Looking Forward, Looking Back

Abstract

This paper surveys the disparate literatures on time, and the relative paucity of metaphors available (based on spatial analogues or mirroring past and future onto one another). Parallels between approaches to the past and future are considered and different intellectual traditions surveyed in futurology, memory, history (chronotopes), archaeology and philosophy. Causation across time, how the past affects the present, how the future may affect present and the past are considered as ways of better understanding how tensed statements in time and of time are essential elements of history and of anthropology. Pluralizing is suggested as a positive step: we should be talking of pasts, futures and even of presents. This has consequences, for example, the Thin Red Line of actuality must be broadened to be perhaps the Thick Reddish Braid. As introduction to a special issue of History and Anthropology I consider the papers that follow and how they contribute to the theme.

Keywords:

• Past
• Future
• Present
• Futurology
• Historiography

[Consider] what it means that humans live in history, in a situation where the future cannot be known and the past cannot be changed and, therefore, where the unpredictable is constantly turning into the irreversible. To live this way is simply an aspect of the human condition; it is a situation that everyone has to grapple with in one way or another, including social scientists and the people whom they study. (Graeber 2012, 25)

The relationship between past and future also runs in the opposite direction, for recollections are open to continual reinterpretation and reassessment. In the world of common sense, the past is often assumed to be fixed and immutable as against the ever-changing flux of the present and the supple fantasies of the future. As Peter Berger (1963, 57) has argued, however, common sense appears to be quite wrong in this assumption: “At least within our own consciousness, the past is malleable and flexible, constantly changing as our recollection reinterprets and re-explains what has happened.” (Cottle and Klineberg 1974, 11)

Thus the subject-matter of history is in an important sense not fact but possibility, not past but future; or, more precisely past possibilities and prospects, past conceptions of the future: futures past. (David Carr (1987, 187) review of Koselleck's ‘Futures Past')

We are drawn to the ancient past not as an entity in itself. Not as a complete self-contained whole, but, to the contrary, because it overspills the past and invades the present and the future. The past lives in the present; and it does so not only in negative ways, as ghosts and wounds that will not heal, but positively, actively, through our thoughts and actions, through our iterations of words, concepts, ancient inventions and ways of doing things, forms of reckoning and forms of eating, in short, forms of being in the world. In a robust and positive sense the deep past lives in the present through numberless quotidian reenactments. (Trautmann, The Clash of Chronologies, 2009, xv)

In this paper I summarize some of the cross-cutting themes which connect historians, anthropologists, archaeologists and, perhaps controversially granted the journal title, futurologists. These themes connect researchers concerned with evidence in different forms and what it tells us. They also connect researchers concerned with the stories or narratives we tell on the basis of our research.

That holds when considering the past, and the future. The work leading to this paper (and the special issue that it introduces) was part of the Care for the Future research programme (funded by the Arts and Humanities Research Council (AHRC UK)). Reflecting on the past enables us to think differently about the future, about the sources that researchers, perhaps far, far in the future, will use when they do their research or could have used had the material actually been archived. (It might help to think about the parallel of doing tomorrow's archaeology today. This prompts questions about what might enter the archaeological record, and what archaeologists will make of it in, for example, 2000 years time?) It may also help as we plan and construct archives of and for the future especially in light of the archival fragility of the digital record in the longer term (see discussion below and Barone, Zeitlyn, and Mayer-Schönberger 2015, for examples).

Metaphors

All too often academic debate is confused by having too many metaphors (and authors move accidentally between them (archives being a case in point, see Zeitlyn 2012a)). When the topic is time, I think the problem is the opposite one but with the same result: confusion. In other words, when the topic is “time” there are too few metaphors available so everyone lands on the same ones but to different effect or purpose.¹ We spatialize, mapping one dimension onto several, and talk about moving or travelling through time (and then trip over ourselves about the asymmetry: we cannot return to where we started, there are no circular paths, time travel does not happen²). The interconnections may be illustrated by Ingold and Vergunst's (2008) reflection that “to follow a path is to remember how it goes, making one's way in the present is itself a recollection of the past  …  onward movement is itself a return” (quoted in Macfarlane 2012, 185). Moreover, we should also note that the idea of “onward” or “direction” itself is spatialized and assumes a present orientation to the future, to the destination where one will be; and, as Heidegger has it, the destination may be more “proximate” (salient) than our immediate location (1978, 142). In short, there are limitations or dangers of confusion arising from spatial metaphors applied to time.

Apart from space, the other major metaphorical resource for considering both past and future is the complementary tense. We play games with parallels (symmetries, asymmetries) between tenses, between the future and the past. (Here I must plead guilty with the rest.) So we talk about anticipating or predicting the past, remembering or forgetting the future. This may be enjoyable and provocative but provocative of what?

Guyer (2007 recently taken up by de Abreu 2013) makes parallels between formal economics and Christian prophecy. On her account, both of these traditions “evacuate the near future” (as Guyer terms it) because their practitioners have their eyes on the horizon, on the long term (cue formulaic invocation of J.M. Keynes “in the long term we are all dead”) or the Apocalypse.

I see a parallel between such emptying of the future and historical myopia: obsessions with innovation and change lead to repeated versions of the same mistakes and what we could term the “evacuation of the near past”. The past, including the recent past is dismissed as not being relevant since now everything is different. This is a form of millenarianism. In the field of high technology, the caricature phrasing is “that's so 1.0 or even 2.0 thinking  …  Now we are 3.1.1” and so forth. (The initial reference here is the distinction between the original World Wide Web (Web 1.0) and the so-called Web 2.0 which arrived with the development of social media platforms. In the interest of full disclosure readers should note that this is a grumble from a middle-aged man using a Nokia 1302 non-smart phone in 2014. However, this is relevant to world events not just technology: consider the world response to Bosnia, Iraq, Afghanistan and Syria. Is there any evidence that we can learn from the past? (Cue formulaic invocation of Marx “The Eighteenth Brumaire of Louis Napoleon”, history repeats itself “the first time as tragedy, the second time as farce” which we must recognize as often a bloody or deadly farce).

Shryock and Smail (2011b, 19–20) suggest some new metaphors for bridging the gap between conventional history and the more than three million span of what they call “deep history”. These are kinshipping, co-evolutionary spirals and fractal perspectives. These may well help cross bridges not only into the past but also into the distant future, far beyond even the scales considered by the Humanity 3000 project (see Inayatullah 2012).

Disconnects in the Literatures

How to map the literature? There is not one literature there are several. We must pluralize to pasts, futures and literatures (and as Reason argues in his paper we must add to this list a pluralized idea of presents, a suggestion I have adopted below). My overwhelming sense is of people talking past each other, of ideas slipping past one another in the night, failing to connect when they could connect. However, the most radical disconnect is between the phenomenological experience of time (kairological time) and reflection or theorizing upon time (what could be glossed as forms of conceptual time). As Moore discusses (2015) watching a seventy year old film can precipitate the experiences displayed for some of the audience: they feel no chronological distance. Somehow recognizing these broad distinctions as separable is less troubling than confusions or parallels within each one. In what follows most of the discussion is about conceptual time.

The work on conceptual time can be partitioned into five different literatures (or academic traditions).

1. Futurology: Barbara Adam starts from a non-deterministic stance in which humans have agency and so can have real effects on the future. The argument of the monograph Time Matters (Adam and Groves 2007) follows from the vision of an open future to which we have a responsibility. Their book therefore maps out our ethical responsibilities not only to our current contemporaries but to our future contemporaries and those of our descendants. Adam and Groves encapsulate much of their argument round the distinction between what they call the “Present Future” (as opposed to the deterministic (or pre-ordained) “Future Present”). In his monograph Samuel Collins also argues for an Open Future in which our understanding of past histories can be used to rethink what is possible (hence the importance of utopias), in other words to think into futures:

This, more than anything else, is the contribution of anthropology to the future: building the anti-time machine by not only preventing us from time travel along linear spatialities, but in collapsing present, past, and future into a palimpsest of virtualities. Thus, an anthropology for the future, rather than of the future. (2008, 125)

Linked to Adam and Groves’ style of approach is that of futurologists who work with businesses and governments as these large-scale institutional actors consider and plan for various futures. Scenario planning helps organizations plan for and anticipate contingencies. They can rehearse possible action packages. This has a parallel in Robert Textor's “anticipatory anthropology” (Textor 1980, 2005; Strzelecka 2013). Increasingly it is being generalized and taken for granted. So Georgina Born discusses the importance of forecasting (and market research) in contemporary capitalism, as:

Central to this dynamic experience of time is the existence of retentions—memories or traces of the past—and protentions—projections or anticipations; thus, the future is experienced as a protention of possible eventualities—a construct of the present—just as the past is experienced through a retention of previous events—also a present construct. (Born 2008, 289–290)³

In Adam and Groves’ terms a “Present Future” indeed.⁴ Where profession futurologists provide services to large multinational companies running scenario planning, one term of art which is useful to recognize is the idea of “backcasting”. Modelled on forecasting it emphasizes the relationship between the present and a desired future. To backcast one asks (if this is what you want to be the case, what should you being doing now?) (Robinson 2003).

2. Histories—chronotopes/chronotypes.⁵ Benjamin (1968) and Koselleck ([1985] 2004) consider different approaches to the creation of histories. These may be deterministic or not, and are often characterized by whether they describe a Fall (or decline from a past Eden) or Progress (towards a future Utopia). Note that these distinctions connect pasts, presents and futures: a present understanding of how the world works affects both one's understanding of the past and that of the future. Trautmann (2009) discusses the nineteenth century rupture in European/North American thought of shifting from a past time horizon of c. 6000 years to that of millions and billions of years. Particularly he discusses this in the context of Europeans confronting Indian intellectuals who already thought in terms of much larger past time spans than was then usual in the Western tradition.

Hirsch and Stewart dedicated a previous special issue of History and Anthropology (2005) to ethnographic approaches to historicity: examining ways in which views of the past and its relation to present and future is culturally constructed. Subsequently Hirsch and Moretti (2010) have explored the cultural and historical specificity of ideas of a Universal Past and Universal History. This is exemplified in Busse's (2005) exploration of historical bases for culture heroes in the Southern lowlands of Papua New Guinea. He uses this to challenge the distinction between mytho-history and what we might (somewhat tongue in cheek) dub history-history, or Universal History, that is the products of conventional Western academic history. Building on this, Hirsch and Moretti (2010) want us to accept the existence of culturally specific historicities. Inspired by Susannah Radstone's idea of “regimes of memory” (discussed below) this could be glossed as arguing for different “regimes of history”.

Palmié, Khan, and Baca in their collection inspired and honouring the work of Sidney Mintz recognize the historical specificity (and constructedness) of both anthropology and history (2009, 3). They are clear that such recognition should not paralyse the disciplines, reducing them to forms of hand-wringing “nothing can be done” naysaying. Indeed they and the authors in the collection take Mintz's work to exemplify how to take a nuanced empirical approach which is sensitive to historical flows and the effects of different scales of impacts and interconnection. Processes of social and historical construction do not occur from nothing, and are not without physical constraint (see also Zeitlyn and Just 2014).

3. Memory. There is a large literature on memory which really covers too much ground to be helpful (so argues Golden 2005). There are studies of personal, historical and social memory, which rely on metaphoric resonances between them. Casey (1983) revealingly explores some of the ambiguities of “keeping in mind”: there is not a single entity for analysis despite what English terminology suggests. Both Ricoeur (2006), and Borges in his story about Funes Memorius the man who could not forget (1964), turn upside down the logic of the saying: “to forgive is to forget” (in which forgiving must proceed forgetting) for them, rather, in order to forgive we must first forget (Bienenstock 2010, 332, see also Bryant 2014). But individuals and societies have different trajectories for this. Survivors of trauma may want to put their trauma out of mind, and act out a kind of forgetting. Examples of this are discussed in several chapters in Historical Memory in Africa: Dealing With the Past, Reaching for the Future in an Intercultural Context (Diawara, Lategan, and Rüsen 2010). Despite such individual forgetting (or silencing, which is not the same thing), socially it may be important that the world does not forget (twentieth century history does not encourage the idea that this is successful).

4. Archaeology (distant pasts)—Several authors discussing archaeological theory (such as Shanks 1993; Olivier 2001; Gosden and Marshall 1999) question the subject's obsession with sequence. They argue that there is not a single sequence (even as suggested by McTaggart's B-Series⁶), there are many. An archaeological example is that of Bronze Age Grave goods where archaeologists have found what is called “fine ware” in male graves and “coarse ware” in female graves. But the different types of pottery yield different sequences, different chronologies and hence different archaeologies (McGlade 1999, 143). This is reminiscent of Gell's comment (reflecting on archives): “it is a category mistake to attribute dates to objects at all; because only events have dates” (1992, 28). Objects have histories not dates.⁷

McGlade also distinguishes chronological from kairological time (1999, 144), a distinction mentioned above. Kairological time is the phenomenological or experiential sense of time, missing from much archaeology for obvious reasons (see Tarlow 2012) as well as being missing from most of the other approaches to temporality for less good reasons.

5. Philosophy: There are several debates in philosophy which are relevant here. Some have developed temporal logics based on various forms of non-standard or multi-value logics. These try to formalize the conundrums posed by the grammatical similarity recognized by Aristotle between (a) “It is raining now” (or “it rained yesterday”) and (b) “It will rain tomorrow” (the problem of “future contingents”). Is it right that one statement has a truth value and the other does not? (see, for example, Prior 1967). On the other hand the philosophy of physics complicates our ideas of the asymmetry implied by the “arrow of time” since most physical processes are symmetric and can run backwards as well as forwards in time (see Price 1996 as discussed by Reason, this issue). Another philosophical approach to determinism, history and the future is often labelled the “Thin Red Line (TRL)” and this I will discuss in the next section since it takes us from the general survey of different literatures to some of the more specific issues discussed by the papers which follow.

Plural Times and the TRL

The “TRL” of actuality threads its way through myriad “roads not taken”: the possible worlds created by choices made and those options avoided.

It is quite common in the UK when talking about the past to think about these “Roads Not Taken”, especially with reference to Frost's poem (1920), and to engage in historical speculation about what might have happened if things had turned out differently. People use simulations to rerun history on the basis of slightly different parameters to see what difference they might have made, engaging in different forms of counterfactual history.

We could call this the study of pasts (or times past or past times) not the past. I emphasize the plural. As mentioned above, I want to make plural (pluralize) not only the past but also both present and future, so rather than thinking about past, present and future I want to think of pasts, presents, futures.⁸ I will briefly consider each before continuing to discuss about how these may interconnect.

Pasts

I have already argued for pluralizing the past on the basis of hypothetical or possible scenarios. There is another way in which the past is or should be seen as plural: multiple viewpoints, different interest groups and different scales have different histories, different understandings of what is going on and how things are unfolding. Influenced by the Indian Subaltern Studies group on the one hand and arguments from feminism and gender studies on the other, we are now more sensitive to ways in which women's history might be a very different thing from men's history, how the accounts of powerful (male) actors might obfuscate (and be misleading) about events of wide significance to many. And of course this is all before we started including ethnicities in the mix of different perspectives. It is not only the large centralized groups that are or should dominate the writing of history (with apologies to my friends and colleagues from such groups). The smaller groups may be beneath the regard of the rulers but in the abstract we need to recognize that they too have their histories and that these may be very different from those of the larger groups.

As David Zeitlyn and Bruce Connell argued some years ago (2003) discussing a specific West African part of history, the situation is a fractal one: at every level there is a different history belonging to a different past. These resemble but are not the same as the histories (pasts) from other levels. Explanatory (hermeneutic) frames will differ, and, as Thomas and Thomas noted, “if {people} define situations as real, they are real in their consequences” (1928, 572) which we could paraphrase as “beliefs have (real) consequences”. One conclusion from this is to move from thinking about the past and its singularity (and with that the idea of a concomitant single history) to thinking about plural pasts (and of course related histories). I have already discussed Hirsch and Stewart (2005) and Hirsch and Moretti's arguments for pluralization, implicit in their moves away from holism. Indeed Hirsch and Moretti's chapter closes with the following words:

our “partial” comparison aimed to highlight the fact that Universal History and Universal Past are intrinsically “partial wholes,” whose holistic narrative conventions about the past embody and promote important “interested” (political and moral) claims and counterclaims that deserve close anthropological attention. (2010, 295)⁹

These partial, interested claims can amount to the making of new pasts.

David Carr makes a similar point in his review of Koselleck's important book Futures Past: “As new concepts emerge or old ones transform we revaluate the past in the light of these new theoretical spectacles hence new pasts emerge as present understanding changes … ” (1987, 197/198). However, as David Reason pointed out to me (p.c. 20 July 2013) this formulation does not clearly distinguish between “different pasts” and “different versions of THE Past”. The latter implies that there are “dated” propositions about the past on which we are all agreed. The former implies that this may not be the case, and that is the line that I am arguing for.

Presents

Having argued for different pasts it is easy to apply the same style of argument to the present, as Reason points out (this issue). Along the same vein Kubler discussed “the plural present” (1962, 129). Different groups have different understandings of the present. It is best seen as multiple not singular. The definite article is misleading: there is not one present with a single truth but a series of linked presents each with its own (and interconnected) set of truths. I should stress, in the interests of full intellectual disclosure, that I do not see this statement as committing me to any form of hard relativism (the clue to empirical commitments as setting constraints on milder forms of relativism is in the word “linked”, see Zeitlyn and Just 2014). On the basis of the interconnections between past, present and future Baert argues for what he calls a “reconstructive present” (1992, 4). For him “the present is that in which the representation or meaning of the past and future are continually altered. People regularly re-assess the past and as a consequence, start thinking differently about the future” (1992, 78). As part of this work of reassessment archives, objects (belongings) and, architecture have important roles to play as Bryant (2014) has explored in contemporary Cyprus where civil war, partition and the partial lifting of the partition make memory and ownership (current and future) fragile, politically sensitive and disturbing (uncanny). Starting with familiar suburban landscapes in Northern England Edensor (2008) talks of “mundane hauntings” when discussing how evidence of other lives, past ways of occupying space, obtrude like palimpsests through the everyday. Thus architecture is a case of pasts affecting presents, and, like it or not, our buildings are full of ghosts (see Collins, 2015). Such is the potential ghostly yet political power of an archive or exhibit that plans to create them may not be realized (Le Febvre, 2015).¹⁰

Similarly Dave Reason has written (p.c. email 27 July 2012 see also this issue):

we tend to assume that there is less of a problem knowing the present than in knowing future or past. However a moment's reflection tells us that we ordinarily and routinely don't know what's going on precisely “now”, and often revise our warrantable accounts of some “present” in response to our appreciation not only of unfolding events but also of a new understanding of the past.

And I would gloss this (for both Baert and Reason) that such a position implies pluralizing the present.

Futures

Multiple futures may seem the thing of science fiction (and quantum physics). Certainly many fiction writers have made much of the possibility of different futures. What will the world be like in the future if this happened, what would it be like if that happened? Philosophically we quickly get into complex issues which have been discussed at least since Aristotle who asked whether there is a truth about who will win tomorrow's sea battle? Is the problem an epistemological one—we simply do not know what will happen (but there is something to know)? Or is the world less structured so, here and now, there really is not anything (any single object of knowledge) to know about tomorrow? Many philosophers see the TRL as ceasing in the present and not extending forwards into the future. For them it threads backwards only. Otherwise we live in a deterministic universe in which we only have the illusion of choice. Some people indeed think that such is the case, especially those practicing particular forms of various religions of the book. However, albeit contentiously, the future seems to be open; we make decisions and act in the world in ways that affect the course of events. On this view, the TRL ends in the present,¹¹ and is only traceable in retrospect. In other words, the TRL does not extend into the future but is created in the process of becoming (Belnap and Green 1994). Only in a deterministic universe can we establish where the TRL will go and shape our present decisions accordingly (but see Bourne 2011 for discussion of the logical implications of presentism, the doctrine that only the present is real and therefore that past and future statements lack truth values). Rosenberg and Harding in their introduction to a volume exploring Histories of the Future, identify an inverse connection between determinism and modernism with its promise of progress. As they have it “the modern was constituted through a rejection of prophecy” (2005b, 4) since prophecy assumes a knowable future in which the TRL can be traced forwards as well as back.¹²

Adam and Groves in their key work Future Matters address this as follows:

A true future orientation … is only possible when the future is no longer pre-given as future present but arises from actions in the present. In our terms this is the difference between the providential future present and progress tied to the creation of present futures. (2007, 53)

Hence work on divination, forecasts and diagnosis appear differently depending on the philosophical position taken (see Zeitlyn 2012b). However, when we change our understanding of the past our understanding of the course of the TRL also changes. The past may not be as unchanging as it appears. Archives and Museums have to actively manage their collections in the light of changing understandings of what their collections contain and what they mean in the present. Reflecting on this, Amad encourages us to view an archive as a bet against the future—betting that these records will be found useful (2010, 1).

Bertrand Lategan builds on Koselleck's focus on the interaction of experience and expectation to argue that sense making (of or about the past) has a future orientation (2010, 147). He accords ontological priority not to the real but to the possible. This opens up connections between pasts and future: “It is exactly when events do not follow the anticipated course or do not comply with the expectations of common (that is, historical) wisdom, that the future potential of history becomes visible” (151, his emphases). An open future is assured by stress on past possibles not the real (in his words, stressing the “priority of the possible”, 152). This is a break with determinism but has its own problems: who is to say what was possible? It allows him to “remember with the future in mind” (157/158). “The contrast between possible and not possible is not determined by the past or by what already exists. It therefore has the capacity to deal both with risk and with what is new” (2010, 158). One implication of this is that we have to revise our idea of “the TRL”: either blurring the line so it becomes a ribbon of possible worlds, or relativistically dropping the definite article, so each viewpoint has its own TRL. Admittedly both such alternatives are difficult. On the one hand there are problems assessing what is possible, and on the other there are problems dealing with intersubjectivity. Lategan urges us to recognize the future inflection of history. In that spirit we might begin to talk of ways in which we can “anticipate the past” just as we can “remember the future” (Farriss 1995).

Finally, in this section, let us consider another way of approaching this by trying to develop an approach which is symmetrical to both pasts and futures. So, for example, on the basis of Susannah Radstone's idea of “regimes of memory” (which encourages us to appreciate the social constructedness of memory¹³) we might start to think of “regimes of anticipation”. These could be described in Hastrup's phrasing as both performative and processual: “We perform a world into being, acting as much upon anticipation as upon antecedent” (2005, 11). Such an approach gives us a way to think about how we “care about the future” and how present cares, attitudes and decisions, shaped by the past, have constraining influences (to say the least, to be deliberately modest) on subsequent futures.

Present actions are future orientated in various different and interesting ways. (They are also past orientated.) It seems that there are important differences depending on whether we are thinking about short-term or longer term futures (see Munn 1992, 107). Note I am interested in this because there seems no philosophical or conceptual difference but there are dramatic differences in attitude. Mundane prediction, anticipation, is pervasive and may be dismissed as seemingly banal (wrongly in my opinion):

I see a car approaching in the distance but know that I can cross the road safely before it gets here.

As you talk, I anticipate that you are about to stop speaking and prepare to respond.

I see the ball coming and raise my hands to catch it. (qv Bourdieu 1990, 80–81)

I turn on the oven so it can warn up before putting in the roast.

I post the card a few days in advance of a birthday so it can arrive in time.

I know Christmas is coming soon so I start looking for presents.

Other examples cover longer durations: planning, preparing and cooking food for a feast and possibly growing crops.

Linguistic performance (for once both written and oral) also mixes short-term pasts and futures. As I read or listen, I keep in mind what I have just read or heard as I process the following parts of a sentence. In long sentences and especially in SOV languages (in which the norm for sentence word order is subject, object, verb; Latin being the classic example) one has to hold most of a sentence in mind before knowing the action involved. And in conversation one anticipates the end of a sentence in order to be ready to respond (as much conversation analysis demonstrates the gaps between speakers are usually extremely small). So our presents as linguistic actors are actually flowing amalgams of recent pasts, presents and near futures. All of which muddies the philosophical waters. In other domains we have to deal with duration as essential parts of the objects of study: music, the spoken word and film or video. Because of this Amad considers “film as an archive-inflected medium” (2010, 21). We can generalize and talk of human experience as time-inflected.

Phenomenologically, the present is not a flowing instant but demonstrably has extension or duration. I will not try to resolve the philosophical conundrums but I note that there is considerable philosophical interest not only on the basis of problems arising from physics (quantum theory and cosmology) but also arising from some of the issues I am referring to (see Price 1996; and collections edited by Callender 2002; Baert 2000). I discuss some recent uses of Bergson's exploration of duration below.

Farming seems to fall on the cusp between short-term and long-term prediction. The routine of the annual cycle robs it of the sense of uncertainty which accompanies other examples of forecasting, although when natural disasters results in crop failure we realize the uncertainty was always there (see Bourdieu's early discussion of Kabyle farming: 1963).

As the extensive literature on different forms of millenarianism makes clear, there are problems for humans who take literally injunctions such as “Live every day as if it were your last”. In the terms I have been using, millenarianism could be described as a confusion between short-term and long-term futures.¹⁴ Although I am not happy at such pejorative language I think there is something in the idea of a dissonance between short-term and long-term orientations which could be helpful in ongoing discussions of various forms of millenarianism. Dave Reason (p.c.) has suggested that we need a “string theory of time” in which every “moment” is a slice through what he calls “the more fundamental tufts and skeins¹⁵ of a temporality constituted of threads of intention”. This is to formulate for Time a concept which does for temporality and temporal relations what “proximity” (Heidegger's usage¹⁶) does for space; we need an idea of “temporal proximity”. In other words, to know who I am/the society I live in now implies I can remember/understand the something of the past (but not too much), and I act now in anticipation of being around to reap the consequences of my actions (or bearing them). So, as was suggested above, all of our “nows” are sliding ensembles configured upon and configuring pasts and futures, what Schütz calls the “vivid present” (1945, 540). And every present moment includes within it, its own casting (representation) of the past and future. Present and future pasts, present and future futures are not necessarily the same, although they often overlap to the extent that they invite the mistaken reading (usually of the past) as being invariant. Graeber is keen to think about radically different futures and pasts. However, even he recognizes “the paradox of fortune (that those things we cannot predict in the immediate future will seem inevitable after they have occurred)” (2012, 37).

This makes me want to ask why short-term prediction, as mentioned above, is mundane to the point of being banal but predictions into the longer term are fraught with difficulty, some (Taleb 2008; Gardner 2010) would say impossibility? Predicting that it is safe to cross the road is different from predicting the winner of a horse race five years in the future. There is a continuum between catching a ball, brewing beer, planting crops, and  …  eventually, the end of the world. Somewhere along it we start to feel that quantitative change has become a qualitative difference. However, I want to underline that, in terms of philosophic logic, such a difference seems hard to justify, as well as possibly chimerical if no clear distinctions between pasts, presents and futures can in fact be maintained.

Correspondingly how do we retrodict or project back¹⁷ into the pasts? Kate Moore has undertaken a study of how contemporary (field research undertaken in 2008) Kwanyama people now living in Namibia understand films of their ancestors (ancestors both literal and metaphorically generalized, sharing a common ethnicity) shot in neighbouring Angola by the Powell Cotton sisters in the 1930s. Her analysis uses the ideas of mimesis and identification (and could also link to Braddock's application of the concept of countertransference to ethnographic enquiry; 2010). Viewers share the “work” with those on the screen, even when this includes entering trance (“work” in Kwanyama); watching a film on material culture a potter makes the gestures she is seeing (Moore 2012, 2015). Such embodied viewing collapses or bridges time differences, although I note that I think we would be a lot more cautious if Moore were reporting her own responses to the films. This fits well with Nadia Seremetakis’ multi-sensorial approach to memory (1996) in which Proust-like the smell and taste of fruit, or the role objects such as embroidery collections, function as series of mnemonics. For the embroideries this was: who gave them, where and when they were made, and through this linking to other events. Hence, she sees memory as part of the engagement of the body with the outside world so never only or purely cognitive. Kate Moore's viewing of the film is profoundly different from that of a local potter for whom the film triggers the gestures of potting. As I have been arguing above we need to think about multiple pasts, presents and futures. The same material evokes different memories, different pasts.¹⁸

Moreover, ritual from the 1930s has consequences eighty years later. Action in the 2010s which led to the inclusion of the Powell Cotton material in the planned Kwanyama local museum may be seen to be contributing to an argument “that Kwanyama identity has long been thus”, concretizing or crystalizing it and locating it in time at least back into the 1930s, so Moore's actions may have consequences eighty years earlier.¹⁹

Discussing images of slaves, and the incompleteness of the visual record of slavery, Stephen Best uses the idea of deferred action: only when an event is recognized (later) as significant did it “take place”, does it become remembered. Memory happens when we remember, and archives are full of present events, the event of the researcher stumbling upon and “creating” the past (2010, 152). When I listen to a sound recording I can hear my grandfather's voice. Looking at a photograph I can see his face. Controversially, the art historian Kendal Walton describes this as looking through the photograph, as seeing across time (1984 ²⁰). At some level of metaphorical remove there is a strange form of sympathetic magic at play here, hearing a voice across the decades, seeing a face, holding a lock of hair. Voices from the grave have emotive force on those who hear them. Conversely, there is redemptive force as we put names to the anonymous dead, as in Jules Michelet's “resurrectionalist history” (which partly anticipated the work of Foucault in the nineteenth century). Michelet's history, brings “the dead to life” especially by giving them names (more accurately: restoring knowledge of their names) (see Zeitlyn 2012a, 2014, 2015).

This train of thought started with some of the differences between short-term and long-term predictions. Such differences raise many types of issues. There are psychological issues, cultural issues, and social issues. These are deeply interconnected which makes it hard to address them separately or meaningfully.

The examples of short-term prediction just given seem very different from actions taken on the basis of prediction and forecasting (mentioned above under Futurology), where it seems we are consciously changing the future. The clearest example of this is of an education ministry or a city administration faced with demographic data which shows an increased number of babies being born. Projecting this into the future it seems there will not be school places for those babies once they are of school age. The response is to build schools. If the schools had not been built there would have been a crisis when the babies reached school age. The schools were built so there was no crisis. Anticipation, prediction, followed by action to avert something, leaves the predicted futures hanging in ways that resemble the roads not taken in the past (see Carr 1987). As in the case above, averting the school-place crisis has become an achievement. The lack of school places has been made into a hypothetical: the potential crisis has had a real effect on actuality, on the actions taken to make the crisis non-actual. So even a non-deterministic future can include a TRL threaded between counterfactuals after all. This establishes some common ground between futures and pasts. The specific philosophical outcome of this practical piece of management is that we have a problem assessing the truth, accuracy or validity of the prediction “there will be no room left in the schools”, especially granted action taken specifically to make the predicted outcomes not come to be.

Interconnections—Causation, Ante- and Retro-causation

How does causation work across time? In the last case it seems a possible future caused present action (building schools) which meant that that particular future did not arise. This could be described as a form of ante-causation, or teleological action. Adam and Groves make the same point in somewhat less contentious language: “engagement with the future is an encounter with a non-tangible and invisible world that nevertheless has real and material consequences” (2007, xv). In the same vein Taddei (2013) discusses forecasts as promises: as speech acts or performatives hence driving action.²¹ This is perhaps another case of the real consequences of beliefs (Thomas and Thomas 1928, 572, see above).

What then of retro-causation, changing the past? One set of examples may be summarized by the wonderful slogan “If you do not like the past, change it”.²² David Lowenthal reviews ways in which the stuff of history, monuments and memorials change and are changed as different present understandings of the past themselves change in ways which are deeply political: think of the way the Spanish Civil War is treated in 2015 by comparison to how it was regarded under Franco. Similarly, Baert talks of a retrospective determination in which the past is made fixed by our reflection upon it: “It is only after its occurrence that people reconstruct symbolically the past so that it acquires the status of ‘inevitable’, ‘unavoidable’ or ‘predictable’” (1992, 81, his emphasis).

Another approach may be taken from Freud. Freudian theory uses a term nachträglichkeit which is helpful to us here. Although originally translated into English by Jones as “deferred action”, more recent work suggests “reconstrual” or “retroactivity” as a better gloss (Thomä and Cheshire 1991 see also Marion 2012). It helps us see how pasts, presents and futures inter-relate: we reconsider (either pasts or futures), we change our understanding and the actions that follow from that understanding. Looking back at the past, we may change how we think the course of the TRL was traced. Looking forwards, into the future, we alter the course it will take (for example, by building schools in the present), so causality can be seen to extend both forward and backwards in ways which can be rhetorically described as altering pasts and futures. Gell (1992) would insist that this is rhetoric, and that it uses a model that presumes linearity of McTaggart's A-Series.²³ Baert expresses a related caution, holding that the past is both revocable (what it was) while remaining conventionally irrevocable (that it was) (1992, 79).

Thomä and Cheshire are clear that the work of reconstrual consists of the present processing of memory traces, of material relating to the past (1991, 413). This resembles Haug's idea of “memory work” or “re-membrance” as used by Radstone (2000, 18) and can also be used to unpack Lacan's “retro-action” whereby “l'effet d'après-coup is effective along the reverse vector also, from the present towards the past” (Thomä and Cheshire 1991, 421/422). Lacan is using a rhetorical extension, describing a present change in the understanding as a causal reconfiguration of the past. It is important to recognize how close the two senses can be. In cases of trauma, violence and dispute, the acts of recognition and relabelling, sometimes long after the event, can be an important part of a process of healing. Healing possibly but also maintaining political struggles: clashes over archives, museums and archaeology, over how to care for the past have present and future significance in, for instance, Palestine/Israel, in Iraq, Afghanistan as well as in Kenya (both dealing with Mau Mau and the riots after the election of late 2007), Cameroon (Deltombe, Domergue, and Tatsitsa 2011) and in the UK in Northern Ireland.

Perhaps the most subversive approach to interconnection between pasts and futures in a temporal chiasmus comes on the one hand through the idea of duration, originally from Henri Bergson via Deleuze (see Turetzky 1998; Hodges 2008 as well as Hirsch and Stewart 2005) and, on the other, from Serres on the crumpling or folding of time (see Abbas 2002). The idea of duration is used both to embrace the idea that simultaneously we think at different rates and over different scales²⁴ and the way in they interconnect: durational time can fold upon itself, supposed effects becoming potential causes. Such ideas have been used productively in different situations to look at how ideas and experiences in more or less distant pasts and futures affect contemporary events. Knight (2012) discusses how the Greek famine during the Second World War affects current responses to economic crisis in Greece. He takes the idea of crumpling from Serres and moves from the juxtapositions this can create²⁵ to talk about “cultural proximity” (Knight 2012, 357). In the current economic crisis the “Great famine” feels close and people behave in ways similar to those reported from the war years. In a very different context, Nielsen uses the idea of duration to explore the paradoxes for those constructing houses in settlements on the outskirts of Maputo where many start but few finish building their dream houses. In the local metaphor they are cultivating a future for their children by building even when they cannot afford to finish the project (2011, 406). Riffing on Bergson he concludes (2011, 416) with a vision of the future gnawing on the past, swelling into the present.²⁶ However, I am not sure we need all the theoretical scaffolding to appreciate the richness of the ethnography he presents. Munn's “temporalization” might do as well; she argues for a view of “time as a symbolic process continually being produced in everyday practices. People are ‘in' a sociocultural time of multiple dimensions (sequencing, timing, past-present-future relations, etc.) that they are forming in their ‘projects’” (Munn 1992, 116).

As mentioned above, the phenomenology of the present starts with duration. It seems to be relatively independent of philosophical debates and clearly may vary between cultural traditions, as a form of temporal dwelling (Bhabha 1994, 10). Grosz has explored (1999, 2004, 2005) some of the ways in which memory works to bring the past into the present as an active working factor. She puts the emphasis on the processional nature of durational time, seeing it as a form of emergence or becoming:

While duration entails the coexistence of the present with the past, it also implies the continual elaboration of the new, the openness of things (including life) to what befalls them. This is what time is if it is anything at all: not simply mechanical repetition, the causal effects of objects on objects, but the indeterminate, the unfolding and the emergence of the new. (Grosz 2005, 110)

What all these authors (and I could cite many more) are grappling with is the paradox of the affective and effective or causal co-presence of past, present, future (see Trautmann's epigraphic quotation) and the shared sense that the present is qualitatively different from both past and future. To add still further to the complexity one way of thinking about multiple futures leads to questions of their relative likelihood. When we ask “which of these futures is most likely to happen?” we encounter a different set of theoretical issues about probability.

Causation and Probability

What does it mean to play the odds? What does it mean to play the odds and lose? Is there any different between these questions? An American-style weather forecast reports 20% chance of rain. On one day I go out without an umbrella and it does not rain. On another day I do the same thing and get soaked. Apart from the obvious, is there any difference in the forecast between these events? Is the unpredicted rain parallel to the unbuilt school discussed above?

These issues arise in Gould's account of surviving abdominal mesothelioma (2007 (1985)). We should note that he survived for more than twenty years after diagnosis for a condition whose post-diagnosis median survival rate is eight months, as he discussed long before his eventual death from another illness. This helps to make clear the difference between medical prognosis, sketching the range of outcomes in a population, and prediction of the outcome for a specific individual. Prognosis is explicitly a projection mediated by statistics of the past onto the future: in Gould's case the doctors tried to hide the bleak prognosis from him. For an individual this is why the language of doctors can seem unhelpful. A prognosis explicitly does not predict what will happen in an individual case.

So Wagner remarks that “In poker, the hand you are dealt is a perspective on the entire set of cards” (2012, 170). He uses probability in order to link to Viveiros de Castro's perspectivism. However, I am not sure if that is correct. A hand of cards is actually a selection (a sample) from which you make inferences about the other players’ hands. If you are lucky that's how it plays out … If it does not, that does not mean your inferences were wrong in principle, only that they were in this particular case. This is the difference between a statistical approach to a collectivity, a set or sample of many deals, and unquantified concern with a single, individual case (which may be idiosyncratic). You may bet your house on a single game of poker and lose to a quirk of the cards. It may be a consolation that it was a very unusual (unlikely) outcome but it is the one that did eventuate. (Those running casinos are happy to pay out to the lucky few, secure that the statistical balance will remain in their favour, so the casino owners but not the players benefit from statistical perspectivism). Some cosmologists take a similar position to the universe, on the basis of what they call the anthropic principle. We can infer quite a lot about the boundary conditions, the starting point for the universe we live in, from the fact that it has at least one planet stable enough for quasi-intelligent life to evolve on. Most values of the boundary conditions do not produce such stable planets, so our universe is unusual among the infinitude of possible universes. We have won that deal, at least for the time being.

The Philosophy of Time confuses matters further. Or rather, it is more accurate to say that work on the Philosophy of Time reveals that our conceptual basis is confused. There are paradoxes associated with wave-particle duality which challenge our conceptual separation between now, the actual, pasts and futures. In a classic experiment light is passed through two narrow slits, and produces a characteristic interference pattern (see Figure 1).

Figure 1. Two-slit interference pattern. https://upload.wikimedia.org/wikipedia/commons/8/87/SodiumD_two_double_slits.jpg.

A photon goes through one or other of two slits and, as it were, bounces (scatters) off the edge of the slit. Over time (as many photons pass through the slits) the results create a familiar, classic, wave diffraction pattern. So an individual photon is somehow “constrained” by past and future tracks of its fellow photons even though the light density may be so low that there is only ever one photon in the device at the same time. In this scenario (which has been repeatedly performed in many laboratories) the individual photons in some sense “know” where their fellows have gone and will go as they “decide” their course through the apparatus. It is very hard to avoid such intentional language when describing the events which build up the interference pattern. Although each photon travels blindly and interacts in a random fashion the net results are predictable, since the physical world is predictable, “as if” they were following the probability distributions. This may be true without us knowing anything about what happens in a single case (which is unknowable for individual photons). This points to the difficult (and different) problems associated with probability distributions, the relationship between an individual case and the population to which it belongs. Any one individual may be an outlier but, just as the American Dream is a misleading myth since we cannot all end up winners, as millionaires, not every photon can be an outlier.

There are other issues arising from the reversibility of most physical processes except thermodynamics (the increase of entropy) over time. Just as we have long been sceptical about the idea of a view from nowhere, Huw Price (1996) argues there is no view from nowhen. Positionality or perspectivism is time anchored as well as being from a viewpoint. Having acknowledged the complexity and difficulty of developing a philosophy of time which can accommodate our asymmetric, unidirectional experience of time and the symmetries of physics, I shall leave these issues for the philosophers, and turn to our relationship to and planning of archives (but with these issues kept in mind).

Archives of the Future? Planning Future Archives

If we write on paper, if we dig foundations to our houses then we can have some confidence that we will leave traces that will be interpretable quite long into the future(s). The way much of contemporary life is organized is not this way—text messages, email and digital documents are, to say the very least, fragile. It is entirely possible that future historians will have less material to study than we have. One case in point might be the records of Cameroonian censuses. I know that the 1986 census returns were typed into computer (I saw the printouts) but I do not think the digital files have been archived. (I hope I am wrong). Another example, might be that of future photographic archives where the switch to digital photography means it is likely that there is (and will be) less legacy or record of the work of Cameroonian photographers from the 2000s than from the 1970s or 1980s (see Tatsitsa on Cameroon as well other articles in this issue: Le Febvre on an archive that was not created, and Collins’ discussion of the study of ghosts and their uneasy relationship to the history of science).

To use Barbara Adam's idea we should reflect on our timeprints (2008). The parallel is with the idea of “ecological footprint”: a timeprint is our impress on history and beyond. This points to a seriously large-scale temporal perspective as exemplified by archaeology, the discipline dealing with human timeprints across millennia, what Shryock and Smail (2011a) call “deep history”. Some provocative examples come from Christine Finn (2001) giving an archaeological perspective on Silicon Valley, from the work of the Long Now Foundation exploring ways of thinking about future pasts and deep, distant futures as well as the earlier work of Durrans (1992) on time capsules. As I wrote in 2012 as a riff on John Maynard Keynes, “in the long-term we are, at best, archaeology”.

From High Theory to Case Studies

As we turn to specifics it is worth reminding ourselves of why we look at controversies. There are several reasons. Most important I think is that they crystallize (or help clarify) features which always present but less visible in mundane, less controversial cases. They help us better understand routine, run of the mill, events where there is no controversy. The disputes make explicit the factors in play, even where no definite resolution is possible.

The articles in this special issue provide concrete substantiation in various ways. I have already discussed Dave Reason's theoretical contribution and Kate Moore's work showing archival film to Kwanyama people in Namibia. Le Febvre provides a case study of how current political issues frame the planning of an archive. These are long recurring themes about who has the right to speak, and to represent. In the highly charged, hotly contested, environment of the Negev it is little wonder that the proposed archive never came to fruition. But the processes she charts illustrate how a contested present has repercussions in its contested pasts.

Other examples may be found in Cameroonian history: an argument about the date of foundation of Banyo becomes a proxy for a dispute about the role of the Northern kingdoms (lamidates) in slavery (Mohammadou 1964; Hurault 1975). In more recent Cameroonian history, the Union des Populations du Cameroun/Bamiléké uprising and its suppression is still a topic sensitive to the administration some 40 years later. A reflex of this can be seen in the work of indigenous Cameroonian photographers (see Zeitlyn 2015; Tatsitsa this issue). The images reveal another side to the policing of the disturbances in West Cameroon: it made the maintenance of everyday life a real achievement. To baptize a baby, to hold a wedding, to bury the dead and to celebrate a death were more than just routine social events. The everyday photographs made by village photographers are testaments to activity and struggle to maintain continuity, tradition and domestic life. Even in happier and more peaceful times it is worth remembering the work undertaken by families out of the political limelight to maintain and continue human life. In short “the everyday” is a social achievement. These are aspects of pasts and futures that all too often historians and archaeologists overlook because the archives often reflect formal structures of power, and the mundane achievement of the everyday is neglected, taken for granted as soon as it is achieved in what Stoler calls a “Disposition of Disregard” (2009, 237). It is only when it fails that it becomes remarked upon (examples might include Sophocles' classic Greek drama of Antigone's struggle to bury her brother, and the exposure of “terrorist” bodies at crossroads or outside police stations, see Deltombe, Domergue, and Tatsitsa 2011).

Matt Hodges review essay discusses different takes on the idea of historicity. He uses Charles Stewarts’ exploration of dreaming on the Greek island of Naxos to raise questions about how the Western academic tradition can approach and comprehend nonhistoricist modes of knowing such as are found in dreams which are taken by Naxiots to include revelations from the past, predictions about the future and understandings of the present. Over the last almost two hundred years these Naxiot dreams have come to act as agents in the politics of how Greek islands relate to one another and the Greek national state. A history of concerns about human and non-human agency has its place in the history of anthropology; Tylor and his nineteenth century contemporaries argued about séances while the discipline of anthropology itself was being developed. As Collins argues (2015) these debates have a real and lively relevance in contemporary and future developments of the “internet of things” in which humans and non-human agencies interact and jointly produce the world we live in.

Conclusions

My conclusion is that things change as we liberalize and pluralize our notions of past, present and future. Recognizing that all are polyvalent, intercalated and multi-perspectival we need to adjust our planning of the archives and museums that accompany us. This can be summarized as an argument for a chronotype which is symmetrical across pasts and futures. A consequence of this is that we have to start thinking of the TRL extending forwards from the presents in a non-deterministic fashion. Perhaps no longer red but pink to signal its change of phase, and broadened into a band or braid: no longer the TRL but the Thick Reddish Braid. Whatever the metaphor, I suggest our thinking forwards should include a miasma of roads not to be taken which scenario planning can crystalize by concentrating on the more extreme examples. Possible but excluded futures cast shadows, exert influence or otherwise interfere with the actual future which is realized as mysteriously as the interference patterns of refraction in the two-slit experiment shown above. Within actuality is reference to the pasts from whence it has come and the futures it contains.

A more historiographical approach to this would arise as historians reflect about the sources they do not have but would like to have. What types of record might there have been but which seem not to have been created or to have survived? This is not to think of hypothetical pasts but hypothetical histories. Similar questions can be asked of anthropologists, provoking consideration of hypothetical, unwritten yet conceivable anthropologies. An example may be found in the planning document for the hypothetical Mambila Dictionary of Biography which started from the thought that if the British can have a Dictionary of National Biography then why should not the Mambila have one too? This raises a host of practical and conceptual problems about the criteria of inclusion for the people to be included in such a putative biographical dictionary project (see discussion and proposals at http://staff.anthro.ox.ac.uk/zeitlyn-david/planning-the-dictionary-of-mambila-biography/).

In contemporary Europe and North America part of our timeprint is a digital one maintained in large part by private corporations and as we learned in mid-2013 mined by security services. Leaving aside the politics and arguments about justifications for surveillance, I fear that this creates a terrible environment for establishing long-term archives. I have already mentioned the strong likelihood of relative few African digital photographs surviving from the early twenty-first century. As we plan archival structures in our presents for possible futures and reconceived pasts we are dealing, in so many ways, with fragile and delicate material. Material which must be cherished so some future audience can make it sing. Or, if not always sing, at least let it speak.

Acknowledgements

The workshops were administered by Penny Fraser and Natasha Samuels (in Yaoundé) and Nadine Levine (in Oxford). Other papers from these workshops have appeared in the inaugural issue of an online journal Vestiges: Traces of Record, http://vestiges-journal.info/. I am extremely grateful to the AHRC for supporting these activities and to the many other participants in the workshops who were not able to provide papers. Among others, these include Hamadou Adama, Ghislaine Glasson Deschaumes, Verkijika Fanso, Haidy Geismar, Alexander Kiossev, Perla Innocenti, Bren Neale, Susannah Radstone, and Leon Wainwright. Summaries of the workshops themselves are available from http://www.mambila.info/Futures/. Statement on access to the underlying research materials: this is a theoretical piece of work based on library research, see bibliography below.

The paper has been greatly improved by comments and suggestions made workshop participants, by the late Dave Reason and subsequently by the referees and the editors of History and Anthropology for all of which I owe much thanks.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This paper has arisen from a series of workshops run under the auspices of the Arts and Humanities Research Council (AHRC), “Care for the Future Programme” [grant number AH/K005170/1].

Notes

[1] One of the motivations for the discussions reported in Shryock and Smail (2011a) is the perceived need for a new set of metaphors to help comprehend “Deep history”.

[2] Surely the clearest evidence that time travel does not occur is that even in Europe (where we have written records covering (patchily) a couple of thousand years) there are no records of time travellers arriving from the distant future. Now of course absence of evidence is not evidence of absence and time travellers may have covered their tracks very cleverly indeed, but I think this is at the very least suggestive! For the more general point about not going backwards, see Hughes (1995, 2 citing Maurice Bloch). Relatedly Lewis argues (1987) that time travel is conceptually possible but does not occur in this possible world. See also Reason (this issue).

[3] She takes the protention/retention distinction from Husserl.

[4] Rosenberg and Harding (eds.) provide a survey of this terrain in their collection “Histories of the Future” (2005a).

[5] I understand chronotypes to be different types of Bakhtin's chronotope, a characterization of an attitude or orientation to time (Harro-Loit and Kõresaar 2010 citing Bender and Wellebery 1991, 4).

[6] In his major contribution to discussion of the philosophy of time McTaggart distinguished A-series of events in time with judgements about them changing as to being past/present/future (so an event's position will change relative to an observer with a specific position in time) from B-Series in which pairs of events are judged relative to another, one occurring before or after the other occurring. See below.

[7] Olivier (2001, 62) makes a similar point about houses and similar structures. Bryant's work in Cyprus provides a vivid example of how such histories trump dates (2014, 683).

[8] A move also made by Hirsch and Stewart in their introduction to a special issue of History and Anthropology (2005, 263).

[9] See Zeitlyn (2009) for a parallel argument about partiality and incompleteness, expanded in Zeitlyn and Just (2014).

[10] For another example which has repercussions for the idea of historicity see Matt Hodges review of Charles Stewart's discussion of the political power of dreams in Greek Island politics (this issue).

[11] See above for discussion of whether the definite article is appropriate here.

[12] We should note that recent work on classical divination casts some doubt on this. Beerden (2013, 22) gives an account of classical Greek and Roman views of the future as strikingly modern, “open but not empty” in which divination was seen as providing “advice” which may influence the choices that people made. She emphasizes that this was not universal: the Mesopotamian view of the future and divination was far more deterministic (220).

[13] Charles Golden in a review essay (2005) argues that the literature on memory is heterogeneous and lacks a single agreed set of conceptual underpinning. As he argues for memory I am arguing for time.

[14] Robert Textor calls this “tempocentrism” (2003, 524–525).

[15] Walter Benjamin reflecting on Proust talks of “convoluted time” (1968, 206).

[16] As mentioned above Heidegger illustrates this with the example of someone walking towards an acquaintance in the street. He asserts that, for the walker, the friend several metres away is “more proximate” than is the pavement underfoot (1978, 142).

[17] Braddock (2010) argues for the relevance of the psychoanalytic concept of countertransference for anthropology and I would suggest by extension to more general intersubjective knowledge of others. It might also be productive to apply the idea of countertransference to present relationships to futures and pasts.

[18] I am very grateful to Julia Binter for suggesting this connection.

[19] Or even more complicating “Moore's actions may have consequences eighty years earlier once the museum is created at some point in the future”.

[20] See also Walton (1990, 331) and subsequent discussion: Martin (1986) and Walton (1986) as well as Maynard (1997) and other references discussed in Pettersson (2004).

[21] He also points to how forecasts can be seen as conservative action—acting to make the future like the past!

[22] William Burton, “The use and abuse of history” quoted in Lowenthal (1985, 263).

[23] As was said above, in slightly different terms, McTaggart identifies the A-Series of absolute positions in time (colloquially dates) across which moves a shifting Now changing as it travels the dates from lying in the Future to being in the Past. He distinguishes this from the B-series in which pairs of events are classed as Before and After each other, and this does not change according to when the judgement is made, the position of Now in the A-series.

[24] An extreme case might be a geologist considering processes over millions of years but still feeling boredom during a phenomenologically interminable committee meeting, see Shryock and Smail (2011a).

[25] Serres uses superimposition (Knight 2012, 357) but I prefer the shorter superposition, partly because it is used in classical accounts of wave refraction and interference which may provide a different metaphorical base for these discussions.

[26] Mishler (2006) uses a similar metaphor discussing personal history narratives: he talks of a doubled arrow of time, in a narrative structured on sequence but started from the end like a detective story, written backwards since the narrator knows “what happens next”.