Participation in Unanimous Decision-Making: The New England Monthly Meetings of Friends
Published in politi.philica.com
This study is the byproduct of a great deal of work, none of it my own. I am indebted to Father Sheeran's excellent book Beyond Majority Rule, which is the best and almost the only careful description of Quaker consensus process by a non-Quaker. The study itself is based on the Herculean work of Frank Bryan and his roving army of college students. Finally, the statistics were compiled over the years by Katharine Lee Clark, the long-time administrative secretary of the New England Yearly Meeting. I would also like to thank Frank Bryan, Thomas Ponniah, and Susannah McCandless for their editing and encouragement.
Unanimity or consensus decision-making systems are those in which every member of an assembly must condone a motion before it carries. Of all forms of radical democracy, this is the one that places the greatest emphasis on minorities and the consent of the governed. Yet despite this theoretical importance, there is very little empirical research on consensus. This paper is an attempt to fill that void.
As it is generally practiced, consensus is not only a form of direct democracy, but it also occurs in a face-to-face setting. Members of the group arrange to meet in person to discuss the matters that concern them, and collective decisions are formally made at these meetings. There are no tiers of representation: each person is solely responsible for bringing their interests and concerns to the table. As such, consensus decision-making resembles New England town meetings, the most famous form of direct democracy, if not always the best understood.
In this paper, I look at participation statistics in one of the best-established communities that uses consensus process. This is the Religious Society of Friends, or Quakers. The basic social unit of the Quaker Church is the Monthly Meeting, so named because it holds a decision-making meeting once a month. These meetings, which in Quaker nomenclature are "Meeting for Worship for Business" form a set of historical events about which we happen to have a good deal of participation data.
Consensus as a Political Method
The Idea of Consensus
As viewed by political science, consensus decision-making is a method in which every member of a group can exercise veto power on a group's collection decisions. It is a species of democracy, the logical extension of the supermajority criteria used in special votes. Consensus is also a species of anarchy: if no one's will is being contravened, there can be coordinated activity but in an important sense there is no governmental authority: there is no coercion of a minority. In this regard, one of the most important appearances of consensus is in the contract establishing voluntary associations. Before majority rule can be agreed to, all groups are de facto consensus assemblies. This initial, brief consensus remains very important to the group's narrative, as the majority can use it to argue, "we all agreed to abide by the outcomes of this particular voting process." Indeed, this use of consensus is so powerful in political discourse that nation-states-which have almost never been formed consensually-often employ a mythology of literal social contracts.
But beyond these uses, political science has tended to treat consensus as an abstract fantasy, not a topic for serious investigation. Even radical theorists like Rousseau, in his day, and Wolff, in ours, admit that consensus is effectively impossible, however desirable it may sound. If a group rejects the classical trinity of monarchy, oligarchy, and majoritarian democracy, it simply ceases to exist from the point of view of political science. Happily, like town meetings, consensus groups do persist and can be studied whether or not the academy believes in them.
The consensus process used by the Society of Friends (Quakers) has been in use for three and a half centuries. It is a highly specific political mechanism, with its own vocabulary, ideology, and traditions. It has to be strongly emphasized that Quakers themselves are very reluctant to describe their consensus process in political terms. Rather, friends tend to view their decision-making process as an integral part of their religious experience. Many Quakers, as well as Father Sheeran (a Jesuit), believe that consensus is so fundamentally spiritual that it cannot be secularized. Wolff agrees, although he allows the theoretical possibility that consensus would work in a secular group with a deeply held ideology-he is thinking of anarchists.
But this concern seems to loom larger in theory than in practice. Consensus is in fact used by many secular groups whose nucleus of shared values seems to be rather small. In particular, consensus has been used by groups like the "spokes-councils" at mass convergence street protests, where there is a common opponent but a very diverse range of positions and interests. It is important to note that consensus can operate in such groups, despite their significant internal conflicts. In light of this, the Quaker disbelief in secular consensus seems unwarranted; perhaps it only mirrors the wider disbelief in consensus of any kind.
I want to invoke the anthropologist's privilege to dutifully note, and then respectfully ignore, a group's self-analysis. However we relate to the spiritual aspects of Quaker unity, it is manifested in a social process that we can observe, and in this case even quantify, to good purpose. At the very least, the behavior of consensus groups should throw some light on less stringent forms of radical democracy.
A Brief History of Consensus
The historical origins of consensus in the West begin with the ascendance of canon law during the "failed Renaissance" of the 12th century. Decretalism, the internal political mechanics of the Church hierarchy, had become an inevitable reality. Yet it was keenly felt to be uninspired, uninspiring, and easily corrupted. At the third Lateran Council in 1170, a voting system was established in which the sanior pars-the most heavily credentialed electors in each district-had to agree unanimously in order for a vote to be considered valid. In later centuries, this principle served as the basis for Church concilar theory. Bishops were, in certain circumstances, constrained from acting without the unanimous approval of the sanior pars.
Quite far from Rome, in many ways, the heretical and revolutionary movements that we now refer to as the Anabaptists developed a similar principle: the "rule of sitting down" (sitzrecht or lex sedentium). When a group of people meet together and agree unanimously on something, the Anabaptists argued, they are expressing the Divine Will. The first concrete appearance of this on the historical stage seems to be at the Martyr's Synod in 1527.
It is possible that these two versions of consensus share a common origin. During the failed Rennaisance, disaffected canon law students were well represented among the wandering scholars called goliardi or the ordum vagorum. By the 13th century, this subculture seems to have blended away into the Cathari and other heretical movements that were undoubtedly the precursors of the Anabaptists. While this is insufficient evidence to posit a connection across three hundred years, the link is tantalizing.
After the 16th century, the broad outlines of how consensus spread can be stated with certainty. The Anabaptist sects were deeply influential on the second reformation sects of the 17th century, above all in England. Quakers became the most successful of these sects (being, in many ways, the most conservative) and found that consensus decision-making was an effective defense against their more zealous co-religionists.
The second reformation was only a religious expression of the wider populist upheaval in British society at the time. It was a time of political and social revolution, and many criticisms were laid against business as usual. Not least among these was the challenge that trial by jury, though established in the Magna Carta, was or had become a simulacrum. Judges made all the decisions, often extra-legally, bullying the juries into compliance.
In 1670, William Penn and several other Quakers were arrested for speaking in London, in an effort to challenge the Conventicle Act. Aware that the Conventicle Act was on shaky ground, the state refused to read the charges, demanding that the jury return a guilty verdict without an indictment. Penn persuaded them not to. The jurors were fined for contempt. One juror, Edward Bushel, refused to pay the fine and was imprisoned "without meat, drink, fire, and tobacco;" apparently he was soon joined by several others. After two months of imprisonment, an appellate court freed the jurors with a writ of habeas corpus, and ruled that jury decisions could not be coerced or subject to punishment.
The "Bushel Case," as it was later known, established jury trials as a foundational element of democracy. They became one of the chief concerns in the 1689 declaration of Rights, and jury trials were one of the key grievances of the American colonists in the late 1700s. Consensus process was and is a basic feature of jury trials, especially those that are considered serious, such as Federal criminal cases in the United States. Strangely, while jury trials themselves have often come under question, their use of consensus-which is relatively unique in common law-has rarely been challenged or even noted in the United States.
In the 20th century, Quakers played critical organizing roles in many cooperatives and intentional communities, as well as in the anti-war and civil rights movements. This provided a great number of possible transmission points to secular activist groups which saw value in conducting business "after the manner of Friends." Consensus appealed to a desire for a more radical democratic process than representative majoritarianism. Moreover, for many young secular radicals, Quakers provided these movements with an ambience of tradition and spiritual authority. Treloar writes that the Quaker roots of consensus "would horrify many non-religious and non-pacifist anarchists." Yet it is my experience that most anarchists are aware of those roots, and find them quaint, not horrifying.
There are four groups that seem to have played especially important roles in disseminating at least the awareness of consensus process throughout the US left. These are the Federation of International Communities (FIC), the American Friends' Service Committee (AFSC), the Clamshell Alliance, and Food Not Bombs (FNB).
In the 1940s, Arthur Morgan, the former chairman of the Tennessee Valley Authority, formed a group called Community Service, Inc. (CSI). This organization became involved with the alternative service work being done by conscientious objectors: primarily Mennonites and Quakers. After the war, CSI became very focused on what we now call intentional communities, and gave rise to the Fellowship of International Communities (FIC). The FIC used consensus, no doubt because it was deeply influenced by Quakers. In the two decades that followed, intentional communities flourished and diversified to an almost unbelievable degree, and many of the FIC's members left to pursue more specific goals. Nevertheless, consensus process was taken up across a very wide range of these communities, and eventually supplanted Skinnerian "planner-manager" systems as the default government for communal life.
During the same period, the AFSC was formed to deal with postwar rebuilding efforts in Europe, but for our purposes it is more important because of its role in the civil rights and anti-war movements. The AFSC began to energetically recruit people from the communities it was working in, gradually creating an organization which embodied certain Quaker values and practices, but was largely made up of non-Quakers, or Quakers who were outside the Society's white-skinned, white-collar, mainstream. This complex, controversial cession of power was, I believe, the main channel bringing consensus "after the manner of Friends" to secular activists.
The AFSC was involved in many campaigns, and AFSC activists went on to play key organizing roles in groups like A Quaker Action Group (AQAG), Movement for a New Society (MNS), and the Clamshell Alliance. The Clamshell Alliance played an important historical role in this regard. It was a mass movement involving thousands of people, and capturing the attention of many thousands more. The Alliance was organized along the format spelled out by the MNS "monster manual," which heavily emphasized consensus process.
The internal politics of the Clamshell Alliance were spectacularly conflicted. MNS had viewed itself as pushing boldly past the Quaker "mainstream." Many participants, however, felt that the Quakerizing insistence on consensus was a bizarre, quasi-religious anachronism, sabotaging the group's efficacy. There were secular anarchists involved on both sides of the consensus issue, and a logomachy about the role of consensus vis-à-vis anarchism began, which continues to this day. Nor was it very polite-anarchists reserve their choicest invective for each other, and the argument at Seabrook was heard at quite some distance.
Shortly following the Clamshell Alliance, Food Not Bombs was founded, also by anti-nuclear activists. FNB is both anarchist-identified and strongly pro-consensus. It represented a full political displacement, as it had no links to Quakerism or any other religious group. FNB soon became a very successful cellular organization, with hundreds of chapters springing up across the country and around the world. Food Not Bombs has become a sort of icon of turn-of-the-century anarchism, and chapters of the group have often become the nexus around which other far-left activist groups organize. They publish Conflict and Consensus, a sort of scaled-down version of the monster manual.
In the last few years, it is arguably FNB's version of consensus that has been widely adopted by groups in the global justice movement. At mass convergences like the Seattle protest, as at Seabrook, there is the doubled challenge of using consensus to organize coalitions that have widely different interests and little history of working together.
It should be noted that the channels of transmission I am describing do not predate World War II. Nordhoff's survey of Utopian societies in 1875 reveals no trace of Quaker business processes, even in the Shaker communities which had Quaker roots. I believe it can be shown that Treloar argues persuasively that the widespread use of consensus by anarchist groups is a very recent phenomenon.
In contrast to this, other authors and activists have identified consensus as the traditional decision-making systems of various indigenous groups, such as the Mayan Zapatistas or the Haudenosaunee (Iroquois) Confederacy. The degree of romantic essentialization involved is most visible when we hear arguments that all indigenous peoples used consensus process or some other form of radical democracy. These sort of claims seem to be made on the basis of their momentary political appeal, but with little consideration of the evidence.
As anthropology, such narratives may well be bunk. But they have certainly built a mythic ground on which other groups can stand. I know of two collectives, for example, that use a consensus process "inspired by the Zapatistas," though it seems clear that the Zapatistas themselves do not use a consensus process as such.
Finally, consensus has emerged independently as an exigency of certain unusual environments, where individual veto powers are implicit, or where it is difficult to count votes. In any event, by the 21st century the diffusion of consensus process has become so widespread that its historical channels seem to diminish in importance.
Consensus as Practiced by Friends
Quaker consensus is first of all a face-to-face interaction. It is direct democracy not in the sense of a referendum, but in the manner of a town meeting. Indeed, it is hard to imagine that unity would ever be reached through an electoral process that was not based largely on protracted communication. Quaker business meetings are facilitated by a clerk, or occasionally two co-clerks, or a clerk and an assistant clerk. The clerk's role is to express the "sense of the meeting;" which in lay terms means synthesizing a proposal from the group's discussion. There are often other officers appointed, the most common being recording clerks and reading clerks. While these positions are important and highly respected, they do not involve any great degree of delegated authority. Indeed, the function of the clerk is not unique to their role:
"The continuing search for unity is the responsibility of all members, but it is the clerk, often assisted by the recording clerk, who must discern the meeting's united spirit and state it in a form the meeting can affirm….Any member may offer a substitute for the clerk's minute, and the meeting may approve, modify, or reject it, in exactly the same manner as if the minute were submitted by the clerk."
In practice it is quite common for members of the meeting to propose a "sense of the meeting" before the clerk has done so. These are sincere efforts to summarize an opinion that everyone in the group can agree with, and thus they tend to have a nuanced, sensitive wording, not at all like the partisan proposals of majority rule. Sometimes, indeed, they are fairly obscure. "We seek further education and discernment on this matter" becomes Quaker longhand for "motion failed."
In my efforts to find factors for participation in Quaker business meetings, I ultimately discarded the gender of the clerks, the presence of single or multiple clerks, and even the presence of clerks themselves; although the few meetings that had no clerk were all in exceptional circumstances. Given the usual heroic treatment of leaders in political processes, this may seem a bit surprising, but it is fairly consistent with the Quaker treatment of the clerk's position as someone more appointed than anointed.
Business meetings are spent in alternating bouts of silent worship and open conversation. In a larger meeting, written minutes are often proposed, read, and either edited on the fly or sent back to various committees, which usually dooms them until a later meeting. There is no vote taken; anyone wishing to block a proposal does so verbally. This usually happens in one of two indirect ways: either someone will express a further concern; or someone will assert that the proposal does not speak for the whole group. In this latter case, it is often asserted that no proposal on such-and-such a matter could speak for the whole group at that time. Thus, by uniting on the proposal to stop discussing an earlier proposal, Quakers try to avoid forcing any one person to be the naysayer. The negative minutes generated by this process are often quite revealing. Here is an example from NEYM in 1999:
"53. The Yearly Meeting was asked by our Peace & Social Concerns Committee and by Connecticut Valley Quarterly Meeting, to support and endorse the Campaign of Conscience for the Iraqi People. This campaign, which is sponsored jointly by the American Friends Service Committee and the Fellowship of Reconciliation, intends to send water purification equipment to Iraq-both as a humanitarian action and as a public witness against our government's policy of non-military economic sanctions against Iraq….
We believe these sanctions should end, and that they represent an economic attack on the Iraqi people.
We do not have clarity at this time to minute our Yearly Meeting's endorsement of the Campaign of Conscience as our corporate expression of this concern….
It is sobering to remember that some Friends were among those supporting the economic sanctions as an alternative to overt war against Iraq. Now, seeing the tragic effects of these sanctions, we must admit our error and work for reconciliation and peace."
Again, these 160 words might be contracted to say "motion failed," but instead, the minute is drawn out to include an assertion that the underlying concern is indeed important, and a critique of how the group has handled it in the past. It also infers, at least to Quaker readers, that the proposal has the support of the majority of the group, but that the process of negotiation and compromise is not yet over. These are the shades of gray that easily vanish if we think of voting as a simple pass or fail process.
When the group accepts a particular sense of the meeting (i.e. when a motion passes), it sometimes does so against the better wishes of a few people. Friends have also developed a variety of methods for registering the concern of minorities who do not wish to stand in the way of a proposal. Dissidents can "stand aside" from a decision, but have their concerns recorded as part of the proceedings. If more than one or two people do this, however, the proposal will generally be discarded. Actual vetoes, or in Quaker jargon "inability to unite with a proposal," are vanishingly rare. It is considered tactless and pointless to propose a sense of the meeting that obviously contravenes someone's wishes.
This politesse and civility seem somewhat strange to those who are used to majoritarian floor-fighting and partisan politics. But there is no advantage to such tactics in a consensus system. Building a large voting block is no help if it means alienating one or two people. Ridicule and "dirty tricks" backfire. By and large, Quaker Meetings are conducted with an almost exaggerated civility. The complete banishment of majoritarian vocabulary, such as "voting," is clearly a correlate of this. But this is not to say that Quaker Meetings are always placid. I have witnessed shouting matches, vituperation, and clerks who blatantly abuse the slender powers that they have. I have seen a treasurer refuse to process checks for a project he disagreed with. Sheeran relates a session of the Philadelphia Yearly Meeting that was occupied by Black militants demanding reparations money-and continued to meet in their accustomed fashion. And the innumerable schisms that have occurred in Quaker history suggest that things can get much rowdier than anything we have seen in the last sixty years.
Schisms, in fact, seem to be a final safeguard of consensus process. In majoritarian democracies, there is almost always one group in power, which can realize its proposals for better or for worse. The minority might consider secession, but the cost of secession can be very high. In a consensus group, no one can realize their visions without unity. It can thus seem attractive for large majorities to secede, or to encourage their opposition to secede, which amounts to much the same thing. While Quakers view this process with something like abhorrence, it seems to serve a very salutary political purpose.
Default Positions in Quaker Consensus
The preceding discussion should emphasize the fact that participation in a Quaker business meeting confers a relatively great amount of political power. The participant can quite easily delay, obstruct, or significantly modify any proposal they disagree with, even if they are the only person in the group who disagrees. Thus the notion of "voter power" in a majoritarian democracy has little relevance to consensus. A voter in an idealized referendum or a town meeting has a relative power of 1/n, where n is the size of the group. In an idealized consensus system, the voter has a power of 1 where vetoes are concerned, and a power of 0 where positive proposals are concerned.
This difference will come back to haunt us, but it presents an important, if esoteric question. When does a proposal become a veto? Does a group default to its previously agreed-upon decisions, or to some formally defined version of inaction?
Imagine, for example, that a Quaker meeting has agreed to provide free meals for homeless persons in their meeting-house. Later, some members of the meeting decide they are uncomfortable with this. They feel the presence of homeless men is making the meetinghouse an unsafe space for (let's say) a survivor's support group that also meets there. What happens? Is the decision instantly "repealed" because there is no longer unity on it, or is the "motion to repeal" itself an item in need of complete unity to pass?
Non-practitioners of consensus often take the former position, arguing that any interruption of existing unanimity nullifies the decision. Sheeran relates an episode in which students staying at Pendle Hill, a Quaker community, argued that they had they authority to disband that institution. Since Pendle Hill was clearly supported by its long-term residents and trustees, who were probably in the majority, the implicit argument is that without unanimous support, the institution must cease and desist. Similarly, Treloar argues that the spokes-council meetings at the Seattle protest in 1999 were not true to the consensus ideal, because they did not disband when they were interrupted: "In consensus process as it is supposed to be practiced, the [persons interrupting the meeting] would have been considered to be blocking any proposals then on the floor and urging a counter-proposal."
A version of the same opinion is enshrined in the early Quaker writings about self-government. The General Meeting of Skipton in 1659 warned Friends "that no footsteps may be left for those that shall come after," but rather that each generation should re-discover and own their own corporate decisions. But it is safe to say that, in actual Quaker practice, such arguments do not apply on an immediate basis. Once made, a decision is assumed to last until the group unites on a countervailing decision. This notion is rarely spelled out, but it has some very common ramifications. First, many meetings are called to order for a set period of time, and unity is required to extend the meeting beyond that length. Second, Quakers are relatively careful to formally "lay down" defunct committees or projects. Third, Quaker clerks are often zealous about insisting that proposals be delivered before deadline and through the appropriate channels before they can be considered. In short, while the situation described by Treloar may not be ‘consensus process as it is supposed to be practiced,' it is consensus practice as it is has been practiced for several centuries.
So participation in Quaker meetings does not normally grant individuals the ability to personally undo the decisions that the group has already made. Nevertheless, the idea of voter power as the reciprocal of group size is essentially useless in consensus systems. And this poses a very interesting question. Bryan found that the logarithm of group size was, by a vast degree, the primary determinant of participation in town meetings. He attributed this finding largely to the decreasing voter power experienced by individuals in larger and larger groups. In a consensus structure, that argument is inapplicable: in one important respect, a Quaker's "voter power" may even increase in larger groups, since the individual can unilaterally override the wishes of more and more people. What, then, determines participation in consensus?
Consensus in the New England Yearly Meeting
The Structure of the Monthly Meetings
Having sketched the broader outlines of consensus, I now want to begin narrowing to our study area. The statistical aspects of the following study are based on a convenience sample: the data collected by Katherine Clark and published annually in the Minutes of the New England Yearly Meeting of Friends. To understand this data and the surrounding discussion, one needs a little familiarity with how Quaker Meetings are structured. Perhaps more than most subcultures, Quakers have laid a wall of special nomenclature around themselves, making it difficult for the outsider to tell what is happening. Moreover, because of the relative anarchy of the Society, there are some exceptions to almost everything here.
The essential "political unit" of Quakers is the Monthly Meeting. This is a local, weekly worship group that holds a Meeting for Business once a month, with some exceptions noted below. Monthly Meetings are located all across the study area (see figure 1). There are two other kinds of local Quaker meeting: preparative or "allowed" meetings and worship groups. These do not conduct business on their own, and are not listed separately in Clark's statistics. Such meetings are outside the scope of this study, and for the sake of clarity I have eliminated them from the accompanying figures. They are, however, an essential part of the Quaker community.
Monthly Meetings are composed of "members" and "attenders." The distinction between these categories has softened a great deal over the centuries. In the 1700s and 1800s, members were subject to a fairly strict code of conduct, and "disownments" were common, especially when members married non-members. In some cases, conflicts over membership caused entire meetings to split, such as the Timothy Davis schism of 1783. Today disownments have become an anachronism, and roughly 25% of self-identifying Quakers within our sample are now "attenders," and they are by no means a less active group.
Membership itself, while it is invested with cultural importance for Quakers, seems to be a dead letter politically. In many cases, the individuals who take leadership positions in a meeting are attenders, not members. Rarely is it possible to distinguish someone's membership status except by asking directly. In the regressions that follow, member/attender ratios were consistently rejected as a factor in participation. In other words, the distinction no longer even correlates to a measurable difference in commitment to a group's activities. Throughout this study, we will be referring to a Monthly Meeting's size as membership plus attenders.
The Monthly Meetings in our sample have emerged over the space of four centuries. Worship groups appear and disappear, quite informally, whenever "two or three are gathered together," to cite the apostle Matthew. Worship groups fall under the care of existing Monthly Meetings if they feel the need of-or are felt to be in need of-some official recognition. When consensus is reached that a worship group should be conducting its own business-either out of geographical pragmatism or theological dispute-the new meeting is "set off" from its host meeting, and becomes a Monthly Meeting in its own right.
When a Monthly Meeting no longer has the energy to sustain itself, it is "laid down," or disolved. In some cases, a meeting is laid down to another meeting, which means that the business responsibilities (and often the remaining members) devolve on the surviving meeting. Through setting off and laying down, the Monthly Meetings of New England have grown a complex family tree, shown in figure 2.
Since the 1800s, Monthly Meetings have been united into Quarterly meetings, which meet quarterly. These form roughly geographic partitions, although there is some overlap between Quarters (fig. 1). Monthly Meetings are occasionally shunted from one Quarter to another. The Quarters themselves have slightly different group cultures, a point we will return to later. Currently, eight such Quarters form the New England Yearly Meeting (NEYM). NEYM meets once a year, for a very intense week-long session of worship, activities, and business at a rotating location. It should be noted that some Quaker meetings in New England do not belong to NEYM-they either belong to New York Yearly Meeting (NYYM) or they are independent. An example of the latter is the Village Street Monthly Meeting, affiliated primarily with the Trotskyist and/or Fascist candidate Lyndon Larouche (1964-c.1979). Unfortunately, we do not have participation data on these meetings, and they are not included in the study.
Following in this same general pattern, there are several assemblies that represent Quakers at an even broader level. Friends United Meeting (FUM), Friends General Conference (FGC), and Evangelical Friends International (EFI), are considered the "big three" coalitions of Yearly Meetings. Ostensibly even more broad in scope, the Friends World Committee for Consultation (FWCC), meets triennially and has representatives from all or almost all the Yearly Meetings around the world. There are many other super-regional organizations of Friends, but "there is no national or international body of Friends which is superior in authority to the separate yearly meetings." In practice, this autonomy is true of the Quarterly and Monthly Meetings as well-while a deviant meeting might be asked to break its ties to the wider community, there is no enforcement of one meeting's decisions on another, nor is there any mechanism for such enforcement. Over the years, several attempts have been made to establish a concrete polity of Friends superior to the Yearly Meetings, but these efforts have always failed-a point we will return to later.
In this longer historical perspective, it is important to note that the New England Yearly Meeting is in the process of re-uniting from a schism. In 1845, NEYM split into two branches, along the lines of a theological cleavage embodied by Joseph John Gurney on the one hand, and John Wilbur, on the other hand. The minority of roughly 600 Wilburites left, taking with them the name New England Yearly Meeting. The majority group of 8000 Gurneyites (who advanced a more politically engaged, anti-quietist version of Quakerism) took the name "Yearly Meeting of Friends for New England." These two bodies re-united in 1945, exactly a century later, but their constituent Monthly Meetings are still sorting out a range of divisions in theology and politics.
It is a casual commonplace to equate the Wilburites with the political right and the 20th-century Republican Party, and to equate the Gurneyites with the political left and the 20th-century Democratic Party. This is no doubt an oversimplification, and I am not convinced that a one-dimensional political spectrum has much utility, anyway. Yet the underlying point is important: theology recapitulates politics, and the Quaker polity deserves our respect as a very unique political arena.
Participation in Monthly Meetings for Business
The annually published Minutes include a statistical summary of the NEYM's constituent Monthly Meetings. In the 1980s, Clark broadened the questionnaire on which this summary is based to include (among other things) a question about attendance at Meetings for Business. Since the survey already included information about the overall size of the Monthly Meetings, this combined information allows us to calculate annual participation ratios for the Monthly Meetings for Business.
In the period from 1993 to 2003, I have complete annual statistics on 591 Monthly Meetings, out of a total of 677. The average membership of these meetings was 67; the average size was 89, counting attenders. The plurality of these meetings (45%) had a single female clerk.
Before continuing, we should note the biases present in this sample. Despite Clark's persistent efforts, many meetings do not fill out the annual survey. These non-respondents are by no means a random sample of the population. In the first place, they are significantly smaller, on average, than the responding meetings. The average membership of the 86 non-responding meetings is only 33-half that of the responders. A much higher percentage of non-responding meetings are programmed (pastoral)—35% as compared to 13%. Finally, the very fact that they do not respond to the survey infers some disconnection from the larger Quaker community in New England-and there is ample reason to believe that inference is often correct. So everything that follows must carry a certain caveat lector: there is another group of meetings, which are generally small, conservative, and detached from the broader NEYM polity, for which we have no participation data.
Of the 591 meeting-years for which we have complete data, twenty-two did not conduct Meetings for Business at all. The average size of these meetings was 29, and it is worth noting that this group includes the seven smallest meetings in our population. Presumably a Monthly Meeting of four people either has no business to attend to, or else they can do so informally. This represents an important exception to Bryan's basic rule that smaller groups have higher meeting participation rates-a sufficiently small group may not need to have meetings in the first place. The largest of these meetings, however, had 233 people (Hartford in 2002)-quite large by Quaker standards.
Among the 569 Monthly Meetings that actually conducted Business Meetings in a given year, the average participation was 22.4% (Fig. 4). This is only slightly higher than Bryan's 20.5% for the Vermont townships. Participation and size were significantly correlated at r = -0.49. The relationship, however, is clearly curvilinear. This can be approximated by using the logarithm of size, giving an r of -0.76. However, the correlation with the reciprocal of size (1/n), or what Bryan calls "voter power," is even stronger: -0.8. It is also, to my mind, more causally intuitive. It is easy to imagine that people consciously perceive themselves as one among a group of ten, or twenty, or a hundred. It is fairly hard to imagine that people consciously use calculus in thinking about whether or not to go to a meeting.
Thus the reciprocal of size explains 63.7% of variance in participation among those meetings doing business. In Bryan's survey of town meetings, he finds that size explains 60.0% of variance. Clearly this reinforces the thesis that scale is the crucial variable in determining participation in direct democracy. Yet the numbers are almost too similar, given the large differences between Quaker Meetings and Town Meetings. The former are taken from small, voluntary, religious associations, which hold their meetings monthly. The latter are taken from substantially larger, secular, political districts, with coercive authority, meeting just once a year. It is rather startling to find that both groups have a mean participation of just over 20%, and essentially identical correlations between participation and size. The implication of a general power law is hard to resist.
These similarities are even more striking if we consider the differences in decision-making process. As we have discussed, Bryan's explanation of voter power depends upon the arithmetic of one person, one vote. But in a consensus system, this explanation is defunct or even reversed. The person who blocks a proposal in a group of ten people can just as easily do it in a group of twenty people, or a group of two hundred people.
Thus, while Quaker Meetings conform closely to the Bryan's pattern of diminishing participation based on size, they do not confirm the proposed rationale for that pattern. In numerous conversations on this topic, I have asked Quakers and secular activists who use consensus why they think participation drops off with size. The answers are quite consistent. They argue that in a larger group, their opinions are less and less likely to be unique: "someone else will say what I would have said." If we accept this explanation, the relevant variable for consensus groups is not voter power or voter decisiveness as such, but how much voters' presence affects the representativeness of the assembly. It would be interesting to see if this factor is not also present in majoritarian groups.
Pastoralism. Let us now proceed to the other, relatively minor factors in our regression. I began with the fundamental dichotomy in Quaker practice: pastoral or programmed meetings, versus non-pastoral or unprogrammed meetings. This distinction echoes the Wilburite / Gurneyite schism, and the even older Hicksite / Orthodox schism. It is also mirrored in the descent of the meetings, and the arrangement of the Quarters. As we will see, it covaries with many other factors.
After controlling for size, pastoral meetings have slightly lower participation rates than non-pastoral meetings. Monthly Meetings with a pastor have a mean participation at Business Meetings 6.6 percentage points lower than unprogrammed meetings.
Pastoral (or "programmed") meetings have a regular minister who scripts a homily and is the center of religious life in the meeting. Unprogrammed meetings have no pastor, and are often viscerally averse to the idea of having one.
Among early Quakers, the rejection of what George Fox called "hireling ministry" was indubitably a theological mirror of democratic values. Fox argued for a completely decentralized theopneusty: there should be no intermediary between the believer and God; no tier of representation or interpretation; no theological elite prescribing to the laity what to believe and how to worship. Instead, Quakers were to be a "priesthood of all believers." This was, and remains, an extraordinarily anarchic, populist vision of religion. Many Quakers, even in the early years, rejected it to appoint official or de facto ministers. Today, pastoral meetings in many cases appear very similar to the services of more mainstream protestant sects, and are typically more conservative in their secular values, as well.
It comes as no great surprise that these meetings should have lower participation rates. Even without the complex interplay of theology and politics, pastoral meetings have someone very visibly "in charge." The members can feel assured that even if they do not attend meetings, business will be taken care of. At the same time, they may feel that even if they do attend meetings, their voice may be overridden by the superior status of the pastor.
This explanation, however, is not unassailable. There are deep cultural and historical differences between the programmed meetings and the unprogrammed meetings. Perhaps the difference in participation is located somewhere more specific in that sea of divergent variables. Again, we must remember that many of the programmed meetings are non-respondents. It is possible that if that group were included, the effect of pastoralism on participation would be different: greater, lesser, or even reversed. We can't know.
One meeting allows us a very cursory controlled look at this question. In 1996, Westport became unprogrammed. We have a bare ten data points on Westport meeting. Based on this tiny sample, however, we could explain 40% of the variance in size-adjusted participation based on the departure of the pastor, presumably with most other cultural variables held fairly steady. (Gonic became unprogrammed in 1996, but unfortunately we have only three data points on Gonic, so it is uninformative.) Of course, such a sample is not statistically significant, but it is consistent with the larger pattern, which is.
Age. It is painfully evident that groups that are new to consensus process are bad at it. Tactics that are designed to build majority blocs are worse than useless in an environment where majorities do not rule automatically. I had expected that older meetings would have moved past these problems. Thus, I hypothesized that older meetings would have higher participation rates. This was not correct.
Calculating the age of a meeting is not straightforward. For example, Mattapoisett meeting has existed as a religious body since 1702. Yet it was only set off as a Monthly Meeting in 1992, after twenty-nine decades of institutional existence. Still, we are chiefly interested in how long a specific group has been using a specific political process. Mattapoisett has been conducting meetings for business for fourteen years, not three hundred four years. Following this logic, I have calculated each Monthly Meeting's "age" as the time since it was set off, had a major schism, or a major reunification.
After controlling for size and pastoralism, age is negatively correlated to participation. More robustly, the logarithm of age is negatively correlated with participation: it does not make that much difference if your meeting is 200 or 300 years old. However, a five-year-old meeting is likely to have higher participation, ceteris paribus, than a century-old meeting.
I understand this result as highlighting the distinction between capability and participation. I continue to believe that consensus process is a learned skill, and that older bodies are better at it. The higher participation levels of young groups may reflect inexperience and insecurity, rather than "better" democracy. In a group that has experienced consensus for generation, the low levels of participation, while perhaps unfortunate, may simply represent a confidence that the system is working.
Grouping of Meetings. Aside from the black-and-white distinction of pastoral vs. unprogrammed, there are two ways to "group" the Monthly Meetings. The most obvious of these is by membership in a given Quarter. The Quarterly Meetings have somewhat specific cultures. Salem is politically engaged and perhaps wealthier than the others; Vassalboro and Falmouth are too far away to attend to Yearly Meeting, so they focus their energies on the Quarterly level. Rhode Island is full of conservatives; Northwest of hippies.
A more historical method of taxonomizing the Monthly Meetings is to arrange them in families. There are, as we see on Figure 2, six major genealogies of Meetings and six smaller ones. These do not overlap precisely with the Quarters, although in a two cases-Dover and Rhode Island-every meeting in the Quarter shares a common ancestor. Figure 3 shows this overlap more precisely.
I began with a model using dummy variables for membership in the Quarterly Meetings. One of these-Vassalboro-was rejected as a predictor, but this only placed Vassalboro in the neutral case. The other seven were all significant to .050 or better; generally much better. The t-values are as follows:
*Significant to .004 or better
There is some danger of over-interpreting these results, but it is worth describing their general pattern, as otherwise they are meaningless to the non-Quaker. The two Quarters with the highest levels of participation are in Maine. For reasons of brute geography, these Quarterly Meetings are disconnected from the Yearly Meeting as a whole. They are, however, highly active in their own sphere, and have produced such luminaries as Rufus Jones and institutions such as the China Camp. Because of the difficulty of reaching the Yearly Meeting Sessions, the Quarterly meetings in Maine are emphasized more strongly than is usually the case elsewhere. In a sense, then, the hierarchy in Maine is shorter; the outermost circle of the organization is nearer to hand.
Predictably, these two Quarters show markedly higher participation rates than any others-27.9% in Falmouth, and 29% in the even more remote Vassalboro. (The other six Quarters together average 19.6%.) This would seem to lend some indirect support to an anti-Federalist suspicion: participation in macro-government tends to wick away enthusiasm for participation in micro-government. However, the lowest levels of participation are found in the Northwest Quarterly Meeting-Vermont and part of New Hampshire. This region is also quite detached from the Yearly Meeting. Sandwich and Rhode Island Quarters, whose participation rates are almost as low as Northwest's, are probably the most conservative region of the Yearly Meeting, and it seems likely that the ur-variable that we see reducing participation in pastoral meetings is present here even in unprogrammed meetings.
I then moved to looking at the Monthly Meetings by "family" group. I assigned dummy variables to the twelve families, and then discarded five of them in the regression models. In the process, I also discarded pastoralism as a variable. It covaries heavily with family, and so it was not needed as a factor in its own right.
This model proved to be more predictive than the breakdown by Quarter. Blood will out, apparently, in terms of democratic participation. I arrived at the following t-values:
Again, the same basic pattern recurs. The Maine Quakers are heavy participants, the Southern coast are low participants. The only new information here, it seems, is that history is somewhat more predictive than geography. With all said and done, dividing the meeting by Quarters allows us to predict 68.1% of variance; dividing them by families allows us to predict 71.2% of variance.
How large can it get?
For myself, the notion that democratic participation is size-delimited quickly leads to the question how small should groups be to function well? Based on the preceding regression, we should expect that when a Monthly Meeting bifurcates, its daughter meetings should experience a doubled boost in participation: they are smaller, and they are younger. The Quaker author Howard Brinton anticipates those results, and their Kohrian conclusion: "The Quaker method works better in small than in large groups….therefore, if a Monthly Meeting becomes overgrown, it should divide." Such advice emphasizes the personal, face-to-face nature of consensus process. If we want to preserve those qualities, we must stay small.
But for many political scientists, including many radicals, the question posed is how large can groups be and still function at all? In discussions of consensus, as in discussions of direct democracy, this is a quite familiar question. It is perhaps especially pertinent because of the use of consensus by very large groups-the Clamshell Alliance, the spokescouncils at mass convergences, and so forth.
The largest Monthly Meetings in NEYM have a little over 500 members. This is smaller than most townships; however, it is much larger than the personnel of many corporations, non-profits, boards, college faculties, and so forth. Many of these groups also meet monthly, and have a strongly shared sense of purpose at least as regards their organization. In most cases, it seems likely that such groups could practice consensus, if they wished to. Indeed, the minutes of most boards reflect that virtually all decisions are made unanimously, although the opportunity for majoritarian voting exists.
In addition to these primary assemblies, Quakers also have the Quarterly and Yearly Meetings, which function in much the same way. Townships, notably, do not have such secondary and tertiary structures. The Vermont state government, which is younger than many of the townships, operates through a constitutional representative democracy, not through direct participation.
I have no data on participation in the Quarterly Meetings. New England Yearly Meeting as a whole is composed of about 5500 members and attenders. At their annual business sessions, the typical turnout is between 60 and 200: a participation rate of 1% to 3.5%. In 2005, I attended a business meeting with 405 members present, a participation rate of 7% for the entire New England region. Consensus is frequently reached at this level, even on controversial issues such as tax resistance. But now we are at the very ceiling of workable scales.
Periodically, Quakers have made efforts to establish "global" assemblies that would, among other things, make decisions in the name of a much wider group of Friends. These have all essentially been failures. Most recently, the Oskaloosa conference in 1929 failed over disputes about "unbelievers" (e.g. liberal Quakers), and the 1977 Wichita Gathering of Friends of the Americas failed over the question of homosexuality. Thus it appears that, for Quakers, consensus is quite feasible in a polity of 5000 or 6000 people, and quite impossible at the level of 20,000 or 30,000 people. In a partial recognition of this, there are two strategies that Quakers have developed in order to conduct super-regional organizing. I think both are of some interest.
The first is the establishment of relatively powerless consultative assemblies like FWCC. The FWCC includes Quakers from all over the theological, political, and geographical spectrum. As a political decision-making body, it is completely impotent-clearly unable to agree on anything controversial, it does not even bother to try. As a forum for communication and fellowship, however, the FWCC is highly successful. Concerns brought up at the FWCC triennials often descend into the business life of smaller Quaker bodies.
The second pattern is the establishment of small, autonomous organizations that serve an umbrella function without committing themselves to the decisions of a wide Quaker polity. A few examples will suffice. The Friends' Committee for National Legislation (FCNL), and the various chapters of the American Friends' Service Committees (AFSC) are each structured as independent non-profits. So are more specific groups like Quaker EarthCare Witness (QEW). These groups are not part of any Yearly Meeting; they have their own membership structure and committee structure, which may include non-Quakers as well as Quakers. In some cases, this internal structure is quite elaborate, with multiple regional chapters and committees.
Although these organizations are independent, they do not operate in a vacuum. Quite the contrary. These groups constantly poll, survey, consult with, educate, and draw members from the larger Quaker community. FCNL's lobbying priorities are based entirely on their surveys of Quaker opinion-although FCNL is an autonomous entity and cannot be obliged to lobby in a particular way.
This separation of foci seems to have many adaptive features. In the first place, it can serve to remove pressure from the Monthly Meetings to reach consensus on every detail of their worldviews, a process that would be divisive and likely impossible. Instead, Monthly Meetings can enter into dialogue with groups like the AFSC or QEW, shaping opinions and offering support of various kinds, without having to commit themselves unreservedly to support for the outside groups' decisions. This development would seem to answer Wolff's concern that a consensus-based organization will dissolve the first time it reaches an intractable dispute.
At the same time, the specific focus of these secondary institutions creates a community of interest in which it is relatively easy to unite on decisions. AFSC members have sometimes reported that it is much easier for AFSC to find unity on political issues than it is for the Monthly Meetings. This is true despite the fact that much of AFSC's membership is non-Quaker, and thus—from the point of view of many Quakers—ought not to be able to reach consensus at all.
This paper has attempted to address to some of the quantitative issues of consensus in Quaker practice. I will present a more qualitative view in a subsequent article, due to size constraints.
Size is confirmed as the most important factor in determining participation in radical democracy. The similarity of results between Bryan's work and the preceding regression suggests the possibility of power laws for participation in radical democracy that are independent of decision-making style. Institutional age appears to be negatively correlated to participation, and institutional origins appears to be more important than geography. The presence of pastoral leadership, which co-varies with history and geography, seems to be associated with lower participation. I suspect that an important variable for future research will be the representativeness of participants.
Appendix: Final Regression
Andersen, Alfred (1997) Fellowship Roots: Where We've Been, Where We Might Go, Fellowship News (http://communities.ic.org/97/felln97.php), accessed July 2006.
Bookchin, M. (2001?) What is Communalism? The Democratic Dimensions of Anarchism. Anarchy Archives (http://dwardmac.pitzer.edu/anarchist_archives/bakunin) accessed August 2006.
Brinton, H. (1952) Reaching Decisions. Wallingford, PA, USA: Pendle Hill.
Bryan, F. (2004) Real Democracy. Chicago, IL, USA: University of Chicago Press.
Coover, V. & Deacon, E. & Esser, C. & Moore, C. (1977) Resource Manual for a Living Revolution. Philadelphia, PA, USA: New Society Press.
Fager, Chuck. (1996) Without Apology. Kimo Press, Media, Pennsylvania.
Hasbrouck, Jay (2005) Anarcho-Primitivism and the Appropriation of Anthropological Narratives. Paper presented to the Association of American Geographers. Denver.
Morse, D. (2001) Testimony: John Woolman on Today's Global Economy. Wallingford, PA, USA: Pendle Hill. Pamphlet #356.
New England Yearly Meeting (1986), Faith And Practice of the New England Yearly Meeting of Friends, 1985. Worcester, MA, USA: New England Yearly Meeting of Friends.
Nordhoff, C. (1960), The communistic societies of the United States; from personal visit and observation: including detailed accounts of the Economists, Zoarites, Shakers, the Amana, Oneida, Bethel, Aurora, Icarian, and other existing societies; their religious creeds, social practices, numbers, industries, and present condition. New York: Hillary House.
Norlind, E. F., (1969) The Atonement of George Fox. Wallingford, PA, USA: Pendle Hill. Pamphlet #166.
Pollard, F. E..; Pollard, B. E.; Pollard; R. S.W. (1949) Democracy and the Quaker Method. London, UK: Bannisdale.
Rothstein, A. & Butler, C.T.L. (2005) Conflict and Consensus. http://consensus.net, accessed Aug 21, 2005.
Sheeran, M. J. (1983) Beyond Majority Rule. Philadelphia Yearly Meeting, West Chester, Pennsylvania.
Stattler, Richard. (1997) Guide to the Records of the Religious Society of Friends (Quakers) in New England. Providence, RI, USA: Rhode Island Historical Society.
Treloar, M. The Tyranny of Consensus, Clamor, May/June 2003, pp. 38-40.
Wolff, R. P. (1980) In Defense of Anarchism. New York: Harper and Row.
 Wolff (1980) pp. 22-27 makes this case persuasively, though, as we will see, he is unfamiliar with such systems in practice.
 The Mayflower Compact, which was signed by 41 of the 102 pilgrims, is the nearest thing to an exception that I am aware of. Here again, the scale is smaller than most villages. At the Constitution museum in Philadelphia, visitors can opt to sign on to the U.S. Constitution retroactively-so far about 100,000 people have, many of them non-US-citizens.
 This is quite a strict taboo. Quakers eschew all reference to "democracy," "voting," "vetoes," and sometimes even "approval" within their own meetings.
 It is worth noting that Quakers themselves hold a very wide range of beliefs on this question, as they do on almost all spiritual questions. For instance, the Pollards (1949) end their fairly unique book on consensus by attempting to link it to mental telepathy.
 Today we would call this civil disobedience, though the term originated with Thoreau. The practice, of course, is much older than either Thoreau or Penn. It forms the theme, for instance, of Plato's Apology and Sophocles' Antigone.
 In England, the decision-making threshold was reduced to 10 out of 12 in the 20th century, due to concerns about organized crime intimidating jurors. In Scotland, the threshhold is 8 out of 15, and apparently always has been.
 This era marks perhaps the fourth or fifth "wave" of intentional communities in American society. The term "intentional community," however, dates from the 1970s*.
 Andersen (1997)
 In an upcoming paper, I am presenting a quantitative analysis of this trend.
 Coover et al. (1977)
 Butler and Rothstein (2005)
 Nordhoff (1875)
 Hasbrouck (2005)
 I have worked or attended meetings in about fifteen Zapatista villages, across a range of the tribes that are involved in the movement. While I do not speak any of the Mayan languages, I feel I can give a reasonably accurate firsthand account of their meetings. There is a wide variety of practice from village to village, and most decisions are made informally, which is not to say rapidly, often over the vocal protest of minorities. A complex system of checks and balances exists, through the use of a parallel hierarchies, so that one's subordinate on one issue is one's superior on another issue. On some occasions, referenda of the village or the entire region are resorted to. I have been personally impressed at the Zapatista's ability to make inclusive and democratic decisions even under great duress, but they do not use consensus.
 For example, on open internet forums such as Wikipedia, an individual may be able to create multiple online personas known as "sock puppets." This reduces the credibility of majority voting.
 New England Yearly Meeting (1986), p. 222
 2000 Minutes, p. 37. One point which did not make it into this minute, but was widely heard at the meeting, was the notion that merely signing onto someone else's campaign was too easy.
 Anarchists and other non-Quaker activists usually use the phrase "block" in place of "veto." Quakers generally find both phrases offensive, and rely on long circumlocutions.
 Sheeran (1983). Some years ago, in the Friends Historical Library at Swarthmore, I dug up the minutes of this meeting. Like most Quaker minutes, they are written in an idiosyncratic elusive style. The only reference to the fact that the building had been taken over by armed men was a single line, something like: "Clerkship of the meeting was transferred to so-and-so." After several hours the militants transferred clerkship back to the Quaker clerk, but whether this was done from exasperation or respect history does not record.
 Sheeran (1989), p. 59. The matter is somewhat complicated by the context of other student power movements.
 Sometimes it is explicit, especially when there is money involved. Hence a clause saying that certain decisions "are to be considered as ongoing financial policies of the Yearly Meeting not requiring annual reconfirmation." (Minutes 2000, p.37)
 Such an anachronism, in fact, that when President Nixon's home meeting was nearly unified on disowning him for waging war, a Friend reportedly said "I had hoped Quakers had gotten beyond disowning people." Nixon died a Quaker.
 New England Yearly Meeting (1986) p. 218
 The geometry of this schism can be seen fairly clearly in Figure 2. Previously, in 1825, there was a schism between the Hicksite and Orthodox meetings. A lesser schism, known as the "Timothy Davis Seperation," can be seen in 1783 in the Pembroke / Sandwich genealogy.
 New England Yearly Meeting (1986), pp. 38-45
 It is possible that Hatford's reported absence of meetings in 2002 is an error, but I have not yet been able to verify that one way or another. It is also possible that it was a "jubilee year," an intentional abstention from business in order to focus on spiritual renewal.
 Significant to .000. Every factor mentioned is significant to .05 or better.
 Significant to .016. The t-value is -2.4, compared to 28.7 for the reciprocal of size.
 Regressed against the reciprocal of size and pastoral status.
 Regressed against reciprocal of size and log of age. All factors significant to .000, except for Greenwich, significant to .002.
 Brinton, pp. 19-20
 Fager (1996), pp. 1-9; I am also grateful to my late grandmother-outlaw, Elizabeth Watson, who played the role of deal-breaker at Wichita, for her lively recollections of that era.
 The constellation of secondary "Friendly" organizations do not always agree with each other. As Quaker meetings divide their support between, say, AFSC, QEW, and Friends General Conference, they are supporting very different and sometimes conflicting strategies of social change.
 Sheeran (1983) p. 60.
Information about this Article
Published on Tuesday 29th August, 2006 at 02:11:51.
Peer review added 16th March, 2012 at 02:23:16
Good work.. nice collection of references… work seems to be original and the idea of search on this topic deserves congratulations to authors…
Additional peer comment added 16th March, 2012 at 02:23:35
Good work.. nice collection of references… work seems to be original and the idea of search on this topic deserves congratulations to authors…
Website copyright © 2006-07 Philica; authors retain the rights to their work under this Creative Commons License and reviews are copyleft under the GNU free documentation license.
Using this site indicates acceptance of our Terms and Conditions.
This page was generated in 0.0540 seconds.