by children with religious affiliation when compared to children with no religious affiliation? These questions led both to scientific and technical challenges in computer science and to heuristic, methodologic and epistemic challenges for human scientists. Working together to tackle one question often led to additional questions, causing us to revisit discussions regarding the research goals of the various research disciplines. In this situation, interdisciplinary efforts preclude a strictly linear research process. Instead, this interdisciplinary research fosters a situation that encourages researchers to go back and forth between objectives and analysis results, and between theoretical questions and empirical work, via a non-linear path filled with reciprocal interactions and recursive loops. Aside from the classical confrontation of scientific approaches, the translation of the categories from one discipline to another, the gathering of consensus on some joint challenges, the shaping of a shared global research design, and the discovery of new ways to think about individual research objectives, the interdisciplinarity also led us to questions regarding the very research infrastructure we sought to construct, the building of which led us to the interdisciplinary approach in the first place.

The work of D. Vinck and P. N. Oberhauser (Chap. 17, this volume) discusses how the drawings are equipped, i.e., how entities are added to the “raw” material in order to make the data usable for the research project. However, doing collective research, with researchers pursuing different research questions, implies having discussion and confronting the various research orientations in order to organize the data in a way that will be useful for the different research designs. The question is even more important when we consider the amount of data and the associated amount of required invisible work because the research infrastructure under construction could unintentionally embed irreversible limitations. The lengthy work of equipping raw data engages the potential use of the data and, as a consequence, creates both possibilities and constraints for further research. Because the research infrastructure is intended to be used by more researchers with different backgrounds and approaches, its scientific flexibility is a key issue. Thus, the internal diversity of the research team, both inside and outside the field of psychology (along with information technology specialists), and its intentional interdisciplinarity was one way we chose to integrate the expected openness of the research infrastructure and its sharable data among researchers with different backgrounds and approaches. This issue has raised questions and generated interdisciplinary discussion. Interdisciplinarity, here, appears to be both an end goal (the openness of the research infrastructure for new users) and a means to that end (a way to ensure the interdisciplinary character of the large-scale digital infrastructure).

However, just as the interdisciplinarity was a resource, it was also a constraint. Depending on the involvement of researchers from different domains, of their own motivation (taking into account the academic pressure placed upon them), and of the relationship between them (sometimes ambiguous collaboration, at other times epistemic dependence), the resulting infrastructure is also the product of negotiation and compromises which sometimes led to a reduction in future possibilities (limitations). The researchers involved also maintained vigilance regarding the contributions and the advice coming from other disciplines. They did this in part because of the difficulty inherent in evaluating the engagement and robustness of these contributions, regardless of whether they came from partners or employees. Along the way, the researchers from one discipline discovered the internal diversity of other disciplines via dialogue and mutualisation; computer science and psychology are heterogeneous domains crossed by a diversity of streams of thought and different paradigms. Depending of the sub-discipline or the specialty, scientific and technical inputs vary. Being ignorant of, or not having anticipated, this internal diversity, the research team sometimes took one direction, unaware that another way was also possible. Along the way, the challenges of interdisciplinarity initially led to some withdrawal and inflexibility, but through learning to work with other disciplines, and thus cultivating a new understanding of other disciplines, interdisciplinarity led to more openness. We must, however, remain conscious that interdisciplinary activity cannot be forced or imposed by one disciplinary group onto another; it develops rather through the process of mutual discovery and the co-learning of a variety of disciplinary languages that are enriched and can potentially reach beyond a single discipline.

Interdisciplinarity in the project was not only a question of learning to translate categories from one discipline to another one, or to confront different approaches. It was also about becoming familiar with each other’s disciplines, their social and epistemic structuring and the necessary ways to cooperate in order to open or close scientific and technical possibilities. As a by-product, interdisciplinary efforts form researchers and research teams that are able to engage new endeavours and collaborations with some mutual expertise on the importance of such cooperation.

New Perspectives

As we close of this presentation of this research project, we turn our attention to new vistas of research that have been opened as a result of our work. They unfold in four directions that we will briefly describe before concluding.

In the Field of the Sciences of Religions

A first direction of future research takes place in the field of the sciences of religions. Interpreting drawings sometimes required a deep knowledge of how a given religious tradition thematises religious concepts, how it symbolizes them in iconic motifs, transposes them in behaviours, etc. For example: In Switzerland, including earrings in a drawing can be a way of adding a feminine feature to a figure of a girl; in Buryatia and Japan, however, earrings are more likely to be an attribute of buddhist divinity regardless of its gender. Without knowledge of the codes used by the religious traditions with which a child may have been in contact, we may not correctly identify the strategies used by this child to signify the divine. Future work involves the study of the iconographic codes in the cultural environment in which the drawings were collected in order to better analyse what may have served as a model for the child.

In the Field of Religious Art

A second direction for future research takes place in the field of religious art. Starting from a given drawing made by a child, the goal will be to look in the history of art to see if we are able to find similar representations of the divine. In the case of an affirmative answer, we will have to ask ourselves if there is a chance that the child has used this particular work of art as a model for his drawing or if it is better to think that an artist before him had simply proposed the same way of representing god. Such an approach leads us to shed light on the history of the sources of inspiration for the representation of the divine in a given culture. It also informs us about the availability of such models in the cultural environment, and encourages us to consider the permeability of this environment with regard to images coming from other cultural backgrounds. In cases where it is implausible that the child could be inspired by works of art that were previously made, we will then be able to emphasize the inventiveness of the child as shown in his/her drawing. It will be an attestation to the notion that similar ideas can emerge in different contexts. From the perspective of the study of artistic activity, the analysis of parallel strategies used by children and by artists for representing the divine will contribute to the discussion of what conditions are necessary in order to proclaim that a child is an artist (Boone, 2007; Twigg & Yates, 2019).

The Study of Creativity and Creative Processes

This reflection leads to a third direction of research: the study of creativity and creative processes. When a child draws god, does he or she reproduce models available in his/her social environment or does he or she independently imagine the mental image that s/he then draws? In terms of reproduction or creativity, it is not always easy to know if a child’s representation of a supernatural agent is an attempt to reproduce a work of art previously seen by the child or if it is an invention of the child, him/herself. In addition, it is often said that religious iconographies are not really open to innovation (Duborgel, 2004; Lubart, 1999). Therefore, a tension between the processes of reproduction and imagination can be observed through the study of religious drawings made by children. This is why it will be particularly interesting to better understand how children’s creativity is expressed in this context.

Developmental Psychology and Developmental Norms

Finally, these considerations open a fourth direction of research in the field of developmental psychology of religious norms. Religious traditions tend to provide iconographic models and to restrict the production of representations of religious topics to these standards. Therefore, such models, considered as orthodox by these religious traditions, receive a normative status. In the field of developmental psychology, with regard to developmental norms, two main research questions can be applied to children’s drawings of god. The first question: At what age do children identify the normative status of certain religious representations, and how do they take this into account when they draw god? What are the strategies they use in order to respect, circumvent, or transgress religious norms? The second question concerns the comparison between religious norms (from the perspective of developmental psychology) and other kinds of developmental norms. There is a whole line of research on the study of norms from a developmental point of view (see for example Gabennesch, 1990; Nisan, 1987; Smith & Vonèche, 2006). The question here is: Does the child’s interpretation of religious norms and canonical representations present some specificity in comparison to the child’s understanding of norms in general? There is a whole field of research that needs to be explored in an intercultural way, particularly in relation to the prohibitions against pictorial depictions of the divine that are advocated in different religious contexts (Wagner et al., 2005): How do children manage such prescriptions?

Conclusion

We come to the end of a book that offers a variety of perspectives on children’s pictorial representations of the divine. The intersecting points of view made it possible to highlight the richness of a material, such as the use of drawings to study children’s development from an intercultural perspective. Much remains to be done to improve image analysis techniques. The results presented already show clearly how the handling of religious representations is located at the intersection of cognitive, affective, and social processes. The four directions of future research that we have just outlined show that much remains to be done. Let us hope that we will find the means to achieve it.