Abstract
Archaeology, like most sciences that rely on stratigraphic excavation for studying the past, tends to conceptualize this past as lying deep underneath the ground. Accordingly, chronologies tend to be depicted as a movement from bottom to top, which contrast with sciences that illustrate the passage of time horizontally. By paying attention to the development of the visual language of disciplines that follow stratigraphy, I show how chronologies get entangled with other temporalities, particularly those of writing. Relying on recent ethnographic work with archaeologists, the analysis reveals that excavation emerges as a double vertical movement of downward destruction and upward reconstruction that coincides with a systematic dissociation of time and space that has important effects for the understanding of the formation of sites. I conclude by looking at some of the implications of this dissociation for contemporary theoretical discussions, particularly those that emerged after the phenomenological push to horizontalize the discipline. Challenging this dissociation, I argue that the conceptualization of time in science should be understood as a process that depends on the body and unfolds in movement.
Introduction
Like most scientists interested in the past, archaeologists tend to face the past as they attempt to carry out their temporal journeys. However, compared with other scientists, archaeologists orient themselves downward rather than backward when facing the past. Instead of turning around they seem to lie forward and face the ground. Accordingly, and following the development of stratigraphic knowledge, archaeologists tend to conceptualize chronologies vertically with earlier events at the bottom and later events on top. This contrasts with the way other disciplines conceptualize chronologies. For example, in history, which has traditionally depended on reading, chronologies tend to be depicted, at least in Germanic and Latin languages, from left to right. Interestingly, when juxtaposed, the verticality of archaeology and the horizontality of history emerge in tension, depending on the context. For example, many archaeologists who collaborate with historians would be willing to state that the advantage of their work is that it manages to go deeper in time compared with history, as it reaches further down into the past.
But archaeology is not alone in the vertical understanding of time. In the history of sciences, other disciplines have understood time vertically. One key example is geology. In this article I trace the history of these conceptual relationships, by paying attention to key texts in the history of science, particularly to their emerging visual language. Visual representations are both useful and necessary means for understanding the way scientists think (Rudwick, 1976; Latour, 1986; Gould, 1991; Moser, 2009). The visual representation of time is of particular significance, given the systematic correspondence between spatial and temporal expressions and the limitations of written text for conveying spatial knowledge. The analysis shows that the verticality of time is far from being self-evident and uncontentious, as it emerges in tension with other temporal trajectories, particularly those of written texts. At the same time, relying on recent ethnographic work with land archaeologists in Scotland, I show that archaeology is caught in a double vertical movement of downward destruction and upward reconstruction that coincides with a progressive abstraction of the vertical temporal dimension from the horizontality of space. 1 By comparing different understandings of time described in the literature, I show that this separation is not universal but has resulted from a particular historical development within western science. Even though this separation of time and space has been criticized in archaeology, particularly by advocates of the landscape approach, I suggest that its influence is still visible in contemporary practices. Moving beyond this traditional separation, and the abstract understanding of time it provides, I argue that the historical analysis of time concepts in science requires an understanding of its practices, as both concepts and practice unfold in movement.
Up and down with vertical chronologies
A vertical understanding of the passage of time, consistent with stratigraphy, was originally conceived among geologists and only later adopted by archaeologists. Like archaeology, most earth sciences that look for things hiding underneath the surface of the earth share this vertical understanding of time. This is revealed both by the arrangement of their chronologies and by the similarity of some of their temporal expressions that talk about the experience of penetrating into something, presumably underneath, when moving into the past. Let me provide an example. In a recent, very successful, paleontological exhibition, called ‘Dans l’ombre des dinosaures’ [In the Shadow of Dinosaurs], held at the Grande Galerie de l’Évolution in Paris, there was a leaflet available for the general public with an introductory text that referred to the experience of time depth. The text written in French came with an English translation. The first phrase of the French text read: ‘Une plongée profonde dans le temps à l’époque où dinosaures et petits mammifères cohabitent sur terre’ (emphasis added). The English translation read: ‘A journey deep into the past, when dinosaurs lived on earth together with small mammals’ (emphasis added). Interestingly, in order to enter the exhibition, held on the lower floor, you have to go downstairs. On the way, you encounter a clock counting backward and some more text, again translated into English, on each side of the clock. On the left, you can read: ‘Notre histoire nous plonge dans des époques très lointaines …’ [This is an adventure in time travel …] and on the right: ‘… laissons-nous guider par le fil du temps’ [… we are about to embark on a journey to the ancient past]. After reading this last text, as you turn to the right, you encounter the stairway that takes you to the exhibition. As you descend you can see painted on the right-hand wall a time-scale that follows the steps of the stairway you are walking on, counting backward (or if you prefer, downward) from 0 to –85 million years. This is the year from when the exhibition starts its narrative about the history of dinosaurs that ends up with the famous Cretaceous-Tertiary extinction, one of the major turning points in the history of life on earth. 2 Once at the bottom of the stairs you can turn around and look upward to the numbers that count from –85 million years to the present. The scale looks like an ascending chronology in front of your eyes. As you turn around again to continue the journey, you encounter the image of a bird, one of the few surviving descendants of dinosaurs on earth, walking backward. Straight after the screen on your right the exhibition starts, –85 million years ago, at the time of dinosaurs.
The text displayed shows a particular understanding of time in palaeontology, which in this case is made accessible to the general public. As a discipline, palaeontology has been fundamental for the development of geology, and for the stratigraphic understanding of the earth. In the attempt to expand the time-scales afforded by biblical texts, the fossil record played a crucial role, providing a systematic means for temporally arranging different strata and correlating stratigraphical sequences across distant areas (see Rudwick, 1996; Toulmin and Goodfield, 1965). What the text reveals, along with the particular display of the exhibition, is the experience of moving downward into a past deep underneath the ground. As in archaeology, to travel in time is to move downward. Interestingly, several other temporal experiences get mixed up as we start walking. This very short journey of less than 3 minutes constitutes a huge conceptual challenge that seems to pass unnoticed by most visitors. As you move forward you start spinning counter-clockwise, downward in numbers and backward in steps, while still moving forward in time. And if you try to imagine all these experiences at once, you seem to go in many impossible directions.
In everyday life, many ways of conceptualizing time get mixed together. These conceptualizations can be consistent or in tension with each other and with our daily actions, which also unfold through time. Such tensions depend, among other things, on the direction and trajectory of the movement. There is one potential tension within vertical chronologies related to the way texts move both in Germanic and in Latin languages. Most chronologies are supposed to express a particular conceptual understanding of the passage of time for their readers. In the case of the sciences we are analysing here, such understanding assumes that time moves from bottom to top. Traditionally, these depictions are inserted in texts that tend to unfold in the opposite direction, from left to right, and from top to bottom. This tendency might not always be evident beyond the particular directions that reading and writing take. However, this might become visible when temporal sequences are introduced into a text. For example, when we count or assign letters of the alphabet to events in time we expect earlier elements on the alphabetic and the numerical sequence to be closer to the top left-hand corner and later elements in the sequence closer to the bottom right-hand corner of the page.
If these tensions are at all problematic, then it should be possible to trace them in the history of vertical chronologies up to the present. Nicolas Steno, in one of the earliest geological works, published in 1669, had to face this tension when coming up with his chronological representation of the geological history of Tuscany. Originally interested in anatomy, he was diverted from his research interests after an unusual find came from the sea. A gigantic shark was caught in the Mediterranean off Livorno and the grand duke of Tuscany, Ferdinand II, ordered that the head be sent to Steno for examination. The teeth of the shark looked very similar to some fossils that at the time were believed to grow inside rocks. These were known as glossopetrae (tong stones). What Steno concluded, contrary to the popular belief of the time, was that the tong stones were not produced inside the rock but rather came from the sea. The layers in which these fossils were found were originally covered in water. Even though earth history had to wait several years for its time-scale to be expanded at the hands of authors like Hutton (1795), Steno seems to have formulated for the first time the principle of superposition, fundamental for the stratigraphic understanding of geological time (Steno, 1916: 229–30).
This vertical understanding of how strata are deposited over time is reflected in a series of drawings of vertical sections that summarized the geological history of Tuscany, which Steno attempted to extend to the entire earth. This history followed an ‘agreement of Nature with Scripture’ (Steno, 1916: 263). Embedded in a non-secular scientific practice, Steno’s conclusion about the origin of marine fossils helped him to find support for the biblical understanding of the earth’s history, particularly the presence of a flood. Six images, numbered from 20 to 25, were used, each representing a particular stage in the history of the earth (see figure 1). Number 25 shows land originally covered in water. Number 24 shows how the lower layers were eroded by the action of underground forces while the upper layer remained in place. Number 23 shows the collapse of the upper layer in the absence of the lower ones. Number 22 shows when Noah’s flood again covered the land and new strata were deposited on the valleys by the movements of the sea. Number 21 shows again how the lower layers were eroded, while the upper layer remained in place. And number 20 shows how valleys looked at the time when the upper layer collapsed. As stated in the Bible, the earth was covered twice in water, first at the beginning and then at the time of Noah’s flood.
History of Tuscany (Steno, 1916).
This sequence of figures was arranged vertically with earlier events at the bottom and later events on top. This is fully congruent with the stratigraphic understanding of the passage of time that Steno was envisioning. The drawings, which represent abstract vertical sections, contain perfectly cut rectangular blocks of strata piled one on top of the other, aligned with the horizon. Inserted within a book, the temporal arrangement of the images seems to run contrary to the temporal unfolding of the text. This becomes more evident when the numbering of the figures is considered from the perspective of its temporal dimension. The chronology and the numbering run in opposite directions. Earlier numbers were assigned to later layers and later numbers to earlier ones. Interestingly, when Steno began to narrate the history of Tuscany, he felt he had to start the other way around ‘from the last to the first’ (Steno, 1916: 262). This makes things particularly awkward as events are narrated from end to beginning. And later on, in the explanation of the figures at the end of the text, he decided to start the other way around. Next to the plate, which listed the figures from 20 to 25, the explanation numbered them this time from 25 to 20, from beginning to end (ibid.: 275–6).
Gould (1987: 51–9), who analysed this series of drawings in trying to understand Steno’s conceptualization of time, mentions that Steno’s narrative of the history of Tuscany started not at the beginning but at the end. However, Gould did not consider this as something worth analysing. As in most academic works on the conceptualization of time, he starts with an abstract distinction that fell short in describing the particularity of Steno’s new form of time conceptualization. Gould was trying to see whether Steno fell into the directional understanding of time’s arrow or the circular understanding of time’s cycle. Even though this distinction, like other abstract philosophical dichotomies, sheds light on the tension between human experience and the conceptualization of time in general, it does not address the multiple trajectories in which time unfolds. 3 In trying to suggest that Steno was closer to time’s cycle, Gould pointed to the different arrangement that Steno’s figures had in the original Latin text. Instead of being sorted all in a single column they were divided in two (from 20 to 22 on the left and from 23 to 25 on the right). For Gould this particular ordering did not represent an economical use of the page but rather revealed Steno’s cyclical understanding of time, in which the first three stages matched the second three. The images of the earth covered in water (nos. 22–5), the images of lower strata eroded by the underground forces (nos. 21–4) and the images of the collapsed upper layers (nos. 20–3) paralleled each other.
Here Gould’s cyclical analysis of Steno’s figures does not pay attention to their particular temporal arrangement. Even though Gould’s analysis is about the ‘discovery of deep time’ (1987: 1), he never analysed the expression deep time in itself. By focusing on the contrast between time’s arrow and time’s cycle, he regarded this concept, systematically used both in geology and archaeology, as given. He never pointed to the connection between a depth under the ground and the past, or to the vertical understanding of the passage of time that results from it. As a result, his omission of the emerging tensions in the development of the visual language of geology is not surprising. Both in Winter’s translation and in the original Latin text the vertical arrangement of geological time runs counter to the temporality of the texts and the way numbers were displayed. This tension gives at least some clues on why Steno decided to narrate the geological history of Tuscany first from end to beginning and only then from beginning to end.
Although it would be very interesting to compare the way Hutton used visual representations to depict time, he did not produce any chronological illustrations in the first two volumes of his Theory of the Earth, published in 1795. Only the third volume, which appeared in 1899, more than 100 years after the previous two, contained chronological images. These were not originally used by Hutton in his manuscript but were arranged years later by Jermyn Street, who followed the already established visual language of geology. The century between the publication of the first volume and the third saw the fast and explosive establishment of this new visual language (Rudwick, 1976). Its emergence followed important technological developments in book production that allowed the use of images at lower costs and within the text. Such developments started to become available by the time Lyell was writing his Principles of Geology, and he seems to have explored their potentials more systematically in his third volume. This contained considerably more visual representations than in the first one (Rudwick, 1990: xiv).
In this exploration, Lyell had to face tensions similar to Steno’s. One example is contained in his third volume just before he analyses tertiary formations. In his text, Lyell paid careful attention to the geological processes of this late period of the earth’s history, which allowed him to show, consistently with his uniformitarian approach, that contemporary causes were sufficient to explain changes in the past. This constituted a movement against catastrophism and the idea of dramatic variation in geological processes. Lyell started by describing how strata tend to be arranged vertically, with earlier sediments at the bottom and later ones on top. In doing so, he provided an image (see figure 2) in which strata are listed from top to bottom using letters of the alphabet. On the next page, while introducing his discussion of tertiary formations, he provided a similar image but with one important difference: the letters were inversely arranged (see figure 3). Interestingly, in the space of one page, first the latest and then the earliest listed letters of the alphabet were assigned to the earliest stratum. Compared with the first image, the arrangement of letters in the second is congruent with the temporal direction of the terminology now introduced, which was listed below the image. Here the letter ‘a’ corresponds to primary rocks.
Order of succession of stratified masses (Lyell, 1991: 15).
Primary to tertiary formations of the Paris Basin (Lyell, 1991: 16).
It is possible to argue that in the second image, Lyell was following the left-to-right arrangement of the text. In any case, the tension seems to be playing a role, which is visible in many other images contained within the third volume that use both numbers and letters in situations in which formations are not inclined from left to right. It is worth noting that Lyell, like Steno, sometimes narrated events backward (or downward) from end to beginning. Interestingly, Lyell attributed this to an empirical necessity to go ‘from the known to the unknown’ (in Rudwick, 1990: xliii), which is congruent with his uniformitarian understanding of the earth. I will come back to this issue later on. What is relevant to point out now is that contrary to Rudwick’s (1976) analysis, even though the visual language of geology was well established by the time Lyell published his Principles, it was not at all unproblematic. Some conventions were changing, even within a single text, as a result of the tensions introduced by the new vertical time of geology.
Another form of visual representation in which this tension can be traced is in the production of geological maps. A few years before the publication of Lyell’s work, geologists both in France and England were attempting to develop the first examples. Again they were forced to solve the tensions imposed by these new vertical chronologies. Two are of particular relevance in the history of geology. The first was Georges Cuvier’s and Alexander Brongniart’s carte géognostique, available probably since 1808, which described the geographical distribution of minerals in the surroundings of Paris. The second was William Smith’s (1815a) Geological map of England and Wales, known as the first geological map in the United Kingdom. These maps, like any other, embodied a particular understanding of space. As David Turnbull (2000) has shown, cartography has never followed a fully standardized procedure. Even though current technologies push us to believe that mapping contains a universal perspective on space, it has never been the result of an objective abstraction and a one-to-one correspondence with the world. Maps are always the result of a particular perspective on space, as we have never been able see the world from everywhere at once (Ingold, 2000: 226). These two maps were no exception. Developed in parallel, with some apparent though unacknowledged exchange between their authors, they resulted from an effort to correlate systematically and for the first time the arrangement of strata with the fossil record (Rudwick, 1996: 277). For example, Smith, who spent several years working on his map, relied on a huge number of fossils that he collected from all over the country (Eyles and Eyles, 1938). This collection, which he kept at home, was arranged vertically in shelves that followed the stratigraphic arrangement in which they were found (Eyles, 1967: 180). And by the time he decided to sell his collection to the British Museum, for financial reasons, Smith assigned two codes to each fossil, one for its geographical location and one for the particular stratum it came from: in other words, one code for its horizontal and one for its vertical alignment. This collection allowed Smith to correlate strata from different regions, which was fundamental to the construction of his geological map. Cuvier and Brongniart likewise attempted to systematically correlate fossils and strata. By the time their map was produced, Cuvier had worked extensively with fossils in the area.
These two maps constituted an attempt to depict in an already established two-dimensional cartographic tradition a three-dimensional understanding of space that summarized the stratigraphic knowledge gathered at the time. Shades, colours, sections, symbolic keys, among other visual resources, were used to depict this three-dimensional understanding of the earth. Here the third vertical dimension is added on top of the originally two-dimensional map, with the result that this third dimension is visualized as separated from the other two. Smith’s Geological section from London to Snowdon that was meant to be read in conjunction with his map is an example (see figure 4). This section attempts to be a realistic abstraction of a real landscape. However, by following straight lines, for accurate measurement, the section integrates only one of the horizontal dimensions with the vertical now introduced.
Extract from Smith’s Geological Section from London to Snowdon (1817).
Returning to the tensions between chronologies and text, Cuvier’s and Brongniart’s attempts to integrate the trajectory of this third dimension differed considerably from Smith’s sections. Contrary to current convention, the key in the first version of Cuvier’s and Brongniart’s map (1810, in Cuvier, 1811) was inverted with respect to their spatial order of superposition. This is probably, as Rudwick (1976) suggests, because the authors described the formations in the text in their temporal order of deposition. However, other influences were probably playing a role. For example, Cuvier and Brongniart (1997: 147) mentioned how miners referred to the upper masses of gypsums as ‘the first’ rather than the other way around. Contrary to miners, Cuvier and Brongniart were among the first to invert this chronological order to match deposition rather than experience. Quite remarkably, following the development of their ideas, the order of the key was inverted in the second edition of the map, published 12 years later (Cuvier and Brongniart, 1822, in Cuvier, 1822). Figure 5 contrasts their order.
Keys and section to the Carte Géognostique des Environs de Paris (Cuvier and Brongniart, 1810 and 1822; see Cuvier and Brongniart, 1997).
In the development of these maps, as in geographical ones, there was a progressive process of abstraction that was meant to serve the goals of the state. In the case of Smith this contributed to perpetuate another verticality, namely ‘the greatness of England in the scale of nations’ (1815b: 10). At least that was the emphasis Smith gave it as he struggled financially to gain a reward for the collection he was selling to the British Museum. This process of abstraction was especially visible in the exaggerated regularity in which the vertical arrangement of soils was depicted using colours. ‘The boundaries between them are drawn boldly with ruled lines, in a style reflecting Smith’s work as a civil engineer, although this entailed major extrapolation from the evidence observable at the surface in outcrops and quarries’ (Rudwick, 1996: 276). According to Smith:
The Society of Arts very wisely foresaw in offering their premium for a mineralogical map (which I have just had the honour to obtain) that one of the greatest difficulties in understanding such an extensive branch of natural history arose from the want of some method of generalizing the information, which could only be supplied by a map that gives, in one view, the locality of thousands of specimens. By strong lines of colour the principal range of strata are rendered conspicuous and naturally formed into classes, which may be seen and understood at a distance from the map, without distressing the eye to search for small characters. This is the advantage of colours over any other mode of representation. (1815b: 11; emphases added)
Verticality and horizontality in archaeology
This brief analysis of some key examples in the development of vertical chronologies shows the particular temporal tensions in which the visual language of geology emerged. Even though these tensions might pass unperceived to the expert’s eye, they are still present in many sciences that rely on vertical chronologies. One interesting example comes from archaeology, namely the well-established Harris’s Matrix (Harris, 1975). Developed in the 1970s for depicting the temporal arrangement of strata, this tool, once ‘revolutionary’ in archaeology, constitutes nowadays part of ‘a new archaeological orthodoxy’ (Chadwick, 1998: 5). In it, soils are abstracted into small rectangles, which are then connected through lines. Three forms of relationship can be depicted. Soils can be above (later), below (earlier), or at the same level (contemporary) with respect to others. The whole set of relationships can be drawn on a single network that gives an idea of how strata are arranged. Usually, strata are assigned numbers that can be referred to a detailed description, which is filled out each time archaeologists encounter a new soil. These descriptions, known as contexts, are entered on special record sheets. Each of them specifies not only the relative position of the soil but also its composition, colour and consistency, among other things. This information allows archaeologists to identify soils and correlate them as they proceed with the excavation. Interestingly, in a Harris Matrix these numbers tend to flow once more the other way around with respect to the usual temporal arrangements of strata. Figure 6 shows a simplified version.
Harris (1975: 115), reproduced with permission.
The reason for this arrangement is very practical and is related to the way archaeologists excavate. As in geology, when it comes to understanding the arrangement of strata, archaeologists need to go from the known to the unknown. 4 Archaeologists have to start with the first soil that is accessible, which is always the one at the surface, and not with those hiding underneath. Otherwise, they would disturb the entire arrangement of strata. This is partly why archaeology is considered an ‘unrepeatable experiment’ (Barker, 1999: 13). They cannot remove a soil and then put it back as it was. They cannot trace the original temporal arrangement without disturbing it. Ultimately, this vertical arrangement is never fully visible, as archaeologists need to start from somewhere. Like their own experience of time, the material traces of the past they search for become irreversible as they move forward (downward), this time not to the future but into the past. On their way, the information collected becomes crucial. The archaeological record is what allows this temporal arrangement to be retrospectively reconstructed in the midst of a rapidly unfolding perceptual process. And when it comes to assigning numbers to the soils, archaeologists also have to move downward, starting at the surface. They cannot start with the last in the sequence, for that would presuppose that they could already anticipate the number of contexts they will encounter on a site. If they tried to do so, they would probably run out of contexts or have some to spare at the end of the season. 5
When paying attention to chronological representations like the Harris Matrix from the perspective of archaeological practice, something fundamental is revealed about the way archaeology is carried out. For sciences that look for their object of study underneath the surface the tension we have described above is also constitutive of their practice. Archaeology needs to start with the latest layer. But only later on, after the job is done and all the layers have been revealed, are archaeologists able to tell the story of a site upward from beginning to end. In so doing, they find themselves counting backward as the fathers of geology did. Accordingly, archaeology is constituted as a double movement. As stated in a site manual for an excavation in which I participated, ‘context sheets are the building blocks of the archaeological written record’. The paradox here is that these building blocks have first to be excavated, one by one. Archaeology always involves downward destruction and upward reconstruction, the latter being again paradoxically embedded in the downward, and left to right, movements of a text.
Even though these tensions might pass unperceived as a result of convention, presumably students becoming familiar for the first time with the basic assumptions of the discipline have to struggle with them, especially considering that usually the starting points for students are texts that assume ‘time depth’ as something given, as occurs most of the time within academia. Gould, for example, confesses the difficulty in explaining to novices the meaning of ‘deep time’ (1987: 3). Possibly, nowadays, this is less visible as some of these topics are introduced progressively earlier in school. However, these tensions will probably never disappear, as long as vertical chronologies are embedded in texts and as archaeology has to move backward (downward) to be able to describe events forward (upward) in time. These tensions would just need to be traced earlier in development.
This double vertical movement coincides with a separation between time and space that is constantly present in the practice of excavation. As we saw above, this can be traced back to the development of mapping in geology. Recording finds in an excavation depends on locating them both horizontally in space and vertically in time. Before starting, a two-dimensional square grid is set up that divides the horizontal plane at regular intervals by using parallel and perpendicular straight lines that cross at right angles. Each crossing point is marked with a wooden stake, which is referred to an ‘absolute’ grid of northings and eastings. At the same time, each square of the grid is named using letters and numbers, and finds refer to these codes. Trays and bags are always labelled with the specific area from which the finds come.
The vertical dimension is recorded differently. As the excavation starts and soils are uncovered, archaeologists have to describe each new layer, paying attention to its vertical arrangement, resulting in a network of relationships between contexts. Bags and trays of finds are also labelled with reference to these contexts. As in Smith’s collection of fossils these two codes, the area on a grid and the context number, allow things to be located three-dimensionally. However, the context number, by contrast to the grid’s coordinates, specifies the temporal location of a find. Grids specify only a non-temporal location in a two-dimensional space. 6
This separation between a horizontal atemporal space and a vertical temporality is also visible in archaeological drawing, still one of the fundamental representational devices of archaeology’s visual vocabulary. As in the early development of geological mapping, plans measure a horizontal two-dimensional arrangement of space while sections measure not only space but time by depicting the vertical arrangement of soils, features and the finds contained in them. By seeking geometrical accuracy they remain separate forms of representation, illusorily depicting an image from nowhere that nonetheless hides a particular point of view. In this the horizontal dimension becomes timeless. Spatiality and temporality, the vertical and the horizontal, seem to coexist in separate types of drawings.
Sometimes this time–space division is visible in how archaeologists think about the trajectories of their finds in the formation of sites. Once, during a large excavation of a medieval site in Scotland, the director told me something revealing about some finds they had encountered in the area we were excavating. This site, inhabited for centuries, was disturbed several times by human action through the excavation, demolition and rebuilding of structures. As a result, at the beginning archaeologists encountered unexpected things that made interpretation difficult, like the presence of medieval pottery in more contemporary layers. According to the director, ‘we found all sort of strange things in terms of dates. The good thing is that things have not moved across places. Finds have travelled up and down quite a lot but luckily not across places [gesturing a horizontal line with the right hand].’ What this remark suggests is that finds move in space when they change their horizontal position and in time when they move vertically. From a horizontal perspective movement is seen as occurring out of time. From a vertical perspective movement seems to be devoid of spatiality. It is like travelling in a time machine that moves a long way forward but ends up arriving at the exact same spot (see figure 7).
Trajectories of materials in Euclidean space: (A) movement in space; (B) movement in time.
It is worth noting that the separation of time and space is only sustained as a result of a process of abstraction. In the case of the grid, finds are marked as belonging to an area by assigning codes to it but, in most cases, their specific location is erased. A similar thing happens with the vertical dimension. The collection of contexts constitutes a web of emerging semiotic relationships. Each time a new context is found the entire web adjusts slightly to accommodate it. Notably it seems to be possible to relate contexts only by simplifying them enough to eliminate their irregularities and their fuzzy boundaries. Soils are abstracted into perfect rectangular blocks, like the ones Steno drew long ago when explaining the history of Tuscany. Together, these coexisting horizontal and vertical semiotic systems allow things to be located in two parallel dimensions, horizontal space and vertical time. They allow past trajectories of finds to be imagined as occurring either outside time or outside space.
Conclusion: A critique of the division between time and space in archaeology
The analysis presented here has allowed us to trace the verticalization of time and its systematic separation from space in archaeology and other related sciences. From the perspective of the conceptualization of space this division is consistent with a widespread understanding of the landscape as a horizontal surface. As Ingold (2007: 25) has recently shown, there is a widespread conception of the earth in science as a flat stage-set furnished with objects. This conception tends to leave outside all that is above and around the experience of an organism in relation to the environment, namely the sky, wind and weather. As we mentioned above, this understanding of the landscape is based on a cartographic view of space from everywhere.
Such a division of time and space persists not only in the ways archaeologists appropriate space but also in the ways their temporal journeys are constituted. The movements of the discipline seem caught in a quest for the oldest, a tendency to look for the initial ground where everything started; the baseboard on which the rest of history stands. The general search for foundational myths in archaeology, or origin stories as Lucas (2005: 54) calls them in his recent analysis of time in archaeology, suggests this. These tend to perpetuate our own view on history. Obviously, this is not exclusive of archaeology. Many sciences of time take similar approaches as they attempt to locate that first present from where to tell the story of the universe, earth, life and the multiple cognitive capacities that make humanity ‘unique’ compared with other species. After moving downward from the present surface, trying to excavate that baptismal ground from where everything else started, archaeologists have to turn around looking upward, as they attempt to narrate the sequence of events. 7 Excavation moves downward, removing surfaces, while reconstruction moves upward, imagining them. 8
The assumption that time can be visualized as a series of vertically aligned surfaces encountered from everywhere, is not only far from universal but becomes problematic as it contrasts with other understandings of time described in the literature. Take, for example, the remarkable work of W. E. H. Stanner (1965) on the Australian Aboriginal concept of the Dreaming, a mythical time or epoch that describes the life of the ancestors. Stanner (1965: 159) described this particular understanding of time as everywhen. Even though concepts like time and history were foreign to Australian Aboriginal people, this idea seems in tension with the view of life as anchored in a distant past and the surface of the earth as a flat stage-set, on which the present is lived. Australian Aboriginal people experience life as being on the inside of an eternal moment of creation in which past, present and future are rolled into one another. This is seen in the lack of tragedy and sense of abidingness that characterizes their life, which contrasts with our western quest to understand the past and to predict the future. Their history does not lie underneath in a vertical succession of flat events but is rather all around them. 9
It is important to mention that I am not suggesting that certain ways of conceptualization are better than the others. Nor am I denying the utility of the verticalization and horizontalization of time and space, which have helped archaeology in its understanding of the past. To the contrary, I am just suggesting that we need to be aware that the way time and space are conceptualized in archaeology is just one among many, which has particular epistemological and ontological consequences. The separation of time and space in archaeology emerges only in the attempt to measure time using the straight linear dimensions of Euclidean space. Following Damm (2001: 38), it is necessary to realize the risks of trying to fulfil the western need of having an official scientific pre-history, which, according to Fabian (1983), almost certainly involves denying others having a voice.
A recent example will serve to illustrate the point. Robinson (2010), in his analysis of how indigenous communities of south-central California understand their history regarding Chumash rock paintings, treats the idea of a deep vertical past as given for the natives he characterizes. Even though he mentions the differences between archaeological and indigenous conceptions of time, he suggests more than once the idea of a ‘deep indigenous past’ (2010: 84). Unsurprisingly, this is translated into a diagram that mirrors Robinson’s archaeological understanding of time, in which the native’s conceptualization of history is drawn as a vertical movement from bottom to top with several critical events. These include, among others, a division between a mythic and a historical phase, with the latter starting after the colonization process (see figure 8). 10
A projection of archaeological time onto South-Central California natives (Robinson, 2010: 102), reproduced with permission.
It is worth noting that in the last decades, after Tilley’s (1994) seminal work, the cartographic view of space from above and its separation from time have been challenged in archaeology by an emphasis on the phenomenological encounter with the landscape. Interestingly, even though this approach has led to concepts such as landscape being used much more widely, this does not necessarily represent a unified movement (see David and Thomas, 2008). The separation of time and space seems still very much alive today, especially now that mapping technologies are becoming ever more popular. Many are still caught in the tension between a cartographic view of space and a vertical understanding of time.
However, if we look carefully, it is possible to see that the phenomenological critique has introduced another tension within the discipline. In its attempt to go beyond the narrow focus on excavation and in its emphasis on vertical chronological thinking, the phenomenological approach has emphasized a particular understanding of place very much informed by the experience of walking (see, for example, Thomas, 2008: 305). This follows an emphasis on the horizontality in which survey techniques are conducted, which from the perspective of practitioners has emerged in tension with the classical downward movements of excavation (see, for example, Bender, Hamilton and Tilley, 1997: 150; see also Bradley, 2003). Even though space is no longer seen from above, the tendency has been to conceive it from its horizontal encounter. This is consistent with a widespread understanding of time in phenomenology as a horizontal movement, which is again more compatible with walking than excavation (see, for example, Ingold, 2000: 189–208). Just to give an example, Heidegger, right at the beginning of Being and Time, defines ‘time as the possible horizon for any understanding’ (1962: 1). While in traditional archaeology time and space are divided into vertical and horizontal trajectories, for the landscape approach these coincide with the horizon. This results in a tension between those who emphasize a deep past and those who concentrate on present experience. The current discussions between the landscape approach and time perspectivism, the second of which stresses the unique time depth of archaeology, exemplifies the tension (see, for example, Murray, 1999: 2; see also Shanks and Tilley, 1987; Thomas, 1996; Lucas, 2005; Bailey, 2007; Holdaway and Wandsnider, 2008). Interestingly, from this new horizontalization of the discipline new risks emerge, particularly the potential denial of all those who move up and down, like miners, divers, climbers, among others, including the scientists who excavate their knowledge. They also encounter the environment vertically, which matches, like those who wish to move horizontally, their conceptual understandings of time.
This last point is crucial. Concepts of time are not just abstract entities but constitute corporeal orientations that follow the historical development of particular forms of practice. Relying on stratigraphic knowledge, experts spend long periods of time paying attention to the ground. In doing so, they excavate using tools, such as the trowel, which again share this downward orientation. Kneeling against the ground, the trowel, the hand, the eyes and the rest of the body engage in a sensory exploration of the properties of soils in movement. As we have shown here, this point of view coincides with the way archaeologists, and other scientists that excavate the past, refer to their experience of time, as in the idea of ‘going deeper in time’. Similarly, but set in opposition, the orientation provided by the new emphasis on survey and the experience of walking coincides with an understanding of time that follows the horizon. In both cases, the way archaeologists understand the past depends on how they conduct their practice, while their practice is never independent of their understanding of time. Both are continuous and co-constitutive (Simonetti, 2012).
This has serious implications for the historical analysis of knowledge in general, and for the analysis of time concepts in particular. This work has attempted to rewrite a chapter in the history of time in science by highlighting and abstracting key authors and events in a sequence that has moved once more from top to bottom. Once anthropologists like me return from the field, they also engage in the process of putting their experiences on paper, looking back at their notes and the history of those they want to describe. But by following closely the relationship between experience and conceptualization, informed by the practice of scientists in the field including my own, this article has attempted to close the gap that has traditionally existed in the West between time and space. Ultimately, understanding the history of time depends on the body as both concepts and practice unfold in movement.