Archaeological Stratigraphy and the Bifurcation of Time: Solido intra solidum

Introduction: Steno’s Legacy

In 1669, the Danish scholar and cleric Niels Steensen, a.k.a. Nicolaus Steno, published a landmark volume for the earth sciences, dealing with stratigraphy and fossils: De Solido intra solidum naturaliter contento dissertationis – or, ‘Dissertation Concerning a Solid Body Enclosed by a Process of Nature Within a Solid’ (Steno, 1916). As the title implies, at the heart of the work was an attempt to explain an apparent paradox: how a solid body, such as a fossil, came to be encased within another solid, such as rock. This was no divine conjurer’s trick, although Steno’s solution to the enigma relied on a similar method of lateral thinking needed to unravel a magician’s act: what you think you see is not what is really there. For Steno, the answer lay in seeing one of the solids – the encasing rock – as a former fluid (e.g. see Simonetti and Ingold, 2018: 26–7).

There is nothing so startling about this once we realize it; after all, one can observe water turning into ice and flowing lava into solidified stone. Thus, arguing that all stratification and incorporation of ‘foreign bodies’ within rocks are indicators of a former fluid state is so obvious, one wonders why it seemed so bold in the 17th century. This is not the place to go into the changing cultural contexts of knowledge production and world views; rather I want to hold our focus on the relation between solids and fluids and, especially, the role time plays in this. In particular, I want to see what the legacy of Steno’s solution has had for the discipline of archaeology and for the way it conceptualizes stratigraphy and the archaeological record.

Archaeological Statics and Dynamics

Both the geological and the archaeological concepts of stratigraphy share a view of strata as a contemporary record of past events. The notion of ‘record’ to describe past remains was common throughout the earth sciences in the 19th and early 20th century, with often explicit analogies to documents (Bowker, 2006; Lucas, 2012). But as a record, it also entailed a particular ontological condition: a fossilized or preserved remnant from the flow of history. By this definition, for something to survive the flux of time or history, it had to be arrested, stopped in its tracks and remain inert, left behind as time marches on. In other words, for the past to have any presence in the present, it can only be as a remnant or vestige. However, the temporality of this residuality has been articulated differently over the history of the discipline (Lucas, 2015).

Up until the middle of the last century, this residuality was viewed largely in the context of missing pieces; Lyell’s famous metaphor of the fossil record as like an annual census, but with gaps for some years, captures this and was deployed to argue for gradualism over catastrophism (Bowker, 2006: 55). In archaeology the question of preservation has been a recurrent concern; many objects survive and are buried but many more decay, perish, or are recycled and re-used. Most are at least broken. Under this view, the temporal record is cinematographic – like a movie but with some frames or whole sections of the movie missing. Our job is to piece the bits together as best we can and to make reasoned inferences about what is missing (also see Simonetti, 2019, on cinematographic time in archaeology).

However, since the 1950s, scholars have begun to emphasize another dimension to the residuality of the archaeological and palaeontological records: their loss of resolution. It was not so much that some events are missing from the archive, but that they have been so irrevocably jumbled up that their temporal sequence has been lost. This is the notion of a record as information and of a fragmented record as one that has a scrambled message. Think of a wishing well with hundreds of coins – how can we know which went in first and which last? Some may rest on top of others, some may have more corrosion than others, but there is no way we can conceivably reconstruct the sequence in which all the coins were deposited. The issue here is that objects or fossils may enter a deposit at different times while the deposit is still forming or ‘fluid’; it is what archaeologists and geologists refer to as time-averaging. Here, the cinematographic metaphor breaks down; it is more like a film which has been subject to multiple exposures – or in the typical archival metaphor used by archaeologists, like those medieval manuscripts which have been erased and written over: the palimpsest (Bailey, 2007; Lucas, 2012).

Yet whether one uses the cinematographic or informatic/palimpsest metaphors to understand the temporality of the archaeological record, there remains an almost implicit acceptance of the distinction between past actions or events, which were dynamic and unfolding in time, and present traces of those events which, as they figure in the record, appear to be frozen and ‘stuck’ in time. The fact is that, in the past, objects like pots, stone tools and animals were all in motion in a living context, but in the archaeological record, they now lie inert and lifeless. Archaeologists in the 1970s talked about this in terms of archaeological statics and dynamics and devoted a lot of time to thinking about how to bridge the two – how to get from a static archaeological record to the dynamic processes that left us this record (Binford, 1983; Schiffer, 1972, 1976). But how does this relate to Steno’s fossil problem? In one case, you have a former fluid become solid, in the other, a former dynamic context turned static; there is certainly an analogy here, but is it purely metaphorical or is there something deeper? I will argue that there is more than meets the eye to this and it lies in the way time is predominantly thought through movement.

The Pompeii Premise

Michael Schiffer, who first coined the terms archaeological statics and dynamics, has, since the 1970s, played a pivotal role in developing our understanding of the formation of the archaeological record (Schiffer, 1987). In his discussion of the bridge between statics and dynamics, he argued for the importance of taking into account all intervening events and processes which can blur or alter the materiality of a past situation so that what we see in the archaeological record is always a distorted version the original. This includes not only preservation but also mixing; Schiffer was well aware that both the cinematographic and the informatic analogies had to be considered. However, Schiffer’s articulation of the issue encountered a conceptual problem that has never been satisfactorily dealt with (Schiffer, 1985). The problem was cast by Lewis Binford, another archaeologist who also played a key role in this field. He described it as the Pompeii Premise (Binford, 1982).

Put simply, Binford argued that in distinguishing between an original, dynamic context and intervening transformational processes, Schiffer was guilty of simplifying the temporality of the past. It was not simply that Schiffer wanted to get back to some specific moment in the past; rather that, according to Binford, the very idea of a specific moment in itself was a fiction. In some ways, calling this the Pompeii Premise is misleading and may make us miss the real point here. At one level, the idea of the Pompeii Premise simply implies viewing the archaeological record as a snapshot of the past – like a frozen moment in time, hence the eponymous use of Pompeii. This is the cinematographic notion of the archaeological record, and one which had been criticized by earlier archaeologists, as well as by Schiffer (Ascher, 1968: Schiffer, 1972; also see Olivier, 1999). But Binford’s point was more subtle; the problem lies not in treating the archaeological record or statics as a snapshot, but the past events or dynamics as if they were. Even if Schiffer did not treat the archaeological record as cinematographic, he was still guilty of treating historical reality as such. In short, in Binford’s critique, the Pompeii Premise applies not to the archaeological record but to past processes.

I am not so sure that even Binford quite realized the full implications of his critique because he was fixated on applying it to the notion of an original event. As Binford quite rightly asks: what is the original event in any archaeological context? Even such a simple example as a burial raises the question: what counts as the original event and what as intervening processes? Any presumption we might make is caught up in certain cultural expectations. For example, a large number of Norse graves in Iceland appear to have been ‘robbed’, but archaeologists have pondered over whether such robbing might in fact be integral to the burial rite and relate to secondary re-burial. Where do we cut the line when it comes to social behavior? Moreover, much social behavior is repetitive, so the very idea of an original act is somewhat irrelevant; it is often the very iterative nature of human action – what we call habit or custom – that enables a palimpsest to be meaningful as opposed to just ‘noise’ (Lucas, 2008).

All this is well and good, but at stake here is also the notion of discrete events, original or otherwise. In short, a cinematographic view of time underpins both the cinematographic and informatic or palimpsest analogues for the archaeological record. How, then, does all this tie back to Steno and his solid fluids? Let us, for now, set aside the issue of cinematographic time and return to more solid (or fluid) ground. In the following paragraphs, I will re-examine Steno’s original dichotomy of solids and fluids in relation to archaeological deposits, to see if this basic distinction actually holds up under scrutiny. This will open a clearer path into the problem of time and movement which, I suggest, is ultimately the key issue here.

Archaeological Deposits as Solid Fluids

Steno’s original solution to the problem of solids within solids was rich in its simplicity. And when it comes to geological stratigraphy, there is no denying the solidity of many fossiliferous rocks. But some geological strata, and certainly almost all archaeological strata, cannot be defined so unequivocally as solids. Moreover, for most archaeologists, the idea of deposits as solid or fluid is not routinely employed, yet in this section I want to explore what happens if we do consider them under these terms. Many years ago, I was excavating in a deep trench on an urban site in England; the trench extended several metres below the surface and I had to stand on a ladder to record the trench wall or section; we also had to put up shuttering and agro-props to stop the sides of the trench from caving in and burying me alive. Despite following all health and safety rules, it was still a little unnerving as I knew the safeguards might not be enough to protect me. Yet it helped that the deposits we were excavating were of thick, stiff alluvial clays – and that it was raining. Sandier deposits would have been much less stable and prone to collapse while, had the weather been hot, the clays would have dried out and started to crumble or shear off.

The point of this story should be obvious: archaeological deposits are not as solid as they may sometimes seem. Indeed, in some cases and under the right circumstances, they will actually ‘flow’. Often what holds deposits together is not their own solidity but the mutual support they provide for each other, such that more ‘solid’ layers will retain more ‘fluid’ ones. And as any archaeologist can tell you, most layers will differ from each other in their solidity or fluidity – that is one of the ways we can actually tell whether we are digging in one layer or another. But we don’t normally use the terms solidity and fluidity when recording such layers – rather we use other terms such as hardness or compaction, texture or granularity. Those layers I recorded in that deep trench would have been recorded in my notes as compact and clayey, where the term clay refers to the particle size of the deposit. There are even simple manual tests you can perform (and which I teach my students) to see if what you are excavating is clay, silt or sand or some combination thereof (see Figure 1). You take a piece of soil, wet it and then try and shape it into successively more plastic forms – the more plastic it is, the more clayey. OPEN IN VIEWER

Plasticity, granularity and compaction are all ways to characterize the solidity or fluidity of a deposit. Clearly, however, this is to understand solidity in a particular way: by its stability of form. It lies in the extent to which a deposit holds its shape, or holds together as a single, discrete entity. Seen from another perspective, this also expresses how much the deposit will ‘move’ – not just when you create a void next to it, as when digging a trench, but also in relation to adjacent deposits. But in this case, it also becomes clear that stability of form is also relative to the surrounding deposits; compact clays may well push against and into looser, sandier layers but meet their match in a masonry wall. Moreover, when two deposits of similar ‘solidity’ meet each other, there is also the possibility of blurring or blending. Another thing all archaeologists quickly learn is that while, in theory, deposits are supposed to be discrete, separate entities, in the field it can often be hard to distinguish them. Boundaries may be fuzzy; instead of a clear line marking where one deposit ends and another begins, there seems to be more of a continuum. I think it is useful to see blending as a kind of mutual permeability, where two objects meeting offer equal resistance and thus can merge. As common, though, is where one is more permeable than the other and, as a result, incorporates the other, unchanged.

The first time I came to Iceland, it was to excavate at a former monastic site. During the excavation, some joker thought it would be fun to play a trick on another digger; they planted the bones from their chicken drumstick, freshy eaten from lunch, into the top of the layer that another archaeologist was working on. Coming back after lunch, when they started scraping the soil, they were quite surprised to uncover these bones, although they quickly realized they had been planted, as their freshness gave them away. You might think that the disturbed ground must have also raised suspicion. Certainly, archaeologists can tell when a metal detectorist has surreptitiously been on site, leaving a trail of holes like molehills. But at the same time, it is very easy to ‘salt’ an archaeological site with fake finds, if you are so inclined; professionally it is deeply unethical of course, but that has never stopped some people. Why is it so easy? Because most deposits are not solids but solid-fluids; their boundaries are variously permeable, their matrix variously lax.

Indeed, a more common practice than seeding a site is to remove finds out of a layer, as in the case of metal detectorists. But unlike the detectorists with their molehills, archaeologists are generally better at covering their tracks. The normal procedure is only to remove a find once all the encasing soil has also been removed (though in some cases, archaeologists leave the find on a soil pedestal for recording purposes). But many is the time when excavation reveals the edge of a really nice object and, being only human, we get impatient and cannot wait. I have done this myself all too often. I gently insert my fingers into the soil and pluck out the artefact. If I feel it warrants recording in situ, I simply return it to its place – I even press down the soil again as if nothing had ever happened. And the deposit happily accepts it back.

But this permeability and laxness of deposits is not simply something encountered during excavation. It defines their very formation and mode of persistence. We see this all the time. Materials like water can flow across deposit boundaries, leaching minerals along the way. Sometimes the fuzzy boundaries I mentioned earlier are simply the product of leaching, leading to graded horizons in the soil profile; sometimes, when water meets more resistance, the minerals collect in a sharp, discrete band between layers. We call this iron panning. Worms and roots also move across boundaries and, when they do so, they too can disturb the soil and mix the finds. The list of post-depositional agents and processes which can cause archaeologists to scratch their heads all underline the permeability of archaeological strata.

These examples should be enough to demonstrate that archaeological strata or deposits are not solids; and in fact, even when they were being formed, they were not really fluids either. Rather they are always both solids and fluids – or solid fluids and, in the case of permeability, the term colloid might be more appropriate to define such strata, whereas it refers to substances suspended within other substances (as opposed to solutions where the two mix or blend). In short, Steno’s solution simply does not hold up for archaeology – and, I suspect, neither would it for much geology, but that is another matter. Maybe these are just metaphors (especially in the case of archaeology), but they are powerful ones which I believe have influenced the way archaeologists think about not only the archaeological record but also the past. To help develop this point, I want to draw out some general conclusions from this discussion of the solid-fluid nature of archaeological deposits.

Two salient qualities emerge from these reflections. One concerns the stability or fixity of a deposit; let us call this its durability, which carries connotations of both hardness and endurance. Indeed, archaeologists might use the term ‘indurated’ to refer to an extremely compacted or cemented deposit. A durable deposit is not only hard; it also retains its shape or form relative to other deposits over time. The other feature concerns the permeability or porosity of a deposit: its readiness to allow other materials – whether deposits or other objects – to flow through it. What both these qualities share in common is a notion of solidity/fluidity as defined by the extent to which they resist or facilitate movement. A durable, impervious deposit resists movement, of itself and of adjacent bodies, while a lax, porous deposit allows both. I am not suggesting that deposits should be classified into these two types, but rather that any deposit might be characterized as occupying a point along the two axes or dimensions of durability and porosity. More importantly for this discussion, however, we can view solid fluids in terms of their general ability to resist or facilitate movement and, in doing so, return to the question of time, raised earlier.

Archaeology, Movement and the Solid-Fluid Dichotomy

Re-casting the nature of stratigraphy in terms of movement brings us back to the old opposition of archaeological statics and dynamics. Past actions and events, which the term ‘archaeological dynamics’ refers to, were not literally fluid and nor have they been literally turned solid, except in rare cases like the bodies encased in ash at Pompeii. But characterizing solid-fluids as resisting or facilitating movement suddenly makes the analogy with statics and dynamics less metaphorical and more material. The past context in which pots and stone tools existed did often facilitate their movement in the course of being used in a range of activities and practices, while their subsequent deposition in the ground has tended to prevent further movement. Of course this is not a rigid distinction, but the point I wish to make is that the reason the analogy between Steno’s solid-fluid dichotomy and the archaeological dualism of statics and dynamics works is because both are based on a notion that thinks of time as movement. And because the static-dynamic opposition, like Steno’s solid-fluid one, is mapped onto the opposition between present and past, resistance or facilitation of movement becomes a way of characterizing the formation of the archaeological and geological record. On the other hand, thinking in terms of resistance or facilitation of movement implies a continuum, not an either/or, which is why the conjunction ‘solid-fluid’ is preferable to a hard and fast division between solid and fluid states. It also questions the rigid dichotomy of statics and dynamics as the present state of things in an archaeological context is not really static; in the soil, they can be subject to small movement and, when removed, they move a lot. But equally objects in the past may have lain still for long periods; think of clothes at the back of a cupboard, boxes in the attic or storeroom.

Yet when we explore time as movement, ironically a new solid-fluid dichotomy emerges. This will become clear if I give some examples of how archaeologists have portrayed movement – not of deposits so much as of people in the past. Here, I will focus on visual representations of population movements or migration as these capture best the dichotomy of solids and fluids. Let me start with a well-known archaeological image, Ammerman and Cavalli-Sforza’s 1971 map of the spread of farming through Europe during the Neolithic (see Figure 2). What it depicts is the movement of farming populations as a wave, represented here as isobars radiating out from a source point – like ripples in water. There is a clear sense here of movement as fluid, along with the assumption that the isobars represent arbitrary slices of time. Regardless of the accuracy of its chronology or even the arguments over whether farming spread through population movement or ideas – all of which remain hot topics, especially as the former has gained new adherents with the advent of the aDNA revolution (Reich, 2018) – the point is that movement is rendered as wave-like. OPEN IN VIEWER

In comparison, take a look at another image, by Anderson and Gillam (2000; see Figure 3), showing the spread of early humans through the Americas. Here, movement is characterized as a string-of-pearls and seems to mimic (although I believe it pre-dates) route guidance software such as that used by Google Maps. It represents movement as a series of discrete steps, like moves on a board game. Interestingly, Anderson and Gillam present two models in this paper: one is the string-of-pearls, the other is the leap-frog model which is very similar visually except that there are gaps between the ‘pearls’ or steps and arrows joining them. This is like jumping between places rather than walking, although of course what they intend to represent is rapid movement followed by periods of stasis. In either case, movement is rendered not as the deformation of a fluid, as in Ammerman and Cavalli-Sforza’s wave-of-advance model, but as more granular, more like the spatial displacement of a solid. OPEN IN VIEWER

One cannot resist the obvious analogy here to the wave-particle duality of quantum physics, but this also takes us right back to Zeno’s paradoxes of motion such as Achilles and the Tortoise or the Flying Arrow, as discussed by Aristotle in Book VI, Chapter 9 of his Physics. It pits motion as a continuum against motion as granular or saltatory. But not just motion of course. Time (and space) are also now inextricably entangled in our solid-fluid dichotomy. Consider the second hand on two different analogue clocks – on one it sweeps its way around the dial, on the other it hops in a jerky movement. Sometimes it flows, sometimes it jumps. This is not as straightforward as it appears however – a point I will return to shortly.

If we now fold this back into our earlier discussion of deposits as solid-fluids we realize we have been caught in a vicious circle. To escape the solid-fluid dichotomy of deposits as once fluid, now solid, we suggested that all deposits fluctuate between the axes of durability and porosity – both of which could be rendered in terms of the degree to which they resist or facilitate movement. But in redefining deposits and their contents in terms of movement, it seems we have just shifted the problem onto a different plane. Is this movement one of fluid deformation or the displacement of solids? Waves or billiard balls? Does the soil from a crumbling section of an excavation trench flow or roll down? How do soil particles or artefacts move? As solids or fluids?

Time to get back into our excavation trench. Now a student is troweling a surface, making it clean and level in preparation for a photograph. But wait a minute – what’s this? I can see distinct arcs of soil or shadow lines radiating out in front of where the student is kneeling, like ripples on a pond. This won’t do – I ask them to go back and remove these traces of their troweling as it will be too distracting on the photograph. And yet, these arcs somehow say something about the way soil moves across site; it is a calm ocean, the waves are slight but the act of trowelling creates a moving crest of soil like a bow wave, following the forward movement of the trowel. These crests are granular; the crumbs of soil tumble over each other and accumulate, growing bigger and bigger – until a shovel scoops them up and tips them into a bucket. The student picks up a full bucket and walks over to a mound of soil by the edge of the trench – the spoil heap – and tips the load onto the mound. Unlike the soil on site, this mound does not move; it settles down, slowly compacts and solidifies unless, or until, it is pushed back into the finished excavation trench or carted away.

Sometimes of course the soil resists the movement of the trowel – especially if it is very clayey or indurated. Then we have to hack or spade it out in clumps. But we don’t leave it like this – no, it is too solid. After all, it may, in homage to Steno, contain within it, other solids – flints, potsherds, coins. So we slice it up, break it down, de-solidify it, so it forms smaller lumps or crumbs – or we may even hose it down in a sieve, making it fluid (possibly once more) so it will yield its treasures, if any.

Ultimately, however, the way these deposits move in all these examples – whether they flow or roll – depends on the degree of granularity relative to the tool that moves them or the eye that watches them. It reminds me of the classic pitch-drop experiment created to demonstrate the fluidity of tar or pitch: put a lump of tar in a funnel and, if you wait long enough, it will drip – on average, one drop every 10 years (Johnston, 2013). This takes the proverbial ‘watching the washing dry’ to a whole new level. But such experiments are used to demonstrate the viscosity of fluids, which is another way of saying how a fluid can appear to be a solid. But conversely, we can point to the granularity of deposits to show how a solid can act like a fluid.

Let us return to those two clocks with different second-hand movements, one that sweeps, the other that hops. Scientists will tell us this is due to whether the mechanism on the second hand is moving above or below our sensory thresholds for distinguishing change (Wittmann, 2017). Doubtless it is, but as an analogy of time and motion, I initially thought such objections do not remove the deeper ontological problem it is meant to illustrate. But now I wonder. Maybe the distinction between the way solids and fluids move, whether like waves or jumps on a board game, is simply about temporal relativity: the speed at which matter or an object moves vis-à-vis the speed at which our senses operate. If the solution to the wave-particle duality comes down to the nature of the apparatus used to record the motion of sub-atomic matter (Barad, 2007), maybe it is the same with movement in general. Matter in itself is neither wave nor particle, it neither flows or jumps – it can do either, depending on the nature of the instrument observing or interacting with it.

Archaeology and Cinematographic Time

Let me conclude by linking this question of movement back to my earlier discussion of cinematographic and informatic metaphors for the archaeological record. Both of these metaphors, as you will recall, effectively portray time and change as founded on discrete events like the string-of-pearls model of movement. In the one case, the fragmentation of the archaeological record was due to missing pieces or segments along the chain; in the other it was due to the mixing or re-ordering of pieces so the original chain is all jumbled up. But in both cases, there is the presumption of history formed of discrete events, like the individual frames on a video (whether this is analogue or digital). In other words, for all intents and purposes we view history and time as something that is granular rather than wave-like. But this is not how we experience time in our everyday life, which is far more fluid. Of course, we can look back on our life and divide it into segments or moments, but as we live it, here and now, it is hard to avoid its fluid nature.

This disjunction between time as experienced and time as represented expresses itself in all kinds of ways; for historical disciplines like archaeology and history, it manifests itself most clearly in the way we divide history into chunks – that is, in periodization. Are periods arbitrary? Yes and no. It is beyond my present purpose to delve into this here (but see Lucas, 2019); suffice it to note that the very ambiguity surrounding periodization largely derives from this disjunction between our experience of time as a continuous flux and our representation of it as segmented or granular. Indeed, even when rendering movement as a wave, as in Figure 2, we still have to rely on isobars to portray movement. The question, though, is how we deal with such dislocation. Is time primordially fluid, such that granularity is a vulgarization, a reduction based on modern clock time, as Bergson and Heidegger argued (Bergson, 1913; Heidegger, 1962)? Or is this notion of flux itself simply a psychological illusion in a world where real time is granular, as in some theories of quantum time where the Planck scale determines the smallest unit of time: a hundred millionth of a trillionth of a trillionth of a trillionth of a second (Rovelli, 2018: 74). Below this threshold, time no longer exists.

Numbers like this are in some ways meaningless because they are so small; half a second, maybe, I can relate to my experience, but anything less and the distinctions carry little experiential weight. Whether it is a hundredth or a millionth of a second, it is all the same to me. But then the same is true the other way around. A hundred years – yes, I can just about relate to that; it is more than I am likely to live but it remains on a human scale. But a thousand years? A million years? Not likely – not without resorting to those well-worn metaphors of the clock: if the span of human history is represented as 24 hours on a clock, then my life represents only the last two to three seconds. But such analogies only underline our inability to relate the experience we have of time passing with such vast stretches of time. At such temporal scales, changes happen so slowly that nothing seems to move at all – just as hundredths of millionths of a second move so fast, it is as if they don’t move at all.

Perhaps this is the reason why the archaeological record appears so static, why our stories of the past appear so jerky and granulated. Cinematographic time in archaeology would seem to be ineluctable, not because time or history is intrinsically cinematographic, but because of the mismatch between the scale of human existence, of being-in-the-world, and the scale at which the archaeological record exists. The only way we can comprehend it is as granulated. The past is time solidified. Maybe Steno had a point after all. And yet maybe not. Perhaps there is a deeper lesson here, which is really about the problem of conceptualizing time through movement. All these issues – and many others, including Zeno’s famous paradoxes discussed by Aristotle – only arise because we think time through movement. Thinking about solids and fluids has only reinforced this association. Yet as other philosophers have argued, notably Deleuze in his work on cinema, there are other ways to imagine time that do not subordinate it to movement (Deleuze, 2013a, 2013b). For Deleuze, cinematographic time is quite different to how I have portrayed it here. But that is another story.