All material on this site is (c) copyright to the respective authors.   ISSN - 1481-3440

The AnthroGlobe Journal

An initiative to broaden international electronic communication in Anthropology


Ritual as a Broken Mirror: The Relationship Between Ritual and Grid/Group Analysis in Religious Groups

Copyright © Edward Croft Dutton 2006

University of Oulu

First posted: 28 February 2006


This article will examine the relationship between the broader structure of a religious group and the nature of the group’s rituals. As such, it will take issue with a fundamental assumption at the centre of Mary Douglas’ ‘Grid/Group Model’ – that the degree of a religious group’s Grid and Group can be understood by an assessment of the group’s ritual. The article will examine Douglas’ work in depth and will also examine an alternative analysis offered by Victor Turner in which he argues that anti-structure is noted in the rituals of highly structured groups. Drawing upon fieldwork with two student evangelical groups, the article will demonstrate that Douglas is mistaken in her assumption. It will demonstrate that the more structured a group is, the less structured its ritual is likely to be, at least in the religious groups assessed. Thus, the results will be shown to validate Turner’s model of religious group operations though it will be demonstrated that Douglas’ model can still be of some use in assessing ritual in itself. Thus, it aims to contribute to a more accurate understanding of the relationship between ritual and structure in religious groups.



Mary Douglas’ “Grid/Group Model” has contributed significantly to the Anthropology of Religion and Religious Studies in general. While Douglas applied the model mainly to tribal religious groups, there have been attempts to apply it to many others with some researchers even attempting to understand types people by means of a development of the model. (E.g: Thompson, Ellis and Wildavsky, 1990). However, Douglas’ model is not with out its critics, not least those who would question how easily the model – which distinguishes between four different kinds of society – can be applied to contemporary, post-industrial societies or how easily it can be applied at all. (Eg: Spickard, 1989 and 1991). Though this problem is appreciated, it is another, perhaps more fundamental, difficulty, in relation to Douglas’ model, which this article will examine.

This article will draw upon fieldwork with two student evangelical groups which, it will maintain, are of the appropriate size to be assessed by the Grid/Group Model.[1] It will look at Douglas’ model in detail and note the implication, central to the Model, that the degree of Grid and Group in a religious group can be ascertained by examining the rituals of that group. The central question that the article seeks to answer is, ‘Is it the case that a group’s ritual directly mirrors that group’s structure?’ The article will demonstrate the difficulties with Douglas argument in this regard. Indeed, it will further show that the degree of structure and control in student evangelical groups would appear to repudiate this finding. It will, therefore, be suggested that it is difficult to use Douglas’ model in the manner in which it has been used but that it can be useful in analysing individual rituals. It can be helpful in providing us with an indirect understanding of a religious group’s Grid and Group in relation to their rituals. However, it will be argued, in the light of the discussion, that Victor Turner’s model appears to be a more accurate means to understand this aspect of religious groups. Drawing upon these discussions, future research will be recommended.

Douglas’ Grid and Group Model

Under-pinning Douglas’ model is the Purity Rule. That is to say, the greater the degree of a control a group asserts over individuals, the greater control it will assert over their bodies. (Douglas: 1970: 42) Douglas contends that a group’s nature can be understood through analysing the extent to which the group is ordered, or its Grid, and the pressure exerted on the individual by the organisation, its Group. ‘Grid,’ then, is defined as ‘order, classification . . . the group’s symbolic system’ (58) while ‘Group’ is defined as ‘the pressures or demands made on the individual.’ (59) Douglas expands these definitions arguing that Grid relates to the extent to which an organisation is systematised and classified. These classifications will remain stable and relatively unchanging unless they are threatened by, for example, new ideas. Thus, the system of classification is sustained by pressure being exerted on individual members to conform to the system of the organisation, that is to say, the level of Group. A high level of Group helps to insulate the organisation’s system against change from outside, ensuring that it remains stable. As such, Grid controls the individual but he can evade this control if there exists a low level of Group. An organisation, analysed at any one time, reflects the relationship between its Grid – the extent to which the organisation is defined and classified – and its Group – the pressure or control which the organisation exerts over its individual members. (60)

I would suggest that Douglas’ understanding of Grid is not entirely free from difficulties in and of itself. On the one hand, it is understood as the symbolic system of the group. It is understood as ‘order’ in the group and this order can itself control the individual. We might suggest that it is impossible for a group to exert pressure unless the group is ordered. Indeed, we might suggest that the very fact that it is ordered at all implies a high degree of pressure and if order causes pressure then it is very close to Group. As such, there appears to be some blurring between Grid and Group if we follow this definition. Perhaps a more helpful way to define these categories is to argue that ‘Grid’ should be understood as what the group conceives whether in terms of symbolic belief or belief about behaviour and ‘Group’ relates to the extent to which the organisation persuades or pressurises people to conceive. This eliminates the points of crossover and defines the difference more clearly.

However, Douglas firstly discusses ‘restricted codes,’ arguing that groups with common values develop restricted codes that are collectively understood while more elaborate codes are used to outsiders. (55) She draws upon Bernstein who has discussed restricted codes in detail. (56) Bernstein contends that industrial society compels its members to indulge in increasingly elaborate and articulate codes if they are making important decisions. Hence, the speech patterns of the industrialist class would be more articulate than those of the working class. (56) Douglas applies Bernstein’s analysis to religious groups. She argues that tightly controlled groups tend to use a restricted code when communicating with each other while employing a far more elaborate code when communicating with outsiders. (55) She also looks at ritual as a restricted code and argues that the less social coherence is valued in a group the less ritual would appear to be valued. (55) Hence a restricted code, at least in the form of a ritual, would appear to be part of a group that holds certain common assumptions.

There are, however, a number of smaller difficulties with Douglas’ model that we might wish to address. Most clearly, it must be remembered that it is a relative model. A group can be seen as highly classified in relation to another group that differs from it. I would argue that only when groups are very different indeed can it be easily asserted that they are differently classified. Douglas implies that ritual activity can be used to assess the level of Grid and Group. Hence, Douglas suggests that the more strongly structured and controlled a society is, the more likely trance-like states in its members are to be perceived as dangerous. (74) I find the view problematic. I would suggest that trance-like states would be quite acceptable in a highly classified group as long as those states were themselves structured in a broader way. Of course, they might be seen to evidence a lower degree of control than in a group where they are not present. However, the broader problem with Douglas, which this paper will demonstrate, is the problem with using ritual in assessing Grid and Group in the broader society.

Douglas’ implicit view appears to be that assessing a group’s ritual will allow us to discern the level of Grid and Group in that society. She understands ritual to be a ‘restricted code’ – meaning that its meaning is only fully understood by group members. As such, she argues that the more controlled a group is, the more ritual activity we would expect it to have. (13) I have no difficulty with this view. She further argues, however, that tribes will vary in terms of the nature of their ritual. This, she implies, can be used to assess the level of Grid and Group. (55) The main, Western example Douglas offers to substantiate this view is West Indians in London in the 1960s. Their ritual is Pentecostal and involves abandon. (84) She suggests that West Indians are poorly ordered because they have ‘no common provenance’ and few contacts with those in power because there were then, at least, few West Indians in positions of authority. (85) I find this argument difficult. Firstly, it is attempting to apply Grid and Group to a broad group, something that Douglas elsewhere criticises. Indeed, she specifically argues that one cannot apply the Grid/Group model to England, for example. (Douglas: 1978: 3) Secondly, if we limit the group down to members of a certain church – rather than trying to compare West Indians with native English – we might discern different results. We would expect the Pentecostal Church to exert considerable demands over members’ bodies relating to drinking, sex and so forth and it would probably meet frequently. By contrast, the Church of England church, assuming it was not itself evangelical, might meet once a week and exert relatively few explicit demands. However, the Church of England ritual would be controlled and the Pentecostal ritual would be less so. It is precisely this difficulty with Douglas’ Model that my own fieldwork will demonstrate. My fieldwork points to a strongly structured group – such as Pentecostals – needing an emotional outlet through their ritual while a weakly structured group needs some kind of structure in its ritual, as will be discussed below. [2]

Turner’s Model

In the light of the fieldwork, it will be argued that Turner’s model of the operations of religious groups is more accurate. Turner’s model can be noted in his examination of Rites of Passage. For Turner, Rites of Passage involve a ‘passenger’ passing through a gap between two cultural realms – ‘a liminal phase.’ In this passage, he experiences a ‘state of transition’ which differs markedly from his previous pre-liminal or his future post-liminal experience. Fellow-passengers tend to experience a strong sense of togetherness in which social distinctions and structure become less relevant. Turner terms this feeling communitas. (Turner: 1969: 95) This allows strong bonds to be fostered, vital for the survival of the community. (96) For Turner ‘Communitas is where structure is not.’ (126) However, he emphasises (95) that those undergoing the liminal ritual in the Ndembu tribe must obey their masters without question. Thus, Turner is interested in communitas within structured parameters.

Turner also distinguishes between the ‘liminal’ and the ‘liminoid.’ The ‘liminoid’ tends to be noted in industrial societies, in Turner’s view. In such complex societies, people’s work lives are segmented into different groups that might have little actual contact. These workers are controlled, more so than in tribal societies, by structured time and rhythm. The liminoid is understood, by Turner, to be the break from such rigidity in the form of leisure time. Thus, the ‘liminal’ is in some way part of societal structure whereas the liminoid is effectively a break with it. (Turner: 1992: 54-56) Even so, the liminal is not understood to be in itself structured. A graduation ceremony would be liminal while a Pop Concert would be liminoid. Liminality is perceived to be less likely in a fragmented, industrial society. However, Turner admits that the liminoid can be found in tribal societies and that the liminal can be found in industrial societies in the form of Church and even academic rituals. (58) He further notes the way in which ‘today’s liminoid is tomorrow’s liminal,’ (58) citing the way in which Pilgrimage gradually became part of the structure of Medieval Christian life while still being clearly liminal.

Turner distinguishes between a number of different types of communitas. Firstly, there is existential-spontaneous communitas. Douglas Davies gives the example of the Pentecost in which a strongly bonded group spontaneously spoke in tongues. (Davies: 2002: 123) Second is NormativeCommunitas. Here there is an ordered group in which communitas helps to create bonds and the communitas is thus within a structure. Finally there is Ideological Communitas, in which a utopian group attempts to achieve a constant sense of oneness. (Davies: 132) It is, I would argue, normative communitas which relates most clearly to the groups examined in this article. Turner’s argument is that communitas occurs within a structured group and this can equally be noted with his discussion of the liminoid. It, in essence, an emotional outlet for an otherwise highly controlled group. Therefore, Turner’s conclusions stand in contrast to Douglas’. For Turner, a group’s ritual is only an indirect reflection of the degree its Grid and Group, in Douglas’ terms.[3]

Field-Work: Student Evangelical Groups

My own fieldwork was with members of two similar student, evangelical groups: Oxford Inter-Collegiate Christian Union (OICCU) and Navigators Leiden Universiteit (NSL). Both of the groups were the largest evangelical group operative and its university. In both cases, their stated aim was to evangelise to other students and to create a community of believers within the university. In general, evangelical groups are noted for a conservative understanding of doctrine and lifestyle ethics, though worship style, as will be seen below, can vary substantially.[4]

I participant observed OICCU from April to June 2003 and again from January to March 2004. I was a participant observer with NSL from September to December 2003. This involved attending many of their meetings and interviewing and surveying samples of their membership. The Grid/ Group Model was employed when analysing the data and the result was clear. OICCU had the highest degree of Grid and Group while NSL had the lowest degree. But, drawing upon Douglas’ understanding of the nature of Grid and Group, OICCU meetings revealed the lowest level of Grid and Group while NSL meetings appeared to highest level. As such, the results seemed to point to a considerable difficulty in Douglas’ use of Grid and Group. Let us, therefore, examine these results in greater detail.

A variety of factors were used to ascertain the degree of Grid and Group in each Christian Union and in their meetings. To understand the degree of Grid and Group in the group as a whole one can look at member’s religiosity and life-style views and attitudes to clothing and language. To understand it in the actual meeting one could examine the degree of religious experience. The following analysis will focus on three of these areas: religiosity, life style choices and clothing. In each case, OICCU was shown to have the highest level of Grid and Group and NSL the lowest.[5]

The Group: Religiosity and Lifestyle

My interviews with sample members of both OICCU appeared to indicate, to great extent, a considerable level of conformity of belief with regard to the issues that appear to be left open to debate. Obviously, in conducting these interviews, I ensured that the samples were representative. They were weighted in terms of various issues that I had examined in random surveys such as religious background, area of origin, gender, where and how they had become a Christian and so forth in order to ensure a balanced sample. Drawing upon these various categories, I was able to develop a highly representative sample in terms of whom I chose to interview. If the person had not previously been surveyed, I would begin the conversation, normally at the CU meetings, with these certain questions of category and, if the person was broadly a part of particular required category, I would arrange to continue the interview at some future point. In each case, I conducted an informal interview based on certain themes, which I had planned in advance to discuss. As both groups had around one-hundred active members, I concluded that twenty-five was an appropriate number to interview in depth.[6]

All those whom I interviewed from both claimed to believe in Hell and to believe that Non-Christians would, in fact, go to Hell. All of those to whom I spoke claimed to believe in the Devil as an actual force in the world. 24/25 in OICCU rejected Evolutionary Theory and all believed in the reality of doctrines such as the Resurrection. One also noted conformity in terms of social belief. In OICCU, all thought it was acceptable for Christians to drink but none to become drunk. Only one felt it was acceptable for Christians to smoke, only one felt it was acceptable to date a Non-Christian and none felt it acceptable to take drugs of any kind. All felt premarital sex was unacceptable. OICCU thus point to a high level of Grid and Group in these specific areas.

One also noted a substantial degree of Doctrinal conformity amongst NSL members. Of a sample twenty-five, all believed fundamental Christian doctrines to be literally true. However, other results were slightly more diverse. Five believed in Evolution, five were not sure and fifteen rejected it. 24/25 believed that Non-Christians would go to Hell but many found this very difficult to express. They emphasised that it was up to God, that it was not their place to say or even that they did not like the word ‘Hell’ and would prefer to call it something else. Many others claimed they had difficulty believing the doctrine in question. Almost all eventually assented to it but only after much discussion. In relation to social belief, there was also more diversity. All felt it was acceptable for Christians to drink and two felt it was acceptable for them to get drunk. In my experience, though, many NSL members who claimed it was unacceptable to get drunk did become what I would term drunk in my presence. Thus I would suggest that the Dutch understanding of ‘drunk’ is perhaps what the British might term ‘very drunk.’ Twenty-three felt it was acceptable for Christians to smoke, in fact most were amazed that the question was asked at all. None would date a Non-Christian. Also, six felt that sex before marriage was acceptable and small minority of male members who claimed it was not admitted to having had premarital sex which was never admitted to in my other groups. Indeed, four felt it was acceptable for Christians to smoke marijuana. Hence, in general NSL’s views appeared to move in the same direction as our other groups on religious and social issues. They would appear to be generally more liberal except with regard to dating Non-Christians which all felt was unacceptable and unwise. (However, the term was defined less exclusively). Thus, NSL would appear to evidence a lower level of Grid and Group than OICCU. Certainly, in terms of social belief we can see that NSL’s are far more in line with most students. Many were also happy to become drunk even if they did not regard it as being drunk.

As such, we are able to note that, in terms of religiosity and life-style, OICCU has the highest level of Grid and Group while NSL has the lowest. Following Douglas’ Model, this finding would lead us to predict that OICCU’s meetings would be the most structured while NSL’s would be the least so. Certainly, we would expect NSL’s meeting to involve the highest level of religious experience or communitas. This, however, was not the finding of the field-work.

The Meetings

A descriptive analysis of the main meeting in each case will demonstrate the degree to which ones findings do not fit in with Douglas’ Model.

 OICCU: 7:14

OICCU had two main meetings in the week. The largest, which occurred on a Wednesday and started at around eight o’ clock, was entitled 7:14, after a particular Bible Passage.[7] This meeting would tend to have large numbers in attendance, often around one hundred and fifty. As with AUCU, this tended to represent the entire active membership. It was conducted inside a particular Oxford Church, St Aldates, which is Church of England. This was an evangelical church and perhaps for this reason OICCU simply used the existing layout and technology. Stackable chairs were placed in a crescent shape facing the electronic band and the lectern at the front. To one side of the focus area was a table upon which there was OICCU and Christian literature and on the other side, at first, would be positioned two members handing out that evening’s programme. At the back of the room was a computer system. A member would sit there and control what was seen on the various computer monitors which hung from the ceiling. At first it would be OICCU advertisements and Biblical Passages such as 7:14. Eventually it would be lyrics to the relevant hymns. At the front of the room, above the lectern, hung a large banner from the ceiling upon which was an image of girl in modern dress reaching for the sky. On it was the quote, ‘My house shall be called a house of prayer for all nations.’ (Isaiah 56:7)

Members would make their way into the church, pick up the programme on the way in and sit down, often but not always in certain groups. At this point, the band would still be practising and thus fragments of certain songs could be heard. Eventually, as more and more members arrived, the band began to articulate some kind of tune. Eventually, the meeting would be addressed by an Exec member from the lectern at the front. He would conduct a short prayer in a somewhat formal style. He would then make a number of introductory remarks often of a somewhat humorous nature of the kind we will later discuss. Thereafter the hymn singing would begin. In ones experience, the first hymn would always be rather lively – such as Lord, Reign in Me - and much of the audience would respond in a very active way, though not all. Around twenty would be noted to dance, jump around, move into the gangway and dance there or dance at the front. Certain members, perhaps ten, would be noted to exclaim certain phrases during instrumental asides such as ‘Thank you Jesus.’ The same people, in the same context, would also be noted to speak-in-tongues. The glossolalia tended to not be especially loud, however, and it was never interpreted. The next hymn would usually be slightly quieter which again appeared to elicit a dramatic reaction from certain audience members. Perhaps fifteen might sit down, close their eyes, sit on the floor and even cry a little. Again one could hear some members exclaiming certain religious words. Usually there would be four hymns. At no point was there silence because the band would continue playing between hymns. One noted the greatest degree of what we might call Charismatic activity during hymns with dramatic changes in tone, such as ‘The Spirit of the Sovereign Lord’ and this will later be discussed.

When the singing had finished people would sit down and there would be some talking momentarily. The President would then begin by making a series of announcements and also a few relevant jokes. Another Exec member, always the same member, would then make the remainder of the announcements. The President would then comment on the final announcement by saying, ‘Awesome’ or ‘I’ll be going to that for sure’ or whatever would be pertinent. At this point a CU member would be invited to the front to read a certain Biblical passage which would flash up on the computer monitors. He would sit down. On one occasion, before the speaker was introduced, a new member, who had recently become a Christian, came to the front to share his testimony, but one must emphasise that this only occurred once. Usually, at this point, the President would then introduce the invited speaker. This was either minister or, in one case, a non-ordained Missionary. The speaker was always male and it would be ensured that he was an evangelical. In general, he was a conservative evangelical but on three occasions during my observation the speaker was from a Charismatic church. He would pray for him briefly and in a quite a formal manner. The President would sit down and the speaker would begin. This would normally last for between forty-five minutes and an hour. When he had finished he would conduct a prayer and sit down. Speakers spoke on a variety of evangelical topics. One minister from Wimbledon spoke on the importance of remaining a Christian even in a ‘worldly’ context for example. There was no opportunity to field questions after the speech. The speaker would be thanked by the President who would then invite a member to the front to conduct another prayer. In ones experience, this would almost always be spoken in a very informal style though not always. Finally, there would a series of further notices given by the President who would also remind people about the Saturday meeting ‘Powerhouse.’ The meeting would always then end with the either the President, or a certain member of the Exec, asking members to, ‘Talk to someone you don’t know before you leave.’ Members would then talk to each other for around twenty minutes and also pick-up various leaflets. They would then leave the Church and many would make their way to College Group, a small meeting of CU members from a particular College with between five and fifteen members on the whole.[8]


The other meeting occurred on a Saturday and began at 12:30. In general, it lasted around two hours and was in a hall attached to New Road Baptist Church. The room itself was almost completely empty. It was not decorated in any way. Chairs were stacked at one end. As one entered, there were tables on one side upon which there were sandwiches and squashes which attendees were welcome to eat. Thus, members socialised together for around half an hour before the actual religious activity began. The Exec member in charge of ‘Powerhouse’ would then invite people to finish their lunch and she would make a few announcements. By this time there would be around thirty people present though one or two would arrive late. A hymn would then be projected onto the wall and a guitarist would begin to play. It was always the same guitarist. Members would stand and sing in the main though around four would like lie on the floor or sit if they so desired. As stated, there were no chairs provided upon which to sit. The group would sing around three songs during one would observe even greater degrees of Charismatic activity than at the other meeting. Members could clear be heard to makes religious exclamations and speak-in-tongues. Many could be seen to get very close to tears and would be, often, very physical in their dancing. This seemed to be the case, however, with the majority of attendees. When the singing was finished members would be left standing or sitting and appeared to take some time to calm down. Eventually there would be a very long silence and soon most members would be standing with their eyes closed. Seemingly at random, a particular would then make a statement about God or what He wanted or conduct a prayer. If it was a prayer, nobody would say there ‘Amen.’ There would simply be a long silence until another member either prayed or prophesied. The language used here tended to be somewhat informal. This would continue for some time until the woman in charge of the meeting invited everyone to sit down.

The Exec member would then invite a particular group member to tell the group about a particular event or Mission that they should pray for. The group would then divide into small groups and pray, at random, for certain aspects of the Missionary activity or whatever it might be. The format tended to be similar to that Aberdeen. Hence a prayer would be made ending in ‘Amen’ and everyone would say ‘Amen.’ The level of formality in prayers varied in this narrow context. At this point, prayers would be brought to close and members of individual colleges would stand to tell the group about the problems with their college’s CU and their college more broadly. Each prayer group would then be allotted a particular college for which to pray and told exactly what they had to pray about. They would then do so. When this has finished other CU members ask the small groups to pray for seemingly quite random things such as future CU events or even a particular boat race that a CU member might be involved in. Thereafter, the Exec member brings the prayer groups to a conclusion and makes a number of announcements. There is then one final hymn, for which members stand. They then sit for a final, relatively formal prayer. The meeting ends here and members begin to converse with one another and generally socialise.

NSL: Lezing

NSL only met once a week. On alternate weeks they would either have the Lezing (a meeting of the whole group in a church hall) or the Dispuut (‘Argument’) meeting, the meeting of a subgroup the members of which were all interested in a certain topic. These meetings would always occur on a Thursday evening. The Lezing (‘Reading’) occurred in a church hall and began at around half past six. On the whole, around one hundred and fifty people would attend. As one entered the church hall, there would be a table to ones right upon which there would a great deal of Christian literature, most of it, unlike in OICCU, for sale. To the left was a structure similar to a pub bar, behind which was a kitchen in which female members were busily cooking that evening’s meal. The rest of the room was taken-up with set tables, around fifteen in all. The room itself was bare except for a large number of climbing plants and vases of flowers. There was no Christian imagery on any of the walls. In one corner of the room, opposite the door, was a piano and in the middle of the wall was a projector screen. Members would make their way into the room and begin to socialise with each other, often purchasing coffee from the bar. Eventually, at the signal of a member of the Bestuur, members would make their way to a table, often with a group of friends, sit down and continue to speak. A Bestuur member – the Organiser - would then climb to the middle of the stairs, which led to the church hall’s balcony, and call for people to listen. He would then make a number of announcements and conduct a prayer. It was always the same Exec member who would do this. Everyone would say ‘Amen’ in response. Further members would climb to the middle of the stairs to make announcements about various up-coming events and finally a certain female member of the Exec, the Secretary, would conduct a prayer especially in relation to the meal that members were about to have.

When she finished members, table by table, would queue for their food and pay. The vast majority present would be conversing at this point. When they had finished their main course almost all members would return to the bar to get something for dessert. When everyone had finished their meals, at the indication of the same male Exec member who had conducted the first prayer, there would be a moment’s silence for ‘silent prayer.’ The tables would then be cleared away. Chairs would be laid out facing the projector screen. The chairs were laid out in a large, rectangular block and they were close together. Around ten members remained seated at the bar even during the singing. Members would continue to socialise together and also to purchase bottled beer from the bar and drink it together. Eventually, they would be invited to sit down. A particular female member, Susanna, would go and sit by piano and hymns would be projected onto the screen. Members would then stand and sing them. One did not note anything that might be comparable to Charismatic activity during the singing. Members would simply stand and sing while others sat at the bar, drinking beer. They did not dance, close their eyes or anything of the kind. The hymns, on the whole, were Dutch translations of English language hymns.


The meetings of both OICCU and NSL clearly point to a higher level of Grid and Group in OICCU’s, which is not what Douglas’ model would predict. Both of OICCU’s meeting involve a loosening of structure in the form of religious experience and this is especially the case in Powerhouse. In 7:14, we note the way in which a large number of OICCU members dance, in response to the hymn, close their eyes and raise their arms in the air. A smaller minority make religious exclamations and even speak-in-tongues. Although this is not interpreted, it still indicates a loosening of control which would not be expected in a tightly structured group if we follow Douglas’ model. This can be noted to an even greater extent in Powerhouse. The majority of the meeting is taken up by singing. There are no chairs in the room and, as such, members sit, stand, lie down or position themselves in whatever position is comfortable. We note, to a far greater extent than in 7:14, dancing, emotional behaviour and perceived Gifts of the Spirit. Although there is some structured praying, in small groups, prayer in this meeting is generally at random and on whatever topic the member wishes to prayer for. In this context, we also note religious exclamations and even glossolalia. Douglas argues that such behaviour is the sign of Low Grid and Low Group society and yet we observe it in the ritual of a relatively highly classified organisation. Indeed, the entire meeting, including the lay out of the room, can be seen as evidence of a low level of structure if we follow Douglas’ model.

Although NSL evidences a far lower degree of Grid and Group than OICCU, its main meeting is far more structured. Like OICCU’s meetings, there are allotted speakers and public prayers. However, it is the hymn singing and corporate prayer that is perhaps of greater interest. Prayer occurs in silence. There are no religious exclamations, Gifts of the Spirit or any other examples of loss of control. The hymns themselves also evidence a high degree of control. There is no emotional behaviour, no dancing, no closing of eyes, no religious exclamations, no glossolalia. The hymn-singing is highly controlled and this is, perhaps, augmented by the differences in the nature of the music provided. Whereas OICCU’s band is composed akin to a rock band, NSL are accompanied by the piano. There is no attempt to stir up emotion in the latter group.

These results empirically contradict the implicit predictions of Douglas’ model. I do, however, think that Douglas’ model could certainly be drawn upon in analysis of religious groups. We have already noted the difficulties with using Douglas’ model in understanding the degree of structure in a society using their rituals. But her model could be used simply to assess a group’s ritual. One could examine the degree of Group in a ritual by looking at the control apparently exerted over members’ bodies. Hence Douglas’ model could successfully be used to analyse rituals. In such circumstances it would be quite to claim, as Douglas does with regard to societies, that trance-like would imply that the ritual had a lower degree of Group than another ritual. Independently, through discussion with the members, one could also assess the level of Grid and Group in the organisation more broadly. Assessing an organisation’s ‘Group’ through ritual analysis would, it would seem, only indirectly tell us about the overall level of control and structure in that group. As such, my results would appear to be more congruous with Victor Turner’s model, as has previously been discussed. OICCU, as an organisation, has the higher level of Grid and Group. It, therefore, requires an outlet – an element of anti-structure – and this can be noted in the degree of communitas in OICCU’s rituals. NSL, as a group, has the lowest level of Grid and Group and, therefore, requires a greater degree of structure in its rituals.


The results of analyses of OICCU and NSL demonstrate that there is an important difficulty in applying Douglas’ Grid/Group model to, at the very least, religious groups. As already noted, Douglas’ implicit assumption is that one can understand the degree of Grid and Group in an organisation by assessing its rituals. This paper finds that one can understand the degree of Grid and Group in an organisation through ritual analysis, but only in an indirect sense. It has demonstrated that the more highly classified a group is, the less classified its ritual activity is likely to be at least in the context of student religious groups. As such, the results are most congruous with Turner’s analysis which suggests that the degree of communitas (or anti-structure) will be highest in the most structured and controlled group. The Grid/Group method is, one might venture, useful in comparing the degrees of control in different groups and in different rituals. It cannot, however, be successfully employed in the way in which Douglas wishes to employ it. A group’s ritual is not always a clear reflection of a group’s broader power dynamics.

Future Research

In stating this, of course, it does not necessarily mean that, in all cases, a highly structured group will tend to have unstructured ritual and this leads to a fascinating future area for research. Those who have developed Douglas (Thompson et al) argue that every individual must have a balance between structure and risk. It would interesting to ascertain if the same is true of religious groups. Thus, if a group is highly structured both in society and ritual, is it likely to be unsuccessful. If this group is successful is that because there is some area of group life, other than ritual, which allows for communitas? A detailed examination of such a group would, therefore, be of great interest in allowing to develop Douglas and Turner and understand with greater subtlety the relationship between structure and anti-structure in a religious group from an anthropological perspective.


Bebbington, D. W., (1989), Evangelicalism in Modern Britain, London: George Allen and Unwin Ltd.

Davies, Douglas (2002), Anthropology and Theology, Oxford: Berg

Douglas, Mary (1978), Cultural Bias, Occasional Paper 34, London: Royal Anthropological Institute of Great Britain and Ireland.

Douglas, Mary, (1970), Natural Symbols, London: Routledge.

Dutton, Edward Croft, (2005), 'Crop-tops, Hipsters and Liminality: Fashion and Differentiation in Two Evangelical Groups,' in Journal of Religion and Popular Culture, IX.

Eade, John and Sallnow, Michael, (1991), ‘Introduction,’ in Eade, John and Sallnow, Michael, (eds.), Contesting the Sacred, London: Routledge.

Hammersley, Martyn and Atkinson, Paul, (1995), Ethnography, London: Routledge

Spickard, James, (1989), ‘A Guide to Mary Douglas’ Three Versions of Grid/Group Theory’ in Social Analysis: 50:2.

Spickard, James, (1991), ‘A Revised Functionalism in the Sociology of Religion’ in Religion.

Stoffels, Hijme, (1990), Wandelen in het licht: waarden, geloofsovertuigingen en sociale posities van Nederlandse evangelischen, Free University of Amsterdam, Kampen: Kok.

Thompson, Michael, Ellis, Richard and Wildavsky, Aaron (1990), Cultural Theory, Oxford: Westview Press.

Turner, Victor, (1969), The Ritual Process, New York: Aldine Publishers.

Turner, Victor, (1992), ‘Variations on a Theme of Liminality’ in Turner, Edith, (ed.), Blazing the Trail, Tuscon: Arizona.



[1] These groups were chosen as part of broader research into liminality and communitas at universities. They are of use to this article because they would be, relatively speaking, in different quadrants following Douglas’ Grid and Group Model.

[2] Spickard refers to a 1974 presentation which argued, very specifically, that trance-like states are likely to be found in highly structured groups. [See: Spickard, (1989), p.159, foot-note 11.] However, neither the presenters nor Spickard develop this finding to look beyond trance-like states or to question Douglas’ basic presuppositions.

[3] One might wish to criticise Turner’s somewhat clear-cut understanding of communitas as being an absence of structure. Eade and Sallnow (1991) argue that Turner's view that liminality leads simply to an absence of structure is not reflected in various analyses of Pilgrimage, which Turner uses as an example of a liminal phase. For a detailed critique of Turner's view of liminality as it relates to universities see Dutton (2005). However, this criticism is not germane to this discussion because a group of Pilgrims are not a highly structured group and are often drawn from various societies. What Turner demonstrartes, however, is that anti-structure can be part of the ritual of structured group. This is demonstrated even if he is over-simplistic, as Eade and Sallnow imply, in his language.

[4] For a more detailed discussion of British evangelical groups see Bebbington (1989). For a similar discussion of Dutch evangelicalism see Stoffels (1990).

[5] Obviously, in order to reach the conclusions presented, I drew upon detailed and dense fieldwork notes. I also conducted interviews with a balanced sample of group members and, as part of a broader assessment, took into account various issues such as differences in religiosity between England and Holland. However, for the purposes of this article, the central issue is the relationship between ritual and group structure.

[6] For a more detailed discussion of fieldwork method see, for example, Hammersley and Atkinson (1995).

[7] II Chronicles 7:14: ‘If my people who are called by my name should humble themselves and pray and seek my face and turn from their wicked ways then will I hear from heaven and forgive their sin and will heal their land.’ This is, according to OICCU members, a very well known verse amongst Evangelical Christians.

[8] No analysis could be made of these College Group meetings because the Exec did not wish me to attend. They felt that my presence in such sparsely attended meetings would be intrusive.


All material on this site is (c) copyright to the respective authors but may be copied or printed FOR PERSONAL USE.