|
Post by Admin on Jun 29, 2022 17:44:22 GMT
Lithic artefacts were recovered from both excavated trenches. Many had experienced taphonomic damage due to fluvial activities, being rolled and/or fragmented, while some retained very fresh edges indicative of little post-depositional modification. Using the techniques outlined in §3.2, a total of 251 artefacts were identified, comprising roughly 20% of the collected lithic material from the excavation. Artefacts were found throughout the excavated gravel, although two clear concentrations were present (table 3). The uppermost 50 cm of gravel (the upper main gravel) contained 113 (45.0%) artefacts, with the greatest concentration being 40–50 cm from the upper limit of the gravel (n = 41, 16.3%). The intermediate large flint layer at a depth of 50–80 cm contained fewer artefacts (n = 26, 10.4%). It is notable that Trench B had no lithic artefacts between 60 and 80 cm. The lower main gravel (incl. sand layer) contained 112 artefacts (44.6), with a concentration located 110–120 cm from the gravel's upper limit (n = 31, 12.4%). Table 3. Absolute and relative frequencies of artefact concentrations throughout the excavated gravel sequence. Data are from both trenches combined, as measured from the surface of the main gravel mass. Note that the associated sediment layers are here for reference and do not perfectly align with the depth increments. depth (cm) no. artefacts % associated layer 0–10 22 8.76 upper main gravel 10–20 10 3.98 20–30 21 8.37 30–40 19 7.57 40–50 41 16.33 50–60 17 6.77 intermediate coarse flint 60–70 5 1.99 70–80 4 1.59 80–90 10 3.98 sand layer 90–100 14 5.58 lower main gravel 100–110 16 6.37 110–120 31 12.35 4.3.1. Detached pieces The majority of the excavated lithic assemblage comprises detached products. A total of 238 flakes and flake fragments were identified within the excavated gravel, which represents 94.8% of the whole assemblage (figure 6). All were made of flint and displayed no evidence of thermic alteration. Most displayed some degree of patination (n = 213, 89.5%), although its extent varied between flakes. ‘Roundness’ and edge damage levels were also variable within the flake assemblage, with 159 (66.8%) displaying some damage along their edges and 175 (73.5%) displaying a degree of rounding (i.e. displaying rounding of ‘2’ or above on a scale of 1–4). A few were exceptionally well preserved and had probably experienced little to no disturbance after discard (‘scale 1’; n = 63, 26.5%). Most displayed light to moderate evidence of disturbance (‘scale 2’; n = 165, 69.3%), while a small number were more heavily rolled (‘scale 3’ or ‘4’; n = 10, 4.2%), and 65 displayed evidence of being snapped/fragmented (27.3%). Complete flakes ranged in size between 9.4 and 56.8 mm in length, 7.0–53.1 mm in width, 1.3–18.6 mm in thickness and 0.2–40.2 g in weight (table 4 and figure 6). Figure 6. Flake artefacts from the 2020 excavations at Fordwich with varying levels of taphonomic damage. For example, flakes (c) and (d) were scaled at rounding level 4 and flake (h) was scaled at level 2. Seventy-two (30.3%) flakes possessed cortical striking platforms and 121 (50.8%) retained some evidence of dorsal cortex. Flakes displaying non-cortical platforms total 132 (55.5%), while those with no cortex on their dorsal surface total 117 (49.2%). Of those with non-cortical striking platforms, 53 (22.3%) retained evidence of platform preparation through faceting (a limited number with multiple facets appear typical of biface production, many are not clear), 36 (15.1%) had no visible evidence of preparation, and this was not clear for a further 43 (18.1%). Of those displaying evidence of previous flake removals on their dorsal surface (n = 156, 65.5%), an average of 2.1 removals were recorded. A few flakes displayed scars suggestive of secondary flaking (i.e. being flake cores; figure 6e), but fluvial processes could equally be responsible.
|
|
|
Post by Admin on Jun 30, 2022 21:18:20 GMT
4.3.2. Cores A total of four flint cores were identified (figure 7; electronic supplementary material, figure S3; Information 1). Three were small in maximum dimensions, measuring between 28.2 and 11.3 mm, and weighing 19.1–5.1 g. Two of these cores displayed cortex covering approximately 25% of their surface, while the other was non-cortical. Platform preparation was minimal, and no small cores possessed faceted platforms. No knapping accidents were present. Two of the small cores displayed four flake scars, while one appeared more highly exploited and had eight visible removals (although see discussion regarding its designation as anthropogenic; electronic supplementary material, figure S3). One small core is highly rolled, obfuscating the flake scars. The larger core weighed 284 g, had a maximum dimension of 84.5 mm, was exploited multifacially, displayed one unifaceted platform, displayed multiple misdirected impact points and two hinge fractures, and had seven visible flake scars (figure 7). Figure 7. Examples of retouched pieces and a core from Fordwich Pit. The double pointed retouched implement is in the top left (a) while the largest of the scrapers can be seen in the bottom left (b). The largest core is on the right (c). See the electronic supplementary material, information for three-dimensional models of the core and double-pointed implement. 4.3.3. Retouched pieces A small number of potential retouched artefacts were identified. These included three scrapers, one double-pointed implement, and two notched flakes (figure 7), while a further flake displayed evidence of continuous retouch on a portion of its edge. The scrapers were fragmented, but the portions recovered were between 34 and 50 mm in length. The double-pointed implement is only 19 mm in length but displays clear direct, abrupt continuous retouch to produce two symmetrical points (figure 7; electronic supplementary material, Information S4). The scrapers displayed direct continuous edge modification (figure 7). All were made from flint and displayed light taphonomic damage, which was distinct from the deep, continuous removals indicative of intentional retouch. The recovery of these implements from fluvial deposits must, however, be stressed; non-anthropogenic flaking processes potentially contributed to the observed edge modification in some instances. 4.3.4. Excavated neolithic artefacts Immediately above the excavated gravel mass was a small collection of Neolithic artefacts, supporting our assertion that the intermediate sand layer and upper gravels found elsewhere at Fordwich (and described by Smith [2]) have not been removed through industrial quarrying processes. These finds are not the focus of the present article and will be described in future work, but include ceramics, flint blades, debitage and a hammerstone. 4.3.5. Surface-discovered lithic artefacts Twenty-two lithic artefacts were found through ad hoc field walking and in the immediate vicinity of the excavated trenches. Surface artefacts do not appear to be in particularly dense concentrations, but to what extent this has been influenced by the prior removal of artefacts as aggregate or collection we do not know. All are flakes, and a combination of Lower Palaeolithic and more recent periods of occupation are represented. This is in keeping with elsewhere in the Stour Valley.
|
|
|
Post by Admin on Jul 1, 2022 17:27:17 GMT
Immediately above the excavated gravel mass was a small collection of Neolithic artefacts, supporting our assertion that the intermediate sand layer and upper gravels found elsewhere at Fordwich (and described by Smith [2]) have not been removed through industrial quarrying processes. These finds are not the focus of the present article and will be described in future work, but include ceramics, flint blades, debitage and a hammerstone.
Twenty-two lithic artefacts were found through ad hoc field walking and in the immediate vicinity of the excavated trenches. Surface artefacts do not appear to be in particularly dense concentrations, but to what extent this has been influenced by the prior removal of artefacts as aggregate or collection we do not know. All are flakes, and a combination of Lower Palaeolithic and more recent periods of occupation are represented. This is in keeping with elsewhere in the Stour Valley.
2020 excavations support Smith's [2] assertion that a ‘large number of amorphous flakes' are present at Fordwich. Indeed, in approximately 4 m3 of excavated gravels 238 morphologically diverse (shape and size) flake artefacts were discovered alongside 13 small retouched and core pieces. While similar concentrations of tools cannot automatically be assumed elsewhere at the site, it is reasonable to assume that several thousand of these small artefacts made their way into the aggregate output during the early twentieth century and were not retained alongside the handaxes. Given the limited number of flake artefacts from Fordwich retained in modern collections, there is clear evidence of artefact selection bias at this site during the first half of the twentieth century. A letter contemporary with the original quarrying at Fordwich suggests that most artefacts were found in the extreme west of the pit [4]. We cannot speak to this point in detail, but our trenches were located to the northwest of the pit and clearly retain strong evidence of flint artefacts.
We found evidence of artefact-based separation between upper and lower elements within the main gravel mass, but there were no clear stratigraphic trends in terms of rolling or pseudo retouch, with ‘fresh’ flakes being found throughout. There is a tendency for the most heavily rolled artefacts (those graded as ‘3’ or ‘4’) to be lower in the sequence, as eight of the ten flakes returning this score were at least 90 cm from the top of the gravel. We would stress that this observation is based on a small sample, and it is only through additional work that any taphonomic stratigraphic alignment can be determined.
Multiple authors stress there to be taphonomic variation within the known handaxe assemblage [2,4,20]. Ashmore [20] assigned abrasion based on a scale of one to five. From this, only four handaxes were regarded as perfectly fresh, 53 were a little worn, 151 were quite worn and 15 were heavily worn (although also see Roe [4]. Smith [2] noted one in 14 of the Fordwich handaxes to be heavily rolled. There is probably overlap between our ‘fresh to minimally worn category’ (n = 63, 25.1%) and Ashmore's [20] separate ‘fresh’ and ‘a little worn’ categories. Assuming this overlap to be correct, our flake data align well with these previous assessments and there is striking similarity in taphonomy between our flake assemblage and the known handaxes.
We have not recovered any handaxes to date and cannot contribute much to technological and morphological discussions on the existing Fordwich bifaces. However, contra Wymer [3], Roe [5] and Bridgland & White [18], we would not rule out the possibility of soft-hammer flake removals having been used to produce some handaxes at Fordwich. Indeed, while many flakes display cortex on their dorsal surfaces, few previous flake removals, and unprepared or minimally prepared flake platforms, and are therefore in agreement with previous studies identifying low numbers of removals on handaxes and the use of hard-hammer percussion [3,5,18], some do not. A few flakes display a prepared and small platform, diffuse bulbs of percussion, and low thickness to width ratios (figure 8), all characteristics associated with the soft-hammer technique [58]. Some display evidence of three or more previous flake removals on their dorsal surface, indicating heavier core-working than often assumed at Fordwich (potentially including heavily flaked handaxes, such as the example in figure 1). Most of the potential soft-hammer flakes were found in the upper main gravel (5 of 8 potential examples), so their presence during MIS 15 is dependent on the younger dates from this level being due to reworking. Potentially, some younger flakes (e.g. MIS 13) could have been reworked into these sediments during their MIS 10/11 disturbance. A sample of three from below the approximately 542 ka sand layer is not, in our opinion, large enough to conclusively argue for the presence of soft-hammer flaking.
|
|
|
Post by Admin on Jul 2, 2022 17:07:15 GMT
6. Discussion A century after Lower Palaeolithic artefacts were first discovered at Fordwich, the site is revealed to retain artefact-bearing gravels dating to approximately 542 000 years ago. These are among the oldest radiometrically dated artefact-bearing sediments in Britain and represent the fifth oldest archaeological occurrence currently known on the island (after Happisburgh, Fakenham Magna, Pakefield and Sapiston). Further, we provide evidence that the 330 handaxes recovered from Fordwich during the 1920s are older than approximately 437 ka, and are most plausibly contemporary with the newly excavated MIS 14 sediments. Assuming an absence of occupation during the cold MIS 14 glacial stage [10,34], Fordwich provides evidence of Acheulean hominins in southeast Britain during the MIS 15 interglacial (approx. 560–620 ka). This makes Fordwich the oldest directly dated Acheulean site in Britain. Fordwich is also revealed to be the only directly dated pre-MIS 13 Acheulean site in northern Europe to display a known handaxe assemblage numbering into the hundreds. After decades of only being mentioned in passing, Fordwich can now be considered a crucial piece of the pre-Anglian Palaeolithic puzzle in northwestern Europe. No handaxes were discovered during these new excavations, but they have revealed technological components previously unknown at Fordwich. We provide the first evidence for the presence of scrapers, with Smith [2], Ashmore [20] and Roe [5] not previously reporting their presence. While limited in number, the scrapers are unmistakable, displaying direct continuous edge modification (figure 7). Two were collected from the base of the upper main gravel, while the scraper in figure 7 was found directly beneath the sand layer that dates to approximately 542 ka. We are hesitant to compare the Fordwich scrapers with well-documented examples from Warren Hill, High Lodge and Moulin Quignon [10,13], due to their limited number and fragmentary nature, but our initial impression is that they are smaller than those usually reported during the British Lower Palaeolithic and less intensively worked than those from High Lodge and Warren Hill. Whether this is derived from our selective sample, fluvial processes, or hominin intention, it remains to be seen. Additional retouched pieces were also found, including two notched flakes. It is, however, the 19 mm long double-pointed retouched implement that is most intriguing (figure 7). Seemingly retouched to create two points, it potentially represents the production of a very small piercing/boring tool. Assuming the younger MIS 10/11 age of the upper main gravel is the result of reworking, then its recovery from the base of these gravels may reflect an MIS 15 age for this artefact. A potential role for fluvial edge damage in the characterization of these implements cannot be ruled out entirely. Similarly, we found evidence suggesting the production of cores smaller than those typically reported from the Lower Palaeolithic of northwest Europe. The example in electronic supplementary material, figure S3 appears to display eight removals from the same platform and was discovered in situ at the top of the lower main gravel (i.e. it is securely dated to approx. 542 ka). While it is easily differentiated from the chipped fluvial gravels at Fordwich, we remain hesitant in assigning it as a ‘miniaturized’ (cf. [59]) lithic artefact. A three-dimensional model of the object is available in the electronic supplementary material, and we invite critical appraisal of it. Three other cores are consistent with northern European Lower Palaeolithic reduction techniques (figure 7). The presence of small Lower Palaeolithic technologies should not be surprising, having been reported elsewhere in Britain (e.g. Beeches Pit, Layer 5a [22], and Boxgrove [unpublished]). A large number of flake artefacts were discovered during the original quarrying work at Fordwich Pit [2], although little is known about their technological characterization. Our excavations discovered 238 flakes and flake fragments. These represent the total number to have been agreed on by all three analysts. While many are unmistakable (figures 6 and 8), and all contain multiple features indicative of being human made, we recognize that the fluvial deposition of these artefacts may mean a number (most likely smaller examples) could potentially be formed by non-anthropogenic processes. Smith [2] suggested that ‘Clactonian’ flakes, which we interpret to mean larger flake tools (in line with historical perceptions of these artefacts [60]) were previously found. Following Roe [5], we have not yet found any evidence of such tools. Many flakes are consistent with having been detached during the manufacture of ‘crude’ minimally flaked handaxes (see above), as is generally suggested for the Fordwich bifaces [4–7]. Equally, some may represent expedient flake cutting tools produced through hard-hammer percussion on cores, despite their small size [61–63]. Their small size is not surprising as locally abundant fluvial flint nodules are generally less than 10 cm in maximum dimension. Some flakes display evidence of extended reduction sequences, with 40 exhibiting three or more previous flake removals on their dorsal surface. There is no depositional structure to the presence of these flakes, with broadly equal numbers in the upper and lower main gravel layers. Thus, we are confident that, at least on occasion, extended reduction sequences were being undertaken at Fordwich during MIS 15. Whether this was linked to more extended handaxe production sequences remains to be seen (see above discussion on soft-hammer flakes). There are no strong technological, morphological or taphonomic differences between flakes in the upper and lower main gravel, lending support to the notion all artefacts date from MIS 15 (table 5).
|
|
|
Post by Admin on Jul 3, 2022 17:07:25 GMT
Table 5. Average technological, taphonomic and morphological comparisons between complete flakes in the intermediate coarse flint and upper main gravel (UMG), dated to MIS 10/11, and those from the sand layer and lower main gravel (LMG), dated to at least MIS 14.
length (mm) width (mm) thickness (mm) weight (g) cortical coverage (%) number of dorsal removalsa step, hinge and overshot terminations (%) rounding scale UMG 22.5 19.6 5.7 3.8 37.1 2.1 25 1.8 LMG 23.3 20.7 5.7 4.0 36.0 2.2 28.4 1.8
Prior to the recent discoveries at Fordwich, scrapers were unknown at the site [2,5]. Their presence below the approximately 542 ka sand layer confirms their MIS 15 production, making them among the oldest known scrapers in Britain. Confirming the MIS 15 age of the previously discovered handaxes is a little more complicated. Smith [2] notes the handaxes to have been recovered from ‘near the brow of the hill’, which would put them closer to our new MIS 14 excavations than the Bridgland et al. [7] exposures. Irrespective, at an absolute minimum the handaxes must date to approximately 437 ka as the main gravel mass at the 1998 exposure is below the sands dated to 423 ± 29 and 437 ± 29 ka. While we have not yet dated the lower main gravel at the Bridgland et al. [7] exposures, it is likely that, as with the lower main gravel at our recent excavations, they date to MIS 14 (electronic supplementary material, figure S2). Indeed, the Bridgland et al. [7] upper gravels and loam are only present in part of the western edge of the quarry (figure 3), while the lower main gravel stretches across the western edge of the Fordwich Pit between the 1998 and 2020 locations (figure 3). This covers the western edge of the pit near the brow of the hill, where most of the handaxes were recovered [2,4]. Our excavations reveal this sediment to contain artefacts dating to the early Acheulean period in Europe. Thus, at present, Fordwich's handaxes are most plausibly contemporary with our newly excavated MIS 14 sediments. The roughly shaped character of the Fordwich bifaces, which are atypical for MIS 12/13, further supports this interpretation [10,18].
6.1. Placing Fordwich among the wider Acheulean of Europe Until now it has been difficult to place Fordwich among the wider Acheulean of Europe. The French site of Moulin Quignon displays evidence of Acheulean hominins in northwestern Europe during MIS 17 [13,34], and thus securely pre-dates Fordwich. From the five handaxes assigned to this early date, several are similar in form to those recovered at Fordwich (i.e. limited flake removals, relatively thick, and irregular in shape). Rampart Field (Suffolk, UK) has a single handaxe associated with an ESR date of 680 ± 26 ka, although it was recovered from greater than 1 m above the dated sediment and an MIS 15 age is preferred by the authors [32]. Warren Hill has a substantial collection of historically collected handaxes, with some argued to have been redeposited from MIS 15 contexts during the Anglian glaciation [10,32. Thus these artefacts could be direct contemporaries of those from Fordwich presenting a rare occurrence in the Pre-Anglian Lower Palaeolithic of northern Europe; two contemporary and substantial Acheulean assemblages. Brandon Fields and Maidscross Hill which both display large numbers of historically collected handaxes could also be from MIS 15 but are less secure in their attribution. Thus while they may be contemporaneous with Fordwich and display sizeable collections for comparison their precise place among the Acheulean of Europe is unclear. As it currently stands Fordwich is the only one of these four British occurrences with directly dated MIS 14 sediments demonstrated to contain artefacts. Thus Fordwich probably represents the joint second oldest Acheulean occurrence in northwestern Europe but it is currently the oldest after Moulin Quignon with radiometrically dated artefact-bearing sediments (see above regarding la Noira, Central France).
It is notable that the crude and highly rolled handaxes from Warren Hill conform with the elongated, thick, and poorly worked handaxes that characterize Fordwich [3–7,10,18]. Similarly, Brandon Fields and Maidscross Hill are also dominated by ‘cruder’ less heavily flaked handaxes, in line with those from Fordwich [10]. Together, these four sites seemingly indicate bifacial large cutting tool (LCT) technological uniformity in Britain during MIS 15 [16]. Currently, it is unclear how this similarity is related to cultural, raw material and reduction factors. The application of ‘quantitative genetic’ approaches will help to decipher the relative contribution of each of these factors [64]. Fordwich is the only MIS 15 context in Britain with confirmed evidence of scrapers. While further discoveries are required to fully explore their technological characterization, they appear closer to the MIS 17 Moulin Quignon scraper than the more highly worked MIS 13 examples from Warren Hill and High Lodge [10,13]. Their presence alongside handaxes further suggests that the MIS 13 separation in handaxe and scraper cultural groups, as suggested by Davis et al. [10], may not have an origin in MIS 15 populations.
The 700 000-year-old Acheulean site of la Noira in Central France pre-dates Fordwich by two MIS stages and is one of the most technologically diverse pre-MIS 13 sites in Europe, providing a sample of bifaces, flakes, cores and retouched pieces [65]. The technological similarity to Fordwich (at least as we currently understand it) is intriguing but, as stated above, a more detailed picture of the flake, core and retouched artefacts is needed in advance of any comparison. Similarly, once we have a more accurate understanding of the small artefacts from Fordwich, comparison with tools from Beeches Pit (MIS 11) could aid interpretation of the nature and extent of any technological continuity in small tool production through this period [22]. Comparison with small artefact assemblages recorded elsewhere in the European and the Levantine Lower Palaeolithic will identify similarities and differences at a continental level [59,66,67].
|
|