- Email: [email protected]
From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks
Quaternary International xxx (2015) 1e8
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks s a, b M. Guardiola a, b, c, *, J.I. Morales a, b, J.M. Verge de Paleoecologia Humana i Evolucio Social, C/ Marcel·lí Domingo, s/n, 43007, Tarragona, Catalonia, Spain IPHES: Institut Catala ria Universitat Rovira i Virgili (URV), Av. Catalunya, 35, 43002, Tarragona, Catalonia, Spain Area de Prehisto c Laboratoy Arch eologie et Peuplement de l'Afrique, Department of Genetics and Evolution, Anthropology Unit, University of Geneva, Switzerland a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
Two main knapping strategies can be used to start bifacial reduction on a lithic cobble or nodule: the alternate strategy, in which first one face is knapped and then the other; and the alternating strategy, in which both faces are removed in the same sequence, interspersing core about-turns between strikes. Flaking reduction of spherical and elliptical blanks (cobbles or nodules) is a common knapping process documented in many archaeological records. Rounded and thick edges require different fracture parameters and give rise to constraints in terms of viable knapping methods. When analysing abandoned cores, it is only possible to see the last strikes, so it is important to know how they were shaped or exploited in the earlier knapping stages in order to understand the entire reduction process. As cortical flakes are the direct evidence of the first reduction phases, we undertook an experimental programme for the purpose of comparing the first flakes generated using the alternate and alternating knapping strategies. We have focused our efforts on identifying and diagnosing distinctive features produced by each strategy in the first or cortical flakes. Our study indicates that several platform attributes can be considered as diagnostic features to differentiate between the alternate and alternating knapping strategies, and that this kind of analysis can be translated to archaeological assemblages. © 2015 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Alternate Alternating Flaking sequence Cortical flakes Nodules/cobbles
1. Introduction Understanding the way in which stone tools were produced in the past is important in order to infer the technological behaviours (Pelegrin, 1985, 1993; Delagnes and Roche, 2005; Braun et al., 2008; Stout et al., 2008, 2010) and capabilities of extinct hominin populations. Bifacial reduction strategies such as the alternate and alternating methods can be considered basic flaking methods that appeared early on in the technological record and were applied throughout hominin evolution (Pelegrin, 2005). Therefore, identifying the application of the alternate and alternating methods, as well as their origin and evolution, could be useful in constructing the referential framework for technological evolution as well as for
de Paleoecologia Humana i * Corresponding author. IPHES: Institut Catala Social, C/ Marcel$lí Domingo, s/n, 43007, Tarragona, Catalonia, Spain. Evolucio E-mail addresses: [email protected] (M. Guardiola), jignacio.morales@gmail. s). com (J.I. Morales), [email protected] (J.M. Verge
all of the derived cognitive, motor skills, technological and cultural implications. Given the continuity and evolution of the reduction sequence (Braun et al., 2005), in a state of abandonment, many tools and cores do not exhibit the necessary attributes to identify how the knapping sequence was started, so primary reduction flakes may be a reliable indicator of the knapping sequences carried out. If it is possible to experimentally identify distinguishing attributes on flakes, then the recognition of different first stage reduction strategies could be translated to the archaeological record. Bifacial tools (e.g. handaxes) and cores (e.g. discoid) can be shaped and exploited using different types of blanks like large flakes, cobbles and nodules, or slab-like fragments as a matrix. From a technical point of view, cobbles and slabs tend to present thick rounded or squared edges that usually make beginning to perform bifacial reductions difficult (Callahan, 1979, pp. 64; Jones, 1994). The result of this starting phase determines the entire shaping process and is dependent on the structure (Roth and Dibble, 1998), shape (White, 1998; White and Ashton, 2003), and
http://dx.doi.org/10.1016/j.quaint.2015.08.039 1040-6182/© 2015 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
2
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
size (Crompton and Gowlett, 1993) of the raw material of the blank. The application of bifacial knapping strategies on tools and cores requires removing flakes from both faces of the blank, and during this process, the cortex should be more or less completely detached, the volume controlled and adapted, and the shape regularised. If the blank has restrictions in morphology, the success of the first reduction stage can be decisive when a specific designed form, especially a biface, is sought. In this scenario, the alternating flaking method has been largely proclaimed as an appropriate strategy for transforming squared and thick edges into bifacial ones (Burton, 1980; Andrefsky, 1998; Baena, 1998; Moore, 2003; Pelegrin, 2005). In addition, the alternating method has been found to be highly capable of invasiveness, especially compared to the alternate method (Moore, 2003). The alternate method can be deconstructed into two different knapping sequences, affecting first one face of the blank, and then the other. In contrast, the alternating method involves working both faces at the same time in a single sequence, involving much more complexity in core management, work planning and the ability to anticipate. In the technological record, some kinds of structured reduction strategies seem to appear in Early Stone Age assemblages (Delagnes and Roche, 2005) as can be interpreted from the concatenation of short series of sub-parallel removals on rounded blanks that can be seen, for instance, in the Gona lithic assemblage (Semaw, 2000). However, the first clear identification of alternating sequences seems to appear somewhat later. The first evidence of the use of the alternating method is claimed to be from the sites of Kanjera South (Kenya) around 2.0 Ma (Stout et al., 2010, pp. 477), the Gadeb sites (Ethiopia) (de la Torre, 2011, pp. 778) ca. 1.4e0.7 Ma, and Gesher Benot Ya'aqov (Israel) ca. 0.8 Ma (GorenInbar et al., 2011, pp. 1909). If the alternating method is understood as a major technical evolutionary innovation for the purpose of reducing thick edges in bifacial reductions, then it could have played a significant role in the technical evolution of large cutting tool techno-complexes and, therefore, it can be presented as something to be taken into account in the structure of Acheulean technological evolution. Based on this scenario, we have designed an experimental programme that aims to analyse the attributes exhibited by flakes produced during the early reduction stages of the alternate and alternating flaking methods. The main goal of the work is to propose some specific variables to be interpreted, like features exclusive to the alternate or alternating methods, allowing these different reduction strategies to later be identified in the archaeological record using the flakes rather than highly reduced or transformed tools or cores. 2. Background and terminology The alternate and alternating concepts have been described, interpreted and applied diversely in the literature, leading to some confusion with regard to the meaning of the terms and the technical strategies involved. In this work, considering the different points of view proposed and applied (Callahan, 1979; Burton, 1980; Whittaker, 1994; Inizan et al., 1995, 1999; Baena, 1998; Ashton and White, 2003; Goren-Inbar et al., 2011), we understand the alternate method to be when the two faces of the tool/core are independently removed, working first on one face and then on the other. On the other hand, we consider the alternating method to consist of knapping both faces in the same sequence, interspersing core about-turns between each strike. There is a problem derived from the existing terminology and related to the difficulty of delimiting reduction stages (Bleed,
2002) and deconstructing reduction processes into strategies, sequences, series, methods or reduction systems. Assessing knapping complexity is a key component in explaining the cognitive changes observed in the evolution of stone technology, but there is a lack of consensus as to which criteria should be used (Stout et al., 2010, pp. 476). This makes comparing flaking strategies very difficult. Faced with this problem, which is not easy to solve, we aim to clearly define the different flaking actions and products used in this work as well as those used to make experimental comparisons. 2.1. The alternate strategy By ‘alternate strategy’ we refer to the process by which one face is partially or completely worked before moving on to the second face (Fig. 1). It is conceptually bifacial, but the hierarchical sequences employed are not. Alternate reduction can be performed with a mixture of methods. The most instinctive of these uses blow by blow removals, using the most highlighted ridges on each face. But in early reduction stages on rounded blanks, taking advantage of the previous removal ridges with sub-parallel or adjacent flaking (Pelegrin, 2005) provides an easier way to reduce the blank and to remove the cortex. Theoretically, only one or very few core turnovers are needed between the two sequences. As a result, in the alternate strategy there are two different production series or shaping sequences, referred to in this work as the ‘first alternate’ and ‘second alternate’. 2.2. The alternating strategy By ‘alternating sequence’, we refer to the process by which both faces are flaked in a single sequence, interspersing core about-turns between each strike (Inizan et al., 1995, 1999) (Fig. 1). Alternating involves core rotation and tilt rectification, but in addition, continuous core about-turns are performed throughout the sequence. During the knapping process it is necessary to change core position and flaking planes continuously, creating a continuous removal series that entails a different type of complexity in visual reconnaissance (the volume of each side), manual rectifications (continuous core aboutturns) and execution (hammer actions). The last scar is the primary hierarchical element because it determines the position and orientation of the core on the next strike. The concavity of the last scar forces the knapper to apply oblique percussions (to the right, to the left, and so on), which are needed to create a new platform for the next blow on the other face of the blank. In this sense, the alternating sequence involves an anticipatory strategy. 3. Materials and methods In this experimental work, we produced two different samples of alternate and alternating flakes. Different experimental series were created using thirty-five cobbles of flint and quartzite (Table 1). The flint nodules come from Zaragoza region (Spain). It is usually homogeneous and non-fissured raw material, and the cortex is perfectly suitable to strike. The quartzite pebbles come from Tagus River, near to Lisbon (Portugal). They are tough and display fine-grained and homogeneous raw material. The cobbles and nodules had regular shapes and elongated elliptical or spherical morphologies and some slab-like nodules with right-angled edges were also selected.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
3
Table 1 Summary of blank shapes (pebbles and cobbles), raw materials and applied flaking methods. Exper.
Morphology
Raw material
Mass (kg.)
Method
Blows
Complete Flakes
P1 P2 P3 P4 P5 P6 P7 Pil.1 Pil.2 Pr-2 N1 N2 N3 N4 N5 N7 N8 N11 N16 N17 N18 N19 N23 N24 N25 N28 N37 N36 N30 N29 N34 N31 N20 N9 N35 Total
Pebble Pebble Pebble Pebble Pebble Pebble Pebble Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule
Quartzite Quartzite Quartzite Quartzite Quartzite Quartzite Quartzite Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint
1.74 3.04 1.92 2.46 2.44 2.44 2.28 2.20 1.41 _ 2.34 2.84 3.34 2.68 3.08 2.18 1.76 2.58 2.82 1.68 1.72 1.38 1.70 0.90 2.18 1.92 2.04 1.18 2.12 1.34 3.10 6.60 2.44 1.28 1.72
Alternating Alternating Alternate Alternate Alternating Alternate Alternating Alternate Alternate Alternate Alternating Alternating Alternating Alternating Alternating Alternating Alternating Alternating Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternating Alternating Alternating Alternate Alternating Alternating Alternating Alternating
13 33 31 38 28 22 16 17 17 34 32 21 28 35 43 26 19 27 36 16 13 37 9 9 15 18 29 27 33 21 36 17 46 26 31 899
11 35 29 35 25 21 17 15 15 33 34 17 29 30 37 26 20 28 34 14 12 35 9 9 13 18 30 27 28 20 38 18 46 26 29 863
Fig. 1. Alternate and alternating methods scheme.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
4
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
The cortex removal sequences were carried out by one of the authors (M.G.), using the free-hand percussion technique. Two stone hammers were used: P1, quartzite (620 g, density ~ 2.45 g/ cm3), for removals requiring tougher percussion (only the first removal in the reduction of each blank); and P2, a hard sandstone hammer (550 g, density: ~2.2 g/cm3) for all the other blows. The flakes were labelled, divided by sequences and all of the aspects involved were meticulously written down: flake number, sequence type, and blow direction were marked on the ventral face of each flake; starting blank zone, rotation direction, about-turns, hammerstone and other observations were also documented. We only prepared the platform when it was necessary in order to understand the platform preparation requirements for each method. The reduction was finished when the entire perimeter of the blank was flaked. As mentioned earlier, three sequences were recorded, ‘first’ and ‘second’ for the alternate reduction method and a single ‘alternating’ sequence for the alternating method. Blow direction was defined as to the left, to the right or perpendicular to the centre of the blank or longitudinal axis (from the knapper's point of view). 3.1. Useful recorded attributes Most of the attributes considered in our database are similar to those used in other experiments (Amick et al., 1988; Toth, 1988; Bradbury and Carr, 1995, 1999; Dibble et al., 2005; Braun et al., 2005, 2008; Marwick, 2008; Eren and Lycett, 2012) and both quantitative and qualitative variables were documented (Fig. 2): - Technical axis (TA). Distance from impact point to the distal end of the flake along the technological axis and perpendicular to the major platform axis. - Maximum width (MW). Maximum width of the flake oriented by the technical axis. Left maximum width (LMW). Maximum width axis is divided in two by the technical axis, we record the left segment. This variable allows us to observe the flake skewness and to approximate mass distribution in two dimensions.
- Left surface (LS). Surface in mm2 to the left of the technical axis. To record this variable, digital photographs were taken and then the surface was measured using ImageJ software (http://imagej. nih.gov/ij/). - Maximum platform width (PtMW). - Platform thickness (PtT). Platform thickness measured at the impact point position. - Platform left width (PtLMW). Distance from the impact point to the left end of the platform. - Platform type: cortical (C), plane (P) and faceted (F). - Butt shape. Seven morphological categories were created to record this variable (Fig. 3): A: flateconvex, B: helicoidal, C: scalene, D: convexeconcave, E: punctiform, F: inverted triangletrapezoid and G: dissymmetrical. - Previous percussion mark. In a preliminary experimental test, this singular feature was observed, and consists of small semicircular fractures at the opposite end of the impact point (Fig. 4). These fractures were caused by the percussion of the previous removal. Three categories were established: no presence (N), left (L) and right (R). - Cortical index. The percentage of dorsal cortex was recorded following the ‘eight-stage system’ (Roth and Dibble, 1998; Braun et al., 2008), in which 1 ¼ 100%, 2 ¼ 91e99%, 3 ¼ 71e90%, 4 ¼ 51e70%, 5 ¼ 31e50%, 6 ¼ 11e30%, 7 ¼ 1e10% and 8 ¼ 0%. - Cortex reserve type. The flakes were distributed into different cortex reserve categories using Toth's technological flake categories (TFC) (Toth, 1982, 1985). This system has normally been interpreted in relation to reduction intensity and flaking strategy studies: types IIII suggest primary flaking and IVVI reflect higher degrees of reduction (Toth, 1985), taking core size into account (Braun et al., 2005, 2008).
3.2. Statistical approach We tested for significant differences and associations in the quantitative and qualitative variables analysed. The distribution of
Fig. 2. Recorded attributes.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
5
Fig. 3. Butt shape type classification: A: flateconvex, B: helicoidal, C: scalene, D: convexeconcave, E: punctiform, F: inverted triangle-trapezoid and G: dissymmetrical.
the paired quantitative variables was tested using the ShapiroeWilk test for normality. The entire set of measurements displayed a non-normal distribution, so a non-parametric statistic ManneWhitney two-tailed U test was used. The ManneWhitney test checks for statistical differences in the medians of two independent samples with a null hypothesis of equal medians. As a nonparametric method, it can test for differences in distributions of any given shape, assuming that the tested groups share the same shape. Correlations between the categorical variables, attributes and methods were explored using a correspondence analysis (CA) for counted data. In order to determine whether the significant dif-
the alternating method and 359 (41.6%) using the alternate method (first sequence ¼ 229 and second sequence ¼ 130) (Table 2). Some of these flakes were broken during knapping, and only when the fractures were clear (easy to refit), were they refitted and analysed. Sometimes, several flakes (usually two), were generated in a single strike (56 cases), in this case we noted the order of each and they were considered individually. In addition, sometimes repeated percussions on the same face were needed in the application of the alternating method; this was indicated by a scoring of ‘r’, which occurred in 120 cases. The data descriptive statistics data are provided in Table 3.
Table 2 Blank and flake production conducted in the experiment.
Alternate Alternating
Blanks
Total Flakes
Flakes/blank
Flakes/kg.
Sequence
16 19
359 504
22.5 26.5
11.05 12.2
1st ¼ 229; 2nd ¼ 130 Alt: 504
Table 3 Descriptive statistics of experimental database. 1st Alternate
TA MW LMW Total area Left area Thickness (prox) Thickness (med) Thickness (dist) PtMW PtMTI PtLMW
1st & 2nd alternate
2nd Alternate
Alternating
Mean
sd
CV
Mean
sd
CV
Mean
sd
CV
Mean
sd
CV
47.15 47.34 23.2 1741.12 828.98 9.22 7.75 4.82 20.52 6.44 10.63
19.53 19.02 10.39 1181.3 596.89 6.64 4.57 2.82 11.1 4.08 7.35
0.41 0.40 0.45 0.68 0.72 0.72 0.59 0.59 0.54 0.63 0.69
47.22 47.11 23.03 1720.28 821.98 9.06 7.64 4.75 20.47 6.22 10.3
19.54 18.86 10.31 1167.48 592.01 6.56 4.45 2.71 11.12 3.72 6.89
0.41 0.40 0.45 0.68 0.72 0.72 0.58 0.57 0.54 0.60 0.67
47.44 47.64 23.31 1767.8 841.41 9.24 7.74 4.85 20.9 6.5 10.7
19.83 19.24 10.49 1223.52 613.65 6.58 4.54 2.82 11.20 4.05 7.31
0.42 0.40 0.45 0.69 0.73 0.71 0.59 0.58 0.54 0.62 0.68
49.31 47.14 22.93 1839.97 883.71 8.82 7.67 4.79 17.89 5.43 9.45
19.59 19.53 10.76 1279.24 669.71 5.80 4.65 2.99 10.40 3.70 7.08
0.40 0.41 0.47 0.70 0.76 0.66 0.61 0.62 0.58 0.68 0.75
ferences observed in the analysis of the methods' characteristic features are statistically causal we tested for their discriminant and predictive power. We considered that the combination of diagnostic features could be checked through the use of a classificatory method to explore its robustness, load and applicability. For this purpose, we performed a binomial logistic regression, a generalised linear model that checks for causal relations when the dependent variable (alternate and alternating in this case) is binomial. 4. Results During our experiments, hand-controlled percussions generated 863 whole flakes: 504 (58.4%) of them were produced using
The results of the ManneWhitney for equal medians applied to the metric variables test are summarised in Table 4. No differences were found between the alternate and alternating methods in some of the most common recorded measurements like flake width and flake area. The proximal, medial and distal thicknesses did not yield significant differences either. However, significant differences were found in the length of the technical axis, with the alternating flakes longer than the alternate flakes. The platform maximum thickness, the maximum thickness from the impact point, and the left platform maximum width differences indicate that alternate flakes have thicker and broader platforms than the alternating flakes.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
6
M. Guardiola et al. / Quaternary International xxx (2015) 1e8 Table 4 ManneWhitney test results for differences between alternate and alternating metrical variables. * Remarks significant differences. p(same) Technical Axis Maximum Width Left Maximum Width Total Area Left Area Proximal Thickness Medial Thickness Distal Thickness Platform Maximum Width Platform Maximum Impact Thickness Platform Left Maximum Width
0.01256* 0.4544 0.0556 0.1607 0.3373 0.2195 0.2366 0.4094 0.00* 0.00* 0.00*
In the attribute associations, the CA analysis was initially performed in multiple ways, but for easier visualisation and interpretation of the results, they are presented using a simple CA analysis performed variable by variable. In order to more efficiently show the category associated with the knapping method, the first and second alternate sequences have been considered independently. Using the exploratory power of the CA, several significant and diagnostic associations between the categories of each variable and the three sequences were found. The cortex distribution on the dorsal faces of the first and second alternate flakes and the alternating flakes exhibits different patterns of cortical index according to the distribution of the categories. While alternating flakes display a heterogeneous pattern of middle to high range cortical ratios (from 30 to 90%), first alternate and second alternate flakes are associated with very low and very high cortex presence, respectively (Fig. 5, A). The platform variable analysis shows clear associations of flat platforms with c, e, b shapes for the alternating method. The alternate sequences are strongly associated with cortical f and g shapes in the case of first, and facetted a and d butt shapes for the second (Fig. 5, B and C). Interestingly, the occurrence of previous
Fig. 4. Origin and location of the previous percussion mark related to a later removal.
percussion marks, when it happens, is highly and almost exclusively associated with alternating flaking (Fig. 5, D). The exploration of the occurrence of synthetic TFC discretisation related to the three sequences correlates type II flakes with the alternate first sequence, type IV flakes with the second alternate and types V and VI with the alternating sequence (Fig. 5, E). In order to test the explicative power of each variable for the alternate and alternating methods, we transformed them and analysed them together in a logistic binomial regression. The Nagelkerke r2 indicator shows that 81.9% of the variance observed in the products of each knapping method can be explained through the analysed patterns (Nagelkerke r2 ¼ 0.819) (Table 5). The model's robustness is shown in the classificatory table with a balanced percentage of correctly classified cases of 93.4%. The method also indicates that the variables with the highest incidence in the regression equation are the complete set of platform quantifications (p ¼ 0.00; 0.013; 0.03; 0.008), the cortical index (p ¼ 0.04) Table 5 Classification table of binomial logistic regression. Alternate Alternate 167 Alternating 19 Overall percentage
Alternating
Correctly classified %
15 318
91.8 94.4 93.4
and platform type (p ¼ 0.00). 5. Discussion and conclusions We experimentally compared alternate and alternating flakes in early stages on rounded blanks, and we provide several distinctive traits with which to distinguish the products of each method. Technologically, and from an experimental point of view, despite its theoretical importance to bifacial reduction, little research has been done on the identification of the alternating method in relation to handaxe production. Archaeologically, being able to identify methods from flakes is very useful, as cores are frequently depleted and do not reflect the technological features of the reduction process. All of these factors have an impact on bifacial technologies. Adaptation to the shapes of different raw materials can be better understood by identifying alternating method flakes. Some of the morphometric traits of the experimentally produced flakes exhibit differences between the alternate and alternating methods. Firstly, the alternating method flakes are longer than the alternate flakes. Although greater invasiveness has been reported with the alternating method (Moore, 2003; White and Ashton, 2003), it is difficult to compare the two strategies because there are substantial differences between the first and second alternate sequence. Based on the shape of the initial edge, the level of invasiveness in the first and second alternate sequences is extremely variable. First alternate removals tend to produce short thick flakes. In contrast, second alternate removals lead to highly variable flake lengths, depending on the percussion conditions. Platform attributes are the best indicators of the alternating method and are identifiable from flakes, especially with regard to the impact point position (normally lateralised) and the presence of the last percussion mark (~25% of alternating flakes; <1% in alternate flakes). Compared to the alternate method, the alternating method yields flatter platforms, normally with scalene, punctiform and helicoidal-shaped butts. In terms of the technological flake categories (Toth, 1982, 1985), alternating method produce a great part of the type V flakes. In fact, the alternating method produces type V flakes from the beginning of the sequence, specifically, from the second removal. Therefore, related to reduction intensity and
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
7
Fig. 5. Correspondence analyses of categorical variables and knapping sequence type: A: Cortex Index, B: Butt shape classification, C: Platform type, D: Previous percussion mark presence, E: Technological flake categories.
levels of transport (Kimura, 2002), the alternating flaking method might produce a distortion of this model. Identifying the alternating method can provide valuable information about technological adaptation and cognitive evolution. To this end, we have defined some recognisable features contextualised in the initial reduction steps on a very common blank shape. In addition to our experiments with rounded blanks, further research needs to be conducted, primarily on larges flakes and knapper variability. In Acheulean bifacial technologies, the alternating method plays a fundamental role in subsequent core
technologies, it is important in Levallois strategies and in blade technologies. In all of them, blank rounded edges (semicircular cross-sections) are one of the most common blank shapes and the alternating flaking method provides an easy and systematic solution by which to reduce them. Acknowledgements This research has been partially supported by the projects 2014SGR-899 and 900 (AGAUR-Generalitat de Catalunya) and,
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
8
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
HAR2013-41197-P and CGL2012-38434-C03-03 (Ministerio de Economía y Competitividad). J.I. Morales has been funded by a predoctoral research fellowship (FI-B2-2014) from the AGAUR of Generalitat de Catalunya. We also want to thank to an anonymous reviewer of the present version as well as two others who reviewed a previous copy. They have improved our work. Finally, we want to thank to the organizers of the European Acheulean Congress, le ne Moncel and Danielle Schevre for hosting our Marie-he research.
References Amick, D.S., Mauldin, R., Tomka, S.A., 1988. An evaluation of debitage produced by experimental bifacial core reduction of a Georgetown Chert Nodule. Lithic Technology 17 (1), 26e37. Andrefsky, W.A., 1998. Lithics: Macroscopic Approaches to Analysis. Cambridge University Press, Cambridge. Ashton, N., White, M., 2003. Bifaces and raw materials: flexible flaking in the British early Paleolithic. In: Soressi, M., Dibble, H.L. (Eds.), Multiple Approaches to the Study of Bifacial Technologies. University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia, pp. 109e123. n a la talla de utillaje Baena, J. (Ed.), 1998. Tecnología lítica experimental. Introduccio rico. British Archaeological Reports International Series, 721. Oxford. prehisto Bleed, P., 2002. Obviously sequential, but continuous or staged? Refits and cognition in three late Paleolithic assemblages from Japan. Journal of Anthropological Archaeology 21, 329e343. Bradbury, A.P., Carr, P.J., 1995. Flake typologies and alternative approaches: an experimental assessment. Lithic Technology 20, 100e115. Bradbury, A.P., Carr, P.J., 1999. Examining stage and continuum models of flake debris analysis: an experimental approach. Journal of Archaeological Science 26, 105e116. Braun, D.R., Tactikos, J.C., Ferraro, J.V., Arnow, S.L., Harris, J.W., 2005. Flake recovery rates and inferences of Oldowan hominin behavior: a response to Kimura 1999, 2002. Journal of Human Evolution 48, 525e531. Braun, D.R., Tactikos, J.C., Ferraro, J.V., Harris, J.W., 2008. Oldowan reduction sequences: methodological considerations. Journal of Archaeological Science 38, 2153e2163. Burton, J., 1980. Making sense of waste flakes: new methods for investigating the technology and economies behind chipped stone assemblages. Journal of Archaeological Science 7, 131e148. Callahan, E., 1979. The basic biface knapping in the eastern fluted points tradition: a manual for flintknappers and lithic analysts. Archaeology of Eastern North America 7 (1). Crompton, R.H., Gowlett, J.A.J., 1993. Allometry and multidimensional form in Acheulean bifaces from Kilombe, Kenya. Journal of Human Evolution 25, 175e199. de la Torre, I., 2011. The Early Stone Age lithic assemblages of Gadeb (Ethiopia) and the developed Oldowan/early Acheulean in East Africa. Journal of Human Evolution 60, 768e812. Delagnes, A., Roche, H., 2005. Late Pliocene hominid knapping skills: the case of Lokalalei 2C, West Turkana, Kenya. Journal of Human Evolution 48, 435e472. Dibble, H.L., Schurmans, U.A., Iovita, R.P., McLaughlin, M.V., 2005. The measurement and interpretation of cortex in lithic assemblages. American Antiquity 70 (3), 545e560.
Eren, M.I., Lycett, S.J., 2012. Why Levallois? A morphometric comparison of experimental “preferential” Levallois flakes versus debitage flakes. PLoS ONE 7 (1), E29273. http://dx.doi.org/10.1371/journal.pone.0029273. Goren-Inbar, N., Grosman, L., Sharon, G., 2011. The technology and significance of the Acheulian giant cores of Gesher Benot Ya'aqov, Israel. Journal of Archaeological Science 38, 1901e1917. Inizan, M.L., Reduron-Ballinger, M., Roche, H., Tixier, J., 1995. Technologie de la e. CREP(4), Nanterre. pierre taille Inizan, M.L., Reduron-Ballinger, M., Roche, H., Tixier, J., 1999. Technology and Terminology of Knapped Stone. CREP (5), Nanterre. Jones, P.R., 1994. Results of experimental work in relation to the stone industries of Olduvai Gorge. In: Leakey, M.D., Roe, D.A. (Eds.), Olduvai Gorge, vol. 5. Cambridge University Press, Cambridge, pp. 254e298. Kimura, Y., 2002. Examining time trends in the Oldowan technology of beds I and II, Olduvai Gorge. Journal of Human Evolution 43, 291e321. Marwick, B., 2008. What attributes are important for the measurement of assemblage reduction intensity? Results from an experimental stone artefact assemblage with relevance to the Hoabinhian of mainland Southeast Asia. Journal of Archaeological Science 35, 1189e1200. Moore, M.W., 2003. Australian aboriginal biface reduction techniques on the Georgia River, Camooweal, Queensland. Australian Archaeology 56, 22e34. flexions sur le comportement technique. La signification culPelegrin, J., 1985. Re ge, pp. 72e91. turelle des industries lithiques, Actes du Colloque de Lie Pelegrin, J., 1993. A framework for analysing prehistoric stone tool manufacture and a tentative application to some early stone industries. In: Berthelet, A., Chavallion, J. (Eds.), The Use of Tools by Human and Non-human Primates. Clarendon Press, Oxford, pp. 302e314. Pelegrin, J., 2005. Remarks about archaeological techniques and methods of knapping: elements of cognitive approach to stone knapping. In: Russon, A.E., Begun, D.R. (Eds.), Stone Knapping: the Necessary Conditions for a Uniquely Hominin Behavior. McDonald Institute for Archaeological Research, Cambridge, pp. 159e169. Roth, B.J., Dibble, H.L., 1998. Production and transport of blanks and tools at the French Middle Paleolithic site of Combe-Capelle Bas. American Antiquity 63, 47e62. Semaw, S., 2000. The world's oldest stone artifacts from Gona, Ethiopia: their implications for understanding stone technology and patterns of human evolution between 2.6e1.5 Million years ago. Journal of Archaeological Science 27, 1197e1214. Stout, D., Toth, N., Schick, K., Chaminade, T., 2008. Neural correlates of Early Stone Age toolmaking: technology, language and cognition in human evolution. Philosophical Transactions of the Royal Society 363, 1939e1949. Stout, D., Semaw, S., Rogers, M.J., Cauche, D., 2010. Technological variation in the earliest Oldowan from Gona, Afar, Ethiopia. Journal of Human Evolution 58, 474e491. Toth, N.P., 1982. The Stone Technology of Early Hominids at Koobi Fora, Kenya: an Experimental Approach (Ph. D. dissertation). University of California, Berkeley. Toth, N., 1985. Oldowan reassessed: a close look at early stone artefacts. Journal of Archaeological Science 12, 101e120. Toth, N., 1988. Behavioral inferences from early stone artifact assemblages: an experimental model. Journal of Human Evolution 16, 763e787. White, M.J., 1998. On the significance of Acheulean biface variability in southern Britain. Proceedings of the Prehistoric Society 64, 15e44. White, M.J., Ashton, N.M., 2003. Lower Paleolithic core technology and the origins of the Levallois method in North-Western Europe. Current Anthropology 44 (4), 598e609. Whittaker, J.C., 1994. Flintknapping: Making and Understanding Stone Tools. University of Texas Press, Austin.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks s a, b M. Guardiola a, b, c, *, J.I. Morales a, b, J.M. Verge de Paleoecologia Humana i Evolucio Social, C/ Marcel·lí Domingo, s/n, 43007, Tarragona, Catalonia, Spain IPHES: Institut Catala ria Universitat Rovira i Virgili (URV), Av. Catalunya, 35, 43002, Tarragona, Catalonia, Spain Area de Prehisto c Laboratoy Arch eologie et Peuplement de l'Afrique, Department of Genetics and Evolution, Anthropology Unit, University of Geneva, Switzerland a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Available online xxx
Two main knapping strategies can be used to start bifacial reduction on a lithic cobble or nodule: the alternate strategy, in which first one face is knapped and then the other; and the alternating strategy, in which both faces are removed in the same sequence, interspersing core about-turns between strikes. Flaking reduction of spherical and elliptical blanks (cobbles or nodules) is a common knapping process documented in many archaeological records. Rounded and thick edges require different fracture parameters and give rise to constraints in terms of viable knapping methods. When analysing abandoned cores, it is only possible to see the last strikes, so it is important to know how they were shaped or exploited in the earlier knapping stages in order to understand the entire reduction process. As cortical flakes are the direct evidence of the first reduction phases, we undertook an experimental programme for the purpose of comparing the first flakes generated using the alternate and alternating knapping strategies. We have focused our efforts on identifying and diagnosing distinctive features produced by each strategy in the first or cortical flakes. Our study indicates that several platform attributes can be considered as diagnostic features to differentiate between the alternate and alternating knapping strategies, and that this kind of analysis can be translated to archaeological assemblages. © 2015 Elsevier Ltd and INQUA. All rights reserved.
Keywords: Alternate Alternating Flaking sequence Cortical flakes Nodules/cobbles
1. Introduction Understanding the way in which stone tools were produced in the past is important in order to infer the technological behaviours (Pelegrin, 1985, 1993; Delagnes and Roche, 2005; Braun et al., 2008; Stout et al., 2008, 2010) and capabilities of extinct hominin populations. Bifacial reduction strategies such as the alternate and alternating methods can be considered basic flaking methods that appeared early on in the technological record and were applied throughout hominin evolution (Pelegrin, 2005). Therefore, identifying the application of the alternate and alternating methods, as well as their origin and evolution, could be useful in constructing the referential framework for technological evolution as well as for
de Paleoecologia Humana i * Corresponding author. IPHES: Institut Catala Social, C/ Marcel$lí Domingo, s/n, 43007, Tarragona, Catalonia, Spain. Evolucio E-mail addresses: [email protected] (M. Guardiola), jignacio.morales@gmail. s). com (J.I. Morales), [email protected] (J.M. Verge
all of the derived cognitive, motor skills, technological and cultural implications. Given the continuity and evolution of the reduction sequence (Braun et al., 2005), in a state of abandonment, many tools and cores do not exhibit the necessary attributes to identify how the knapping sequence was started, so primary reduction flakes may be a reliable indicator of the knapping sequences carried out. If it is possible to experimentally identify distinguishing attributes on flakes, then the recognition of different first stage reduction strategies could be translated to the archaeological record. Bifacial tools (e.g. handaxes) and cores (e.g. discoid) can be shaped and exploited using different types of blanks like large flakes, cobbles and nodules, or slab-like fragments as a matrix. From a technical point of view, cobbles and slabs tend to present thick rounded or squared edges that usually make beginning to perform bifacial reductions difficult (Callahan, 1979, pp. 64; Jones, 1994). The result of this starting phase determines the entire shaping process and is dependent on the structure (Roth and Dibble, 1998), shape (White, 1998; White and Ashton, 2003), and
http://dx.doi.org/10.1016/j.quaint.2015.08.039 1040-6182/© 2015 Elsevier Ltd and INQUA. All rights reserved.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
2
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
size (Crompton and Gowlett, 1993) of the raw material of the blank. The application of bifacial knapping strategies on tools and cores requires removing flakes from both faces of the blank, and during this process, the cortex should be more or less completely detached, the volume controlled and adapted, and the shape regularised. If the blank has restrictions in morphology, the success of the first reduction stage can be decisive when a specific designed form, especially a biface, is sought. In this scenario, the alternating flaking method has been largely proclaimed as an appropriate strategy for transforming squared and thick edges into bifacial ones (Burton, 1980; Andrefsky, 1998; Baena, 1998; Moore, 2003; Pelegrin, 2005). In addition, the alternating method has been found to be highly capable of invasiveness, especially compared to the alternate method (Moore, 2003). The alternate method can be deconstructed into two different knapping sequences, affecting first one face of the blank, and then the other. In contrast, the alternating method involves working both faces at the same time in a single sequence, involving much more complexity in core management, work planning and the ability to anticipate. In the technological record, some kinds of structured reduction strategies seem to appear in Early Stone Age assemblages (Delagnes and Roche, 2005) as can be interpreted from the concatenation of short series of sub-parallel removals on rounded blanks that can be seen, for instance, in the Gona lithic assemblage (Semaw, 2000). However, the first clear identification of alternating sequences seems to appear somewhat later. The first evidence of the use of the alternating method is claimed to be from the sites of Kanjera South (Kenya) around 2.0 Ma (Stout et al., 2010, pp. 477), the Gadeb sites (Ethiopia) (de la Torre, 2011, pp. 778) ca. 1.4e0.7 Ma, and Gesher Benot Ya'aqov (Israel) ca. 0.8 Ma (GorenInbar et al., 2011, pp. 1909). If the alternating method is understood as a major technical evolutionary innovation for the purpose of reducing thick edges in bifacial reductions, then it could have played a significant role in the technical evolution of large cutting tool techno-complexes and, therefore, it can be presented as something to be taken into account in the structure of Acheulean technological evolution. Based on this scenario, we have designed an experimental programme that aims to analyse the attributes exhibited by flakes produced during the early reduction stages of the alternate and alternating flaking methods. The main goal of the work is to propose some specific variables to be interpreted, like features exclusive to the alternate or alternating methods, allowing these different reduction strategies to later be identified in the archaeological record using the flakes rather than highly reduced or transformed tools or cores. 2. Background and terminology The alternate and alternating concepts have been described, interpreted and applied diversely in the literature, leading to some confusion with regard to the meaning of the terms and the technical strategies involved. In this work, considering the different points of view proposed and applied (Callahan, 1979; Burton, 1980; Whittaker, 1994; Inizan et al., 1995, 1999; Baena, 1998; Ashton and White, 2003; Goren-Inbar et al., 2011), we understand the alternate method to be when the two faces of the tool/core are independently removed, working first on one face and then on the other. On the other hand, we consider the alternating method to consist of knapping both faces in the same sequence, interspersing core about-turns between each strike. There is a problem derived from the existing terminology and related to the difficulty of delimiting reduction stages (Bleed,
2002) and deconstructing reduction processes into strategies, sequences, series, methods or reduction systems. Assessing knapping complexity is a key component in explaining the cognitive changes observed in the evolution of stone technology, but there is a lack of consensus as to which criteria should be used (Stout et al., 2010, pp. 476). This makes comparing flaking strategies very difficult. Faced with this problem, which is not easy to solve, we aim to clearly define the different flaking actions and products used in this work as well as those used to make experimental comparisons. 2.1. The alternate strategy By ‘alternate strategy’ we refer to the process by which one face is partially or completely worked before moving on to the second face (Fig. 1). It is conceptually bifacial, but the hierarchical sequences employed are not. Alternate reduction can be performed with a mixture of methods. The most instinctive of these uses blow by blow removals, using the most highlighted ridges on each face. But in early reduction stages on rounded blanks, taking advantage of the previous removal ridges with sub-parallel or adjacent flaking (Pelegrin, 2005) provides an easier way to reduce the blank and to remove the cortex. Theoretically, only one or very few core turnovers are needed between the two sequences. As a result, in the alternate strategy there are two different production series or shaping sequences, referred to in this work as the ‘first alternate’ and ‘second alternate’. 2.2. The alternating strategy By ‘alternating sequence’, we refer to the process by which both faces are flaked in a single sequence, interspersing core about-turns between each strike (Inizan et al., 1995, 1999) (Fig. 1). Alternating involves core rotation and tilt rectification, but in addition, continuous core about-turns are performed throughout the sequence. During the knapping process it is necessary to change core position and flaking planes continuously, creating a continuous removal series that entails a different type of complexity in visual reconnaissance (the volume of each side), manual rectifications (continuous core aboutturns) and execution (hammer actions). The last scar is the primary hierarchical element because it determines the position and orientation of the core on the next strike. The concavity of the last scar forces the knapper to apply oblique percussions (to the right, to the left, and so on), which are needed to create a new platform for the next blow on the other face of the blank. In this sense, the alternating sequence involves an anticipatory strategy. 3. Materials and methods In this experimental work, we produced two different samples of alternate and alternating flakes. Different experimental series were created using thirty-five cobbles of flint and quartzite (Table 1). The flint nodules come from Zaragoza region (Spain). It is usually homogeneous and non-fissured raw material, and the cortex is perfectly suitable to strike. The quartzite pebbles come from Tagus River, near to Lisbon (Portugal). They are tough and display fine-grained and homogeneous raw material. The cobbles and nodules had regular shapes and elongated elliptical or spherical morphologies and some slab-like nodules with right-angled edges were also selected.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
3
Table 1 Summary of blank shapes (pebbles and cobbles), raw materials and applied flaking methods. Exper.
Morphology
Raw material
Mass (kg.)
Method
Blows
Complete Flakes
P1 P2 P3 P4 P5 P6 P7 Pil.1 Pil.2 Pr-2 N1 N2 N3 N4 N5 N7 N8 N11 N16 N17 N18 N19 N23 N24 N25 N28 N37 N36 N30 N29 N34 N31 N20 N9 N35 Total
Pebble Pebble Pebble Pebble Pebble Pebble Pebble Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule Nodule
Quartzite Quartzite Quartzite Quartzite Quartzite Quartzite Quartzite Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint Flint
1.74 3.04 1.92 2.46 2.44 2.44 2.28 2.20 1.41 _ 2.34 2.84 3.34 2.68 3.08 2.18 1.76 2.58 2.82 1.68 1.72 1.38 1.70 0.90 2.18 1.92 2.04 1.18 2.12 1.34 3.10 6.60 2.44 1.28 1.72
Alternating Alternating Alternate Alternate Alternating Alternate Alternating Alternate Alternate Alternate Alternating Alternating Alternating Alternating Alternating Alternating Alternating Alternating Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternate Alternating Alternating Alternating Alternate Alternating Alternating Alternating Alternating
13 33 31 38 28 22 16 17 17 34 32 21 28 35 43 26 19 27 36 16 13 37 9 9 15 18 29 27 33 21 36 17 46 26 31 899
11 35 29 35 25 21 17 15 15 33 34 17 29 30 37 26 20 28 34 14 12 35 9 9 13 18 30 27 28 20 38 18 46 26 29 863
Fig. 1. Alternate and alternating methods scheme.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
4
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
The cortex removal sequences were carried out by one of the authors (M.G.), using the free-hand percussion technique. Two stone hammers were used: P1, quartzite (620 g, density ~ 2.45 g/ cm3), for removals requiring tougher percussion (only the first removal in the reduction of each blank); and P2, a hard sandstone hammer (550 g, density: ~2.2 g/cm3) for all the other blows. The flakes were labelled, divided by sequences and all of the aspects involved were meticulously written down: flake number, sequence type, and blow direction were marked on the ventral face of each flake; starting blank zone, rotation direction, about-turns, hammerstone and other observations were also documented. We only prepared the platform when it was necessary in order to understand the platform preparation requirements for each method. The reduction was finished when the entire perimeter of the blank was flaked. As mentioned earlier, three sequences were recorded, ‘first’ and ‘second’ for the alternate reduction method and a single ‘alternating’ sequence for the alternating method. Blow direction was defined as to the left, to the right or perpendicular to the centre of the blank or longitudinal axis (from the knapper's point of view). 3.1. Useful recorded attributes Most of the attributes considered in our database are similar to those used in other experiments (Amick et al., 1988; Toth, 1988; Bradbury and Carr, 1995, 1999; Dibble et al., 2005; Braun et al., 2005, 2008; Marwick, 2008; Eren and Lycett, 2012) and both quantitative and qualitative variables were documented (Fig. 2): - Technical axis (TA). Distance from impact point to the distal end of the flake along the technological axis and perpendicular to the major platform axis. - Maximum width (MW). Maximum width of the flake oriented by the technical axis. Left maximum width (LMW). Maximum width axis is divided in two by the technical axis, we record the left segment. This variable allows us to observe the flake skewness and to approximate mass distribution in two dimensions.
- Left surface (LS). Surface in mm2 to the left of the technical axis. To record this variable, digital photographs were taken and then the surface was measured using ImageJ software (http://imagej. nih.gov/ij/). - Maximum platform width (PtMW). - Platform thickness (PtT). Platform thickness measured at the impact point position. - Platform left width (PtLMW). Distance from the impact point to the left end of the platform. - Platform type: cortical (C), plane (P) and faceted (F). - Butt shape. Seven morphological categories were created to record this variable (Fig. 3): A: flateconvex, B: helicoidal, C: scalene, D: convexeconcave, E: punctiform, F: inverted triangletrapezoid and G: dissymmetrical. - Previous percussion mark. In a preliminary experimental test, this singular feature was observed, and consists of small semicircular fractures at the opposite end of the impact point (Fig. 4). These fractures were caused by the percussion of the previous removal. Three categories were established: no presence (N), left (L) and right (R). - Cortical index. The percentage of dorsal cortex was recorded following the ‘eight-stage system’ (Roth and Dibble, 1998; Braun et al., 2008), in which 1 ¼ 100%, 2 ¼ 91e99%, 3 ¼ 71e90%, 4 ¼ 51e70%, 5 ¼ 31e50%, 6 ¼ 11e30%, 7 ¼ 1e10% and 8 ¼ 0%. - Cortex reserve type. The flakes were distributed into different cortex reserve categories using Toth's technological flake categories (TFC) (Toth, 1982, 1985). This system has normally been interpreted in relation to reduction intensity and flaking strategy studies: types IIII suggest primary flaking and IVVI reflect higher degrees of reduction (Toth, 1985), taking core size into account (Braun et al., 2005, 2008).
3.2. Statistical approach We tested for significant differences and associations in the quantitative and qualitative variables analysed. The distribution of
Fig. 2. Recorded attributes.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
5
Fig. 3. Butt shape type classification: A: flateconvex, B: helicoidal, C: scalene, D: convexeconcave, E: punctiform, F: inverted triangle-trapezoid and G: dissymmetrical.
the paired quantitative variables was tested using the ShapiroeWilk test for normality. The entire set of measurements displayed a non-normal distribution, so a non-parametric statistic ManneWhitney two-tailed U test was used. The ManneWhitney test checks for statistical differences in the medians of two independent samples with a null hypothesis of equal medians. As a nonparametric method, it can test for differences in distributions of any given shape, assuming that the tested groups share the same shape. Correlations between the categorical variables, attributes and methods were explored using a correspondence analysis (CA) for counted data. In order to determine whether the significant dif-
the alternating method and 359 (41.6%) using the alternate method (first sequence ¼ 229 and second sequence ¼ 130) (Table 2). Some of these flakes were broken during knapping, and only when the fractures were clear (easy to refit), were they refitted and analysed. Sometimes, several flakes (usually two), were generated in a single strike (56 cases), in this case we noted the order of each and they were considered individually. In addition, sometimes repeated percussions on the same face were needed in the application of the alternating method; this was indicated by a scoring of ‘r’, which occurred in 120 cases. The data descriptive statistics data are provided in Table 3.
Table 2 Blank and flake production conducted in the experiment.
Alternate Alternating
Blanks
Total Flakes
Flakes/blank
Flakes/kg.
Sequence
16 19
359 504
22.5 26.5
11.05 12.2
1st ¼ 229; 2nd ¼ 130 Alt: 504
Table 3 Descriptive statistics of experimental database. 1st Alternate
TA MW LMW Total area Left area Thickness (prox) Thickness (med) Thickness (dist) PtMW PtMTI PtLMW
1st & 2nd alternate
2nd Alternate
Alternating
Mean
sd
CV
Mean
sd
CV
Mean
sd
CV
Mean
sd
CV
47.15 47.34 23.2 1741.12 828.98 9.22 7.75 4.82 20.52 6.44 10.63
19.53 19.02 10.39 1181.3 596.89 6.64 4.57 2.82 11.1 4.08 7.35
0.41 0.40 0.45 0.68 0.72 0.72 0.59 0.59 0.54 0.63 0.69
47.22 47.11 23.03 1720.28 821.98 9.06 7.64 4.75 20.47 6.22 10.3
19.54 18.86 10.31 1167.48 592.01 6.56 4.45 2.71 11.12 3.72 6.89
0.41 0.40 0.45 0.68 0.72 0.72 0.58 0.57 0.54 0.60 0.67
47.44 47.64 23.31 1767.8 841.41 9.24 7.74 4.85 20.9 6.5 10.7
19.83 19.24 10.49 1223.52 613.65 6.58 4.54 2.82 11.20 4.05 7.31
0.42 0.40 0.45 0.69 0.73 0.71 0.59 0.58 0.54 0.62 0.68
49.31 47.14 22.93 1839.97 883.71 8.82 7.67 4.79 17.89 5.43 9.45
19.59 19.53 10.76 1279.24 669.71 5.80 4.65 2.99 10.40 3.70 7.08
0.40 0.41 0.47 0.70 0.76 0.66 0.61 0.62 0.58 0.68 0.75
ferences observed in the analysis of the methods' characteristic features are statistically causal we tested for their discriminant and predictive power. We considered that the combination of diagnostic features could be checked through the use of a classificatory method to explore its robustness, load and applicability. For this purpose, we performed a binomial logistic regression, a generalised linear model that checks for causal relations when the dependent variable (alternate and alternating in this case) is binomial. 4. Results During our experiments, hand-controlled percussions generated 863 whole flakes: 504 (58.4%) of them were produced using
The results of the ManneWhitney for equal medians applied to the metric variables test are summarised in Table 4. No differences were found between the alternate and alternating methods in some of the most common recorded measurements like flake width and flake area. The proximal, medial and distal thicknesses did not yield significant differences either. However, significant differences were found in the length of the technical axis, with the alternating flakes longer than the alternate flakes. The platform maximum thickness, the maximum thickness from the impact point, and the left platform maximum width differences indicate that alternate flakes have thicker and broader platforms than the alternating flakes.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
6
M. Guardiola et al. / Quaternary International xxx (2015) 1e8 Table 4 ManneWhitney test results for differences between alternate and alternating metrical variables. * Remarks significant differences. p(same) Technical Axis Maximum Width Left Maximum Width Total Area Left Area Proximal Thickness Medial Thickness Distal Thickness Platform Maximum Width Platform Maximum Impact Thickness Platform Left Maximum Width
0.01256* 0.4544 0.0556 0.1607 0.3373 0.2195 0.2366 0.4094 0.00* 0.00* 0.00*
In the attribute associations, the CA analysis was initially performed in multiple ways, but for easier visualisation and interpretation of the results, they are presented using a simple CA analysis performed variable by variable. In order to more efficiently show the category associated with the knapping method, the first and second alternate sequences have been considered independently. Using the exploratory power of the CA, several significant and diagnostic associations between the categories of each variable and the three sequences were found. The cortex distribution on the dorsal faces of the first and second alternate flakes and the alternating flakes exhibits different patterns of cortical index according to the distribution of the categories. While alternating flakes display a heterogeneous pattern of middle to high range cortical ratios (from 30 to 90%), first alternate and second alternate flakes are associated with very low and very high cortex presence, respectively (Fig. 5, A). The platform variable analysis shows clear associations of flat platforms with c, e, b shapes for the alternating method. The alternate sequences are strongly associated with cortical f and g shapes in the case of first, and facetted a and d butt shapes for the second (Fig. 5, B and C). Interestingly, the occurrence of previous
Fig. 4. Origin and location of the previous percussion mark related to a later removal.
percussion marks, when it happens, is highly and almost exclusively associated with alternating flaking (Fig. 5, D). The exploration of the occurrence of synthetic TFC discretisation related to the three sequences correlates type II flakes with the alternate first sequence, type IV flakes with the second alternate and types V and VI with the alternating sequence (Fig. 5, E). In order to test the explicative power of each variable for the alternate and alternating methods, we transformed them and analysed them together in a logistic binomial regression. The Nagelkerke r2 indicator shows that 81.9% of the variance observed in the products of each knapping method can be explained through the analysed patterns (Nagelkerke r2 ¼ 0.819) (Table 5). The model's robustness is shown in the classificatory table with a balanced percentage of correctly classified cases of 93.4%. The method also indicates that the variables with the highest incidence in the regression equation are the complete set of platform quantifications (p ¼ 0.00; 0.013; 0.03; 0.008), the cortical index (p ¼ 0.04) Table 5 Classification table of binomial logistic regression. Alternate Alternate 167 Alternating 19 Overall percentage
Alternating
Correctly classified %
15 318
91.8 94.4 93.4
and platform type (p ¼ 0.00). 5. Discussion and conclusions We experimentally compared alternate and alternating flakes in early stages on rounded blanks, and we provide several distinctive traits with which to distinguish the products of each method. Technologically, and from an experimental point of view, despite its theoretical importance to bifacial reduction, little research has been done on the identification of the alternating method in relation to handaxe production. Archaeologically, being able to identify methods from flakes is very useful, as cores are frequently depleted and do not reflect the technological features of the reduction process. All of these factors have an impact on bifacial technologies. Adaptation to the shapes of different raw materials can be better understood by identifying alternating method flakes. Some of the morphometric traits of the experimentally produced flakes exhibit differences between the alternate and alternating methods. Firstly, the alternating method flakes are longer than the alternate flakes. Although greater invasiveness has been reported with the alternating method (Moore, 2003; White and Ashton, 2003), it is difficult to compare the two strategies because there are substantial differences between the first and second alternate sequence. Based on the shape of the initial edge, the level of invasiveness in the first and second alternate sequences is extremely variable. First alternate removals tend to produce short thick flakes. In contrast, second alternate removals lead to highly variable flake lengths, depending on the percussion conditions. Platform attributes are the best indicators of the alternating method and are identifiable from flakes, especially with regard to the impact point position (normally lateralised) and the presence of the last percussion mark (~25% of alternating flakes; <1% in alternate flakes). Compared to the alternate method, the alternating method yields flatter platforms, normally with scalene, punctiform and helicoidal-shaped butts. In terms of the technological flake categories (Toth, 1982, 1985), alternating method produce a great part of the type V flakes. In fact, the alternating method produces type V flakes from the beginning of the sequence, specifically, from the second removal. Therefore, related to reduction intensity and
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
7
Fig. 5. Correspondence analyses of categorical variables and knapping sequence type: A: Cortex Index, B: Butt shape classification, C: Platform type, D: Previous percussion mark presence, E: Technological flake categories.
levels of transport (Kimura, 2002), the alternating flaking method might produce a distortion of this model. Identifying the alternating method can provide valuable information about technological adaptation and cognitive evolution. To this end, we have defined some recognisable features contextualised in the initial reduction steps on a very common blank shape. In addition to our experiments with rounded blanks, further research needs to be conducted, primarily on larges flakes and knapper variability. In Acheulean bifacial technologies, the alternating method plays a fundamental role in subsequent core
technologies, it is important in Levallois strategies and in blade technologies. In all of them, blank rounded edges (semicircular cross-sections) are one of the most common blank shapes and the alternating flaking method provides an easy and systematic solution by which to reduce them. Acknowledgements This research has been partially supported by the projects 2014SGR-899 and 900 (AGAUR-Generalitat de Catalunya) and,
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039
8
M. Guardiola et al. / Quaternary International xxx (2015) 1e8
HAR2013-41197-P and CGL2012-38434-C03-03 (Ministerio de Economía y Competitividad). J.I. Morales has been funded by a predoctoral research fellowship (FI-B2-2014) from the AGAUR of Generalitat de Catalunya. We also want to thank to an anonymous reviewer of the present version as well as two others who reviewed a previous copy. They have improved our work. Finally, we want to thank to the organizers of the European Acheulean Congress, le ne Moncel and Danielle Schevre for hosting our Marie-he research.
References Amick, D.S., Mauldin, R., Tomka, S.A., 1988. An evaluation of debitage produced by experimental bifacial core reduction of a Georgetown Chert Nodule. Lithic Technology 17 (1), 26e37. Andrefsky, W.A., 1998. Lithics: Macroscopic Approaches to Analysis. Cambridge University Press, Cambridge. Ashton, N., White, M., 2003. Bifaces and raw materials: flexible flaking in the British early Paleolithic. In: Soressi, M., Dibble, H.L. (Eds.), Multiple Approaches to the Study of Bifacial Technologies. University of Pennsylvania Museum of Archaeology and Anthropology, Philadelphia, pp. 109e123. n a la talla de utillaje Baena, J. (Ed.), 1998. Tecnología lítica experimental. Introduccio rico. British Archaeological Reports International Series, 721. Oxford. prehisto Bleed, P., 2002. Obviously sequential, but continuous or staged? Refits and cognition in three late Paleolithic assemblages from Japan. Journal of Anthropological Archaeology 21, 329e343. Bradbury, A.P., Carr, P.J., 1995. Flake typologies and alternative approaches: an experimental assessment. Lithic Technology 20, 100e115. Bradbury, A.P., Carr, P.J., 1999. Examining stage and continuum models of flake debris analysis: an experimental approach. Journal of Archaeological Science 26, 105e116. Braun, D.R., Tactikos, J.C., Ferraro, J.V., Arnow, S.L., Harris, J.W., 2005. Flake recovery rates and inferences of Oldowan hominin behavior: a response to Kimura 1999, 2002. Journal of Human Evolution 48, 525e531. Braun, D.R., Tactikos, J.C., Ferraro, J.V., Harris, J.W., 2008. Oldowan reduction sequences: methodological considerations. Journal of Archaeological Science 38, 2153e2163. Burton, J., 1980. Making sense of waste flakes: new methods for investigating the technology and economies behind chipped stone assemblages. Journal of Archaeological Science 7, 131e148. Callahan, E., 1979. The basic biface knapping in the eastern fluted points tradition: a manual for flintknappers and lithic analysts. Archaeology of Eastern North America 7 (1). Crompton, R.H., Gowlett, J.A.J., 1993. Allometry and multidimensional form in Acheulean bifaces from Kilombe, Kenya. Journal of Human Evolution 25, 175e199. de la Torre, I., 2011. The Early Stone Age lithic assemblages of Gadeb (Ethiopia) and the developed Oldowan/early Acheulean in East Africa. Journal of Human Evolution 60, 768e812. Delagnes, A., Roche, H., 2005. Late Pliocene hominid knapping skills: the case of Lokalalei 2C, West Turkana, Kenya. Journal of Human Evolution 48, 435e472. Dibble, H.L., Schurmans, U.A., Iovita, R.P., McLaughlin, M.V., 2005. The measurement and interpretation of cortex in lithic assemblages. American Antiquity 70 (3), 545e560.
Eren, M.I., Lycett, S.J., 2012. Why Levallois? A morphometric comparison of experimental “preferential” Levallois flakes versus debitage flakes. PLoS ONE 7 (1), E29273. http://dx.doi.org/10.1371/journal.pone.0029273. Goren-Inbar, N., Grosman, L., Sharon, G., 2011. The technology and significance of the Acheulian giant cores of Gesher Benot Ya'aqov, Israel. Journal of Archaeological Science 38, 1901e1917. Inizan, M.L., Reduron-Ballinger, M., Roche, H., Tixier, J., 1995. Technologie de la e. CREP(4), Nanterre. pierre taille Inizan, M.L., Reduron-Ballinger, M., Roche, H., Tixier, J., 1999. Technology and Terminology of Knapped Stone. CREP (5), Nanterre. Jones, P.R., 1994. Results of experimental work in relation to the stone industries of Olduvai Gorge. In: Leakey, M.D., Roe, D.A. (Eds.), Olduvai Gorge, vol. 5. Cambridge University Press, Cambridge, pp. 254e298. Kimura, Y., 2002. Examining time trends in the Oldowan technology of beds I and II, Olduvai Gorge. Journal of Human Evolution 43, 291e321. Marwick, B., 2008. What attributes are important for the measurement of assemblage reduction intensity? Results from an experimental stone artefact assemblage with relevance to the Hoabinhian of mainland Southeast Asia. Journal of Archaeological Science 35, 1189e1200. Moore, M.W., 2003. Australian aboriginal biface reduction techniques on the Georgia River, Camooweal, Queensland. Australian Archaeology 56, 22e34. flexions sur le comportement technique. La signification culPelegrin, J., 1985. Re ge, pp. 72e91. turelle des industries lithiques, Actes du Colloque de Lie Pelegrin, J., 1993. A framework for analysing prehistoric stone tool manufacture and a tentative application to some early stone industries. In: Berthelet, A., Chavallion, J. (Eds.), The Use of Tools by Human and Non-human Primates. Clarendon Press, Oxford, pp. 302e314. Pelegrin, J., 2005. Remarks about archaeological techniques and methods of knapping: elements of cognitive approach to stone knapping. In: Russon, A.E., Begun, D.R. (Eds.), Stone Knapping: the Necessary Conditions for a Uniquely Hominin Behavior. McDonald Institute for Archaeological Research, Cambridge, pp. 159e169. Roth, B.J., Dibble, H.L., 1998. Production and transport of blanks and tools at the French Middle Paleolithic site of Combe-Capelle Bas. American Antiquity 63, 47e62. Semaw, S., 2000. The world's oldest stone artifacts from Gona, Ethiopia: their implications for understanding stone technology and patterns of human evolution between 2.6e1.5 Million years ago. Journal of Archaeological Science 27, 1197e1214. Stout, D., Toth, N., Schick, K., Chaminade, T., 2008. Neural correlates of Early Stone Age toolmaking: technology, language and cognition in human evolution. Philosophical Transactions of the Royal Society 363, 1939e1949. Stout, D., Semaw, S., Rogers, M.J., Cauche, D., 2010. Technological variation in the earliest Oldowan from Gona, Afar, Ethiopia. Journal of Human Evolution 58, 474e491. Toth, N.P., 1982. The Stone Technology of Early Hominids at Koobi Fora, Kenya: an Experimental Approach (Ph. D. dissertation). University of California, Berkeley. Toth, N., 1985. Oldowan reassessed: a close look at early stone artefacts. Journal of Archaeological Science 12, 101e120. Toth, N., 1988. Behavioral inferences from early stone artifact assemblages: an experimental model. Journal of Human Evolution 16, 763e787. White, M.J., 1998. On the significance of Acheulean biface variability in southern Britain. Proceedings of the Prehistoric Society 64, 15e44. White, M.J., Ashton, N.M., 2003. Lower Paleolithic core technology and the origins of the Levallois method in North-Western Europe. Current Anthropology 44 (4), 598e609. Whittaker, J.C., 1994. Flintknapping: Making and Understanding Stone Tools. University of Texas Press, Austin.
Please cite this article in press as: Guardiola, M., et al., From blunt to cutting: Distinguishing alternating method flakes in early stages on rounded blanks, Quaternary International (2015), http://dx.doi.org/10.1016/j.quaint.2015.08.039