Some information flows in the Samoan fono

by Jacob Lee
Idealized spatial organization of the fono

Idealized spatial organization of the fono

The fono, politics and space

A fono is a political meeting in the Western Somoan village. The fono is a spatially organized event: high ranking orators are seated in the front, low ranking orators and other low status and low rank persons in the back. High status chiefs and special guests are seated on the sides (tala).

A person’s location with respect to these three areas during an event may signal a variety of informational contents, some of them stereotypical and some of it sensitive to the particular situational context of the event. Within each area, the position of individuals can also signal various informational contents, particularly as relating to status. The precise boundaries between these three areas may not be well defined, and seating in an ambiguously defined location may itself convey interesting information.

What enables these sorts of information to be conveyed?

Cognitive Schema of the fono

Alessandro Duranti answers this by positing a cultural cognitive schema defining an idealized fono seating arrangement spatially organized in terms of various socially relevant classes of persons:

By matching the ideal plan for a particular occasion with the actual titleholders who occupied various positions in the house, one could obtain a first reading of the political situation and make a few predictions about the way in which the discussion might unfold. Thus, according to the kind of fono that was being held, a particular set of orators would be expected to sit in the front row. In such a system, every slight variation from what is considered the ideal plan is potentially significant. For this reason, as suggested above, the ideal plan acts as a cognitive schema that provides a key for the participants to interpret the contingencies of the day. The relationship between the ideal seating arrangement and the actual one gives a first approximation of the potential conflicts, tensions, and issues of the day. (Duranti p. 65)

This schema indicates various sorts of stereotypical information. If a person p is seated in the front F then one conventional item of informational content indicated by this fact is that p is a high-ranking orator.  Other informational contents are indicated too, of course. More interesting are occasions when an actual fono deviates (or is believed to deviate) from this schema. Subtle variations in seating arrangement, in conjunction with knowledge of the political situation, can indicate various interesting sorts of information. Throughout Duranti’s analysis, he draws us to the subtle political dimensions of the relative positions taken up in the political theater of the fono:

Samoans are in this respect true masters of spatial finesse, as demonstrated by the poisition occupied by the matai (JL: chief) who shares the title with Savea Sione, namely Savea Savelio. He sits in a position that is similar to Savea Sione’s but slightly “farther back.” This he explained to me as a sign of restraint: He should not take a foregrounded role in the fono proceedings given that the actions of the one we might call his alter ego, Savea Sione, were under severe scrutiny by members of the assembly. (Duranti p. 68)

and how an intimate understanding of this space provides informational cues for contextually relevant political positioning and interaction.

An understanding of the locally engendered meaning of the seating arrangement for the day suggests that Moe’ono, as well as the other matai in the fono house, had ways of expecting, ahead of time, Tafili’s attack and her role at the meeting. If she is present and has chosen to sit in the front row, the place reserved for the more active members of the assembly, everyone knows that Tafili has come ready to speak and, most likely, to argue. Thus, even before a word is exchanged, Tafili’s spatial claim provided Moe’ono with clues about the forthcoming discussion and gave him some time to prepare himself for it. In this case, the regionalization of the interactional space available to participants can communicate just as much as words. (Duranti p. 72)

Information Flows

Jon Barwise and Jeremy Seligman argue that information flow crucially depends on regularities within distributed systems. Such information flows are present, but they will not necessarily be available to any particular cognitive agent nearby. Such agents must be attuned to those regularities to translate information about the occurrence of an event of one type into information about the occurrence of an event (possibly the same) of another type. This becomes particularly evident when one is displaced into natural and cultural environments outside our own experience and knowledge. One might not know, for example, that an increase in the number of insects indicates water nearby; a certain discoloration of the skin may indicate that a patient has a certain inflammatory skin disease, but only to a person attuned to the constraints between this kind of skin discoloration and the presence of that particular disease. Duranti describes his first encounter with the fono:

The first time I entered the fono house, I only saw people sitting around the eges of the house and noticed that some portions were unoccupied whereas other portions seemed crammed with people. (Duranti p. 64)

It was only after mapping many different fono events, and matching seating with titles and other relevant information, that Duranti was able to appreciate how much information about the political  events of the day was present in the seating locations of its participants. Standing back from this, we must recognize that every participant in a fono is thus situated, having their own information, and being attuned to some constraints active in the situation, and not others, and so on. This is particularly easily seen when on one occasion, Duranti intentionally seated himself  in a low status position in the fono, when as a guest he was usually accorded a high status position. In this case Duranti sat in the back of the fono (low status position) with low status men, and women. He was was served food last by young servers, and didn’t get any fish, until one of the high status chiefs noticed and directed the servers to bring some of his fish to Duranti:

My experiment was over. I had been able to show the relevance of the locally defined spatial distinctions (front vs. back region) for establishing the status of a participant in a public event. At the same time, I had proven to myself that the system was flexible. Different “parts,” namely the servers versus the matai, or the kids versus my adult friends, were acting on different premises. For the kids who brought in the trays with the food and for the untitled adults who were preparing the portions, it was safer to follow the basic spatial distinctions. They had no way of knowing the details about who was doing what on a particular occasion. The spatial arrangement in the house constituted a first key to know how to operate with a minimum assurance of appropriateness…In most cases, the seating plan works very efficiently to convey a first sense of order. Whether or not that order conforms to the relative statuses of the participants as displayed on other occasions is not something that low-status people must be concerned with. Their socialization teaches them that any hierarchy must adapt to contingencies, must fit their task…It is up to the more knowledgeable members in the gathering to complement or rectify the reading provided by the bare layout of the human bodies in space…The distribution of knowledge about how to act on any given situation is thus functional to the distribution of power within the community. On the one hand, the lower-status people act on more general and hence more easily amendable models, that is, models that need additional information in order to operate appropriately. Higher-status people, on the other hand, not only have access to more specific information about the nature of the activity and the expected and expectable actions, they also control this more specific knowledge by putting it to use when they choose to do so. (Duranti p.59-60)

The fono is a distributed system, with many different parts. Information about one part of the fono can give us information about other parts of the fono. But the information flows in this distributed system are relative to to the cognitive schemas by which the fono is conventionally understood by each of the participants. The information flows to which any participant is attuned depends on their assessments of how others are attuned to informational flows in the event. Duranti is able to assess the reasons for his not receiving any fish from the servers because his position in the fono conventionally indicated a low rank and status, and because the servers were not sensitive to other information relevant to interpreting Duranti’s behavior in any other way*.

In order to see Duranti’s choice of seating as an exception to the idealized arrangement presumes not only knowledge about the idealized arrangement, but crucially requires additional information that is inconsistent with that idealized arrangement. In this case, the fact that at one other such occasions Duranti had been seated toward the front. Like Duranti when he first began to participate in the fono, any stranger encountering the scene for the first time, especially one who was not familiar with the dynamics of agency and signification in the fono, would not know ‘what was going on’. That Duranti was unusually seated likely would not have impressed itself on such an observer as a fact worth pondering further. But the intelligent and culturally and situationally literate observer would have seen Duranti’s position as unusual, and possibly interesting, if noticed. The seating of any actual fono is set against the idealized arrangement given by the cognitive schema; the difference between them is a scaffold for signification.

Since deviation from the ideal is not infrequent, and is often interpreted this way or that, we might very well suppose that some deviations from the ideal are in fact conventional and well understood. Others may be less well-understood. When Savea Savelio sits slightly back from Savea Sione, Savea Savelio, and presumably many of the other participants understood, though Duranti may not have seen the reason until later. Yet,  while there may have been some conventional interpretation of Duranti’s sitting in the back, it is safe to say that that interpretation of his seating misunderstood Duranti’s unconventional objective. We may well doubt that anyone besides Duranti, or anyone he let in on it beforehand, correctly understood that his sitting in the back was  a behavioral experiment.

*It is also possible that the servers speculated on the reasons, but did presume to act on those speculations.


A. Duranti, From Grammar to Politics: Linguistic Anthropology in a Western Samoan Village, Berkeley: University of California Press, 1994.

12 Responses to “Some information flows in the Samoan fono”

  1. Cool. But Duranti’s analysis appears to be predicated on three considerations: (1) a shared cognitive model — assumed to be equally shared by everyone; (2) status differences of which everyone is also aware; and (3) political subtleties of which only some, presumably the higher status and more experienced participants are aware. I wonder if the model could be improved by considering (4) the effects of spatial positioning on lines of sight. I am influenced here by having read somewhere (I will try to chase it down) about a classic modeling problem: waves of applause in a concert or theater setting, where those sitting in the front rows cannot see what is behind them but may have a strong effect on those at the back who can see what those in front do.

  2. Hey John, I thought this rang a bell so I checked – you blogged this almost exactly a year ago (how time flies) as ‘the standing ovation problem‘. The book is Complex Adaptive Systems: An Introduction to Computational Models of Social Life, John H. Miller and Scott E. Page (2007). Asher and Noen made a fun connection in the comments to the ‘boid algorithm’. This is clearly one of our blogly themes! But I’ll need more headspace than I’ve currently got available to engage this as it deserves….

  3. Again just shooting from the hip on the run, I’m reminded of the distribution of political factions in the seating chart of the National Assembly during the French Revolution. Wikipedia offers this interesting observation of the process involved: “One deputy, the Baron de Gauville explained, ‘We began to recognize each other: those who were loyal to religion and the king took up positions to the right of the chair so as to avoid the shouts, oaths, and indecencies that enjoyed free rein in the opposing camp’.” The stakes were high here, eventually life and death, and the nuances of positioning were accordingly saturated with meaning.

  4. (1) a shared cognitive model — assumed to be equally shared by everyone; (2) status differences of which everyone is also aware; and (3) political subtleties of which only some, presumably the higher status and more experienced participants are aware.

    I think that this captures the base model. But I think that (1)-(3)can be weakened to account for diversity of (overlapping) models and understandings.

    You are right that the information flows that a participant has access to depends upon how they are situated within the event. Audience members in an auditorium facing forward in the back have a very different perspective than those seated in the front. There are also of course other pragmatic issues involved in choice of seat, and who is not to say that on this or that occasion, a person’s choice of seat was unconcerned with appearances, or was dictated rather by some mundane and non-political circumstance. And surely, very general issues of proxemics pertain to any such event.

    I do not know if Alessandro Duranti was especially interested in information flows per se when he wrote what he did, though he was certainly interested in how participants learned and used the information available to them in the fono event. Or let me say that there are a few ways in which the phrase ‘information flow’ is understood.

    On the one hand, we can think of information about something moving in space and time from one location to another, such as when an item of information moves from one person to another. On the other hand, an information flow is understood as the movement of information within an abstract logical space. Jon Barwise once wrote that information moves at the speed of logic. I would say that it moves at the speed of inference, or computation. The point is that information flow is not like the flow of information vehicles. After all, an event can carry information about another event that has not yet occurred. These two sorts of flow are distinct, but interdependent.

  5. But I think that (1)-(3)can be weakened to account for diversity of (overlapping) models and understandings.

    Easy to say; but that takes the modeling problem to a new level of complexity. One of the most provocative remarks in Page and Miller’s Complex Adaptive Systems is that the agents in most models are either too stupid (simply following prescribed heuristics) or too sophisticated (using game theory to calculate every move). How to model agents that have both different perspectives (presumably dependent on such things as spatial and status relationships) and different levels of interpretive/decision-making capacity (could this be treated as an outcome of experience, with agents learning over time?) seems to be a hellaciously difficult problem. Crack that one and you could get famous very fast.

  6. So much depends on our objectives. Is our objective an agent-based simulation of the fono, or is our goal to understand the fono as most of its participants understand it, well enough that the interactive behavior of the participants makes sense, or is it to learn how to navigate the fono, at least well enough that one could step into an orator’s shoes comfortably, … or is our objective something else? I think the answer to this question will determine what sorts of complexity we can tolerate, or must tolerate.

  7. Great point. From a Predator drone’s perspective the situation is quite simple: all the targets are clustered close together. From a feminist’s perspective the situation may again be simple: all the empowered people are men. And from a virus’ perspective the situation may be wicked complex, depending on the immunological variance of the local population, nutrition rates, vectoring opportunities and so on.

  8. My, God. Complete agreement. Where do we take this next?

  9. I hope I am excused, at least for the time being, for not having as intimate a familiarity with the ethnographic and anthropological literature as would justify the next few things I say. As I said before, so much depends upon one’s objectives and I think that the breadth of anthropological inquiry, at least to the extent I know it, reflects how wide a net the discipline has cast. Nonetheless,in listing various ways in which one may approach a social and cultural phenomenon like the fono, it will be noticed that many of them, although more or less of respectable anthropological heritage, are *not* about understanding the dynamics that move a social-cultural system from one state to the next; in short, their objective is not a theory of history, at any scale. That is not to say that no insights into such questions is possible to be had from such inquiry; that is obviously untrue. But neither is it sufficient, or at least so it seems to me.

    As you may know, as an undergraduate at UCLA my emphasis of study was archeology. I learned that while archeologists themselves have a wide spectrum of research objectives, at least one very prominent thread within archeology is the study of the growth, transformation, and sometimes collapse of societies, a natural consequence of archeologists inevitably confronting evidence of change through time. This concern with the evolution of society I took away with me, even as I moved away from archeology itself, and even as the scale of social phenomena of interest has sometimes shrunk from the rise and fall of civilizations to the causal chains making up the events of everyday life.

    Here I must digress and confess that I am not yet quite clear on the scope of my own ambitions and objectives; or rather confess to having eagerly fallen prey to ever digging deeper into the micro at the expense of the macro. The futility of this is obvious, but more than amply demonstrated by the immense computing power brought to bear by Blue Brain Project in simulating a single cortical column (there are about a million in the human brain). We live in an incredibly fine-grained universe.

    But to return to my main point before I trail off any further, it is in view of this concern with essentially historical processes that I wish my post to be read, though I do not cast my discussion in such a way. One thing that distinguishes say a solar system from a biological systems is the degree to which information processing is intrinsic to its behavior and long-term evolution. This is acutely more true of human systems. But how does information figure into human life, and how can theories of information inform our understanding of historical/causal social processes? To what extent can we understand the course of events at a fono, such as some of those described by A. Duranti, as the consequence of situated reactions to information flows? I do recognize however that there is some reason to be wary; after all, one can quite easily deceive oneself that one is doing something new and exciting merely by dressing up old ideas in new clothing.


Leave a Reply!

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: