Doubting among the Zafimaniry

In 2003, I organized a series of group discussions on psychological matters – thought, language, memory, dreams, ancestors, etc. – among Zafimaniry villagers, a group of forest dwellers in Madagascar who, for historical reasons, are fairly distinct and relatively isolated from other Malagasy. In spite of the presence of a church school, the villagers can be considered as either unschooled or as very little exposed to schools, since the actual school is now hardly ever in proper functioning order. In spite of this, I found that on the whole, my Zafimaniry interlocutors were exceptionally coherent and self-assured in the views they expressed. In this post, though, I would like to describe two different types of shared public manifestations of doubt that arose in the course of these common intellectual journeys.

Type 1 doubt – Doubt as a step on a progression towards truth

Once, we had been discussing the relation of language to thought in people, and I had discovered, somewhat to my surprise, that the Zafimaniry villagers shared a clear theoretical position concerning the question. The more vocal among them told me that thought and language were very different matters and that language was not necessary for thought. It seemed from the general approval with which some statements were greeted that everybody agreed about this and there was no air of doubt that I could detect among the small crowd that had gathered in the house where the experiment had taken place.



The only discordant voice came from me as I expressed a certain amount of disingenuous surprise bordering on scepticism. This attitude caused my interlocutors to want to convince me of their point of view, and for this they used the example of the deaf and dumb man who lives in the village and who (so argued my interlocutors) was clearly capable of thought, while at the same time being deprived of language. Indeed, people had often expressed great admiration about this man’s ability to communicate. I was genuinely impressed by this example and the rhetorical use that was being made of it but I, nonetheless, continued the discussion further by raising a further question. I asked if animals, who also obviously could not speak, were capable of thought.



Again, the answer was loud and clear: Yes they were! In order to convince me of this my interlocutors used another example. Pigs, they said, when they see someone come out of a house with a basket full of taro, come rushing to the spot because they think that there will soon be peelings that they will then be able to eat these as they fall to the ground.

At this stage I then encouraged the joint intellectual journey to move yet further by asking if this was the case for other animals such as chickens. The answer I was given was the same as that for pigs and similar examples were also used to convince me that all animals could think. By then, however, the whole dialogue was becoming something of a joke as the question was thought was being considered for ever more lowly animals. In the boisterous linguistic melée that was developing, some of my interlocutors adopted my part in the dialogue, not by taking my side of the argument, but by forecasting my next question only to answer it as soon as they had formulated it. “What about fleas?” they asked and their answer was “clearly fleas thought since they hid in the seams of garments so as not to get caught”. This thought experiment delighted everybody and so I played my last card.

“Do trees think?” I asked. The assembly went thoughtfully quieter, partly because trees are of central importance for the Zafimaniry and therefore not a joke. Most people said that trees could not think but they did not argue the point. However, the participant who had most clearly enjoyed the intellectual game, an old friend of mine, ingeniously proposed the opposite of what seemed the general opinion and he argued his point. “Yes” he said “trees think and this is shown by the fact that their roots, when finding themselves on rock, seek out the wet soil by growing towards it.

Did he believe that trees thought? I doubt it. What I understood of his meaning, with the multiple cues that speakers make visible in the course of a conversation, was: “this is an intriguing possibility and the line of argument we have been following could well lead to that conclusion, but then we have now reached a pretty ridiculous level of talk and I am not really committed to what I am saying”. We had entered doubt and I believe this mood of doubting was shared. However, this conclusion of mine is only based on interpreted micro clues, too impermanent to take in consciously and analyse, far less to describe them. Such clues are the only evidence possible for the existence of this type of doubt in such context. This is probably why the ethnography of doubt is so rare in the published record.

The ephemeral, inexplicit character of the phenomenon does not, however, mean that it is not worth while reflecting on what is involved. The best way to do this is to consider that this sort of doubt exists within a dialogic movement from certainty towards the unknown. This is what was happening in the discussion outlined above.

During the early part of the interchange, the villagers seemed, as a group, certain of what was being asserted. Their certainty was based on a combination of two factors. Firstly, it was based on their trust in the testimony of other people who had assured them in the past that things were so. Because of this and because of their trust in the opinion of others who were present in the room, people who either did not know this before or, more probably had never thought about the matter, were certain that pigs are capable of thought. Secondly, those who had been clear about the matter used empirical evidence, either evidence from their own observation or the reports and interpretations of others, to support the received wisdom and thus made its propositions more convincing. This use of evidence from the senses, above all sight, is characteristic of Malagasy reasoning as it is probably of reasoning in most places in the world. In this case, the empirical support provided by the reference to pigs expecting their food clearly did its rhetorical job successfully. The empirical evidence was not however just left to speak for itself; instead it became a tool in a process of induction and deduction that led ultimately to more dangerous and uncertain ground. Ground that was more remote from the certain truth of the testimony that had been relied on and of the truth guarantee that had been provided by relevant empirical evidence. This distance from certainty had moved the discussion to the territory of doubt. Whether trees thought was not something people had said in the past and this was being asserted only by one person in a somewhat tentative tone of voice.

What is essential for understanding the occurrence of this type of doubt is to place it in the developmental movement of shared thought. This is central for realising where this doubt has come from but also for where it might be going. The villagers clearly enjoyed the perilous nature of the assertions they or others were proposing as they moved ever more forward in their joint reasoning into less secure territory. Their attitude, however, also made it quite clear that, at the same time that as they were moving in this direction they were quite prepared, expecting almost, for their doubt to be dispelled in future, either because of new testimony from others, whether present or not or through the use of new empirical evidence. This type of doubt was thus seen as but a necessary moment in the exiting and, in this case at least, jolly progression towards firmer truth. This moment of doubt was thus not all that dissimilar to the use of doubt as a key scientific tool advocated by a whole line of philosophers, Descartes perhaps being the most well known among them.

Type 2 doubt – Doubt as a conversation-stopper

From the type of discussion I have described above, it had emerged that, for the Zafimaniry, and probably for most other Malagasy, thought is very much a matter of aligning intention and action. They have a very pragmatic and down to earth approach to mind. I was therefore intrigued how this could connect to other aspects of their culture, especially their relations to dead ancestors which had always seemed to me contradictory. Would I be able to move smoothly our discussions about questions of mind, which had been of a scientific character to these matters which many anthropologists characterise as religious or mystical?

I doubted it, because the Zafimaniry often declare propositions, and act in ways, which the anthropological literature would normally label as ancestor worship. People make offerings to the dead in order to obtain their blessing. They fear the ancestor’s displeasure. When it is believed that this displeasure has been incurred the dead are believed to manifest themselves to the living by asking for certain rituals to be performed for them. These demands are most usually made in dreams.

At first, there seemed to be few difficulties.

I repeatedly tried to move the discussions towards such matters by asking about the mind in sleep. In one such attempt, the discussion about the relation of sleep and mind started with my being told with great certainty and unanimity that in sleep one was “as if dead”. After death all activity stopped and one was certainly not able to think. Similarly in sleep the mind switched off and that was it. It seemed as if nobody present had any doubt about these matters. Although I had heard this sort of thing before, it was nonetheless surprising both in relation to death and sleep, since such statements seemed inconsistent with the beliefs and practises just alluded to. Rita Astuti, working in another part of Madagascar has examined with great precision a similar apparently contradictory situation.

Because of the prima facie contradiction between the propositions just expressed and my knowledge of Zafimaniry ideas concerning ancestors, I then asked about dreams and how, for example, it was possible for ancestors to come in dreams to ask to have a ritual performed for them if one was then “as if dead”.

The mood changed in a way that I cannot fully document empirically, but I have little uncertainty about this. In any case the rather boisterous atmosphere which had characterised the discussion up to then disappeared and people became much quieter. I believe that the contradiction that lay behind my questions was sufficiently salient so that a good number of people understood what I was driving at and this is what explained the change of mood. Several people had already said at this point, that, when talking about dreams, we were moving to an area that was “difficult” ("sarotra"). However; I also got an intriguing answer from two or three senior people. The most vocal of these put the matter in this way: “Although people who are asleep are like dead, the spirit of other people could come and be active in the sleeper's head”. This answer does indeed remove the apparent contradiction which I had been exploring.

I could not tell with what degree of certainty this proposition was expressed, but particularly interesting was the reaction of several other people. They approved of what was being said with one of the several Malagasy ways of saying the equivalent of the English 'Ha Ha', a vocalisation which is also accompanied by several semantically significant movements of the head. In this case what, I assumed, was meant was something like “thank you, this matter is now clear now that you have explained it”. The explicit meaning of the nodding is, therefore, that doubt has been removed, but it also implied a modification in the relation of discourse to fact which I interpret as meaning that, by now, in the dialogic process, we have reached such a very difficult area that neither evidence from the senses nor induction or deduction are safe and that, as a consequence it is best to turn for information to authority figures who can be trusted.

What made this clear was the contrast between, on the one hand, this stage in the discussion and, on the other, what had gone before and, on the other, with the other discussion described above. In these all kinds of people joined in: women and men, old and young (with the possible exception of young newly in-married women). Indeed in these more matter-of-fact exchanges older women dominated. By the time we had moved to the discussion about dreams, however, only senior male authority figures spoke and, as we have seen, the other people present expressed their deferral to their opinion.

I pushed the discussion further and, as a good rationalist, kept on asking more and more tricky questions about dreams and the spirits of the living and of the dead, and the way these manifested themselves in dreams. For example, I said, untruthfully, that I had dreamt of my son the night before, who everybody knew was alive, and asked whether that meant that he had somehow come to me from England. The mood caused by this type of question was totally different to the enthusiastic and somewhat amused explorations about such things as the minds of pigs.

People were uncomfortable and I was told somewhat sententiously that these were too difficult matters and that one could not be sure. We were back with doubt but a very different type of doubt.

Another revealing event occurred while this discussion about dreams and spirits was going on when, unexpectedly, a respected older man from a nearby village, who was on his way to somewhere else, came and joined us, probably intrigued by the gathering. People told him what was going on, and asked him his opinion on the questions we had been discussing. On many points he contradicted what had been said but he was heard with respect, and nobody pointed this out. It is not possible to discuss here many of the issues he raised. I suspected at the time, and still do, that he was making up theories on the spot in order to impress. In particular he asserted that the spirits which came in dreams were not fanahy (the word normally used, and which we had employed up to now) but ambiroa. I was not, and am not clear to this day, whether what he meant was that a different kind of entity was involved or that the proper name for the same entity was ambiroa.

What is sure, however, is that his views on this, and on other matters were either esoteric or at least heterodox, since the word fanahy is the one normally used in the village, as indeed in other parts of Madagascar, although people also sometimes talk of ambiroa (several anthropologists and others, including myself, have learnedly attempted to distinguish between fanahy and ambiroa, as well as several other terms. I now believe that the proper thing would have been to say that these words and the notions to which they allude are not clear).

In spite of the oddity of what he was saying and the fact that he was contradicting what had been said and approved shortly before, the reaction to his views was very similar to the way the views of the other elders had been received. Again people nodded and said 'Ha Ha', implying that they had learned something new and of great value. The flat contradiction with what had been said before was just left in suspension, as it where. It was not that the new actor in the discussion had imposed his view, since, in spite of the expressed approval, everybody, as far as I could see, carried on talking about fanahy after his intervention.

What I believe had happened is that we had reached a type of discourse and a topic which was in doubt. This doubt, however, is of a fundamentally different type to the doubt expressed about the question whether trees could think. There was nothing to be done about removing it. The approval that had been expressed following what the senior men had said turns out, on reflection, to be much less straightforward than I had, at first, taken it to be. It was not a matter of believing that after their intervention we now knew what things were really like. If that had been the case the contradictions between the elder of the village and the passing incomer would have caused confusion.

What I now think was being expressed was something like this: "We are in area in which we are in doubt and where we shall remain in doubt.Those in authority are expressing an opinion but we cannot pass judgement on their views as we are in an area beyond our competence.We listen to them with respect but that does not remove our doubt, nor should it”.

The contrast in the types of doubt illustrated here is, above all, a contrast in what leads to the sharing of doubt and where the sharing of doubt leads in future reflection and action.

The two cases discussed both evoke the word doubt as it is normally understood in English and perhaps both involve similar internal states. But, if we put them within the context of the social and linguistic flow in which they occur and are shared, they have little in common. In the first case, the sharing of doubt is part of a dialogic process which encourages the movement for a joint quest for truth. Doubt is thus a tool to stimulate a forward movement. In the second case we seem to be dealing with a growing shared agreement, on the part of ordinary people, an agreement not to resolve doubt but to stay bathing in it and to delegate the active search for truth to authority, perhaps because one fears where this quest might lead. Doubt, in the second case is not a matter of forward movement; rather it is a device for stopping it.


  • comment-avatar
    José-Luis Guijarro 20 May 2010 (19:05)

    Either am I a real moron who doesn’t understand the point made by Bloch above, or my (very) intuitive anthropological frame is absolutely useless. In both cases, my interpretation of Bloch’s argument is that there are two sorts of [i][b]doubting[/b] processes[/i] in the current discourse of the Zafimaniry villagers: one that makes them proceed towards more understanding and another that stops them going on in the communicative intercourse. Now, from my probably misguided point of view, I don’t really consider this to grant the idea that there are two such processes with different goals and mechanisms … or whatever! It seems to me that in the second case, the [i]same[/i] (or, at least, very similar) doubting process is stopped because they enter into what to them is a mysterious and dangerous field which may be better tackled by people who somehow have got the tools (intellectual, magic, educational or whatever) to do so. It’s a kind of taboo world in which you better be prepared if you don’t want to run into trouble. To me the claim made by Bloch rings similar to the (facetious!) one that would assert that doubting about Hamlet is different than doubting about Napoleon, or doubting about cooking carries different effects than doubting about astronomy. This seems so ludicrous that I am sure I am missing something very crucial in Bloch’s message.

  • comment-avatar
    Olivier Morin 24 May 2010 (23:26)

    You managed to describe your experience with the Zafimaniry in a way that makes it feel quite familiar, and I am sure many readers would agree that these two kinds of doubt are by no means peculiar to your Malagasy friends. While reading your posts, I couldn’t help but think of the various discussions raised by the ‘New Atheists’ (Harris, Hitchens, Dawkins, Dennett, etc.). Some opponents have argued against them on factual grounds – maintaining that the Bible is literally true, that science gives us every reason to think that Mary was a virgin mother, that the Earth was created in 6 days, etc. But these, it seems to me, were a minority. Most of the opponents of the New Atheists were essentially not arguing against them on factual grounds (even though, in the heat of the debate, some of them may have tried to play that card). The roots of their opposition lie, I think, elsewhere. They were attacking the new atheists’ doubts about religion, not so much for being wrong-headed, as for being out of place. For the New Atheists, when a religious institution makes an outlandish claim, that is an occasion where one asks for arguments (your type 1 doubt, the truth-seeking enquiry kind of doubt) For many believers, on the contrary, these claims are an opportunity to express appropriate feelings of puzzlement, wonder, and respect for things beyond our grasp (your type 2, conversation-stopping kind of doubt). When confronted in a debate, a religious person may very well try to rationalize and argue for her dogmas in a type-1 kind of way, but that really isn’t the point. The point in the first place was to express deference to some authorities by nodding and saying Ha-Ha to whatever it is they are claiming. Deferring to religious authorities, after all, carries very little cost in most cases. Most of the claims made by religious leaders are obvious, intuitive (be good to your friend, forgive your enemy) or they bear on matters you and your church already agree upon (opposing or endorsing gay marriage, or abortion). The counter-intuitive claims that spice up the whole thing can be disregarded or explained away (“the story of creation is a metaphor for …”). This argument has already been made by others, and the new atheists have answered it. They point out that entrusting self-appointed representatives of God with important matters is something we should actively discourage (but are these matters so important ? see above for an answer). They also note that someone who does not believe in her religious dogma as she believes that the Earth is round, but merely takes it as an opportunity to express deference and humility, has nothing to fear from the New Atheists. As Dennett would say, they do not believe in God as much as they “believe in belief”. These, I think, are two valid and important points. But why have they have failed to convince so many religious people ? Why is it that most deferrent believers-in-belief refuse to admit that they are Mere Deferent Believers-in-Belief, and leave it at that ? (Actually, there is a religious movement whose members explicitly claimed not to believe in God and other dogmas, but nevertheless thought that such belief was an excellent thing, about which one should follow the guidance of the church : the far-right Catholic followers of Charles Maurras. A smart idea? not so sure.) There are several reasons why so many religious people won’t let themselves be described as type-2 doubters. One might be the fact that their religious tradition (in most cases a religion of the book) happens to insist on seeing the flock as a group of believers – not wishy-washy believers-in-belief, type 2 doubters, semi-propositional faithfuls: believers. But there is also the fact that, for purposes of conversion and ideological struggle, religious people have devised sophisticated rationalisations for claims that were, originally, nothing more than good food for type-2 doubt. These rationalisations are the real target of the New Atheists. But religious people, with good reasons, take the refutation of the rationalisationsto be an attack on their attitude – on the humility and deference their (belief-in-) beliefs are meant to express. Hence their reaction to doubts that, in other contexts, could be expressed, reflected upon, and dealt with in a type 2 fashion. This is what happens when you demonstrate someone – with compelling evidence – that their lover has ugly eyes. They won’t be convinced. They will argue back. What if the lover was really ugly? What if the relationship was alienating? Then you were right – but being right or wrong wasn’t the point.

  • comment-avatar
    Maurice Bloch 25 May 2010 (15:32)

    The problem with taking my argument into the discussions between the religious and the new atheists is that the latter occurs within an environment where people think they know what religion is, even though they have not analysed the notion analytically or historically (thus making inter alia the classic mistake of thinking that the phenomena which have been called religious, are centrally about declarative beliefs). What I was trying to do in the blog was to say: forget about what, using our folk/institutional categories is called “religion” or “science” or even “atheism” instead look at the nature of the epistemic and ethnographic discourse whether it occurs in what people think is religion or science or games. Olivier notes, in passing, that both sides in the controversy may use at different moments both moments of doubt type 1 and doubt type 2. This makes my point. Thus, as is well shown by a number of writers such as Eva Brian Malley, Eva Keller and Olivier actually, some self defined “religious” American fundamentalist, some of the time, argue in all the ways that are typical of science at its best. We know that “scientists” use discourses of doubt type 2 frequently.