Studies of lithic raw material play an important role in exploring prehistoric behavioral strategies because understanding how prehistoric people used lithic raw material can help us to explore their technological organization linked to mobility and land use (e.g., Binford, 1979; Kuhn, 1995). In numerous studies of lithic raw material, an issue of raw material “quality” has often been discussed. The “quality” of raw material can be an important basis in the selection of raw material. Besides, the raw material “quality” is used for establishing the models that quantify attractiveness of raw material sources and estimate kinds of lithic tools produced (e.g., Andrefsky, 1994, 2009; Wilson, 2007; Browne and Wilson, 2011). On the other hand, most of the discussions on raw material “quality” were qualitative. Flaked stone tools are often made of sedimentary and volcanic rocks, such as obsidian and chert, which are rich in silica content and are microcrystalline and glassy. A common recognition among archaeologists is that these properties increase the “quality” of lithic raw material (e.g., Bradbury et al. 2008; Sharon, 2008).
In the Levantine Paleolithic, chert (flint) is mainly used as raw material of flaked stone tools. Recent raw material studies actively adopted field surveys for outcrops and analyses of lithic artifacts and raw material samples collected at the outcrops (Finkel et al. 2016, 2018, 2019, 2020b, 2022; Groucutt et al. 2017; Parow-Souchon et al. 2021). Increasingly, the identification of raw material sources was conducted through various geochemical studies (e.g., INAA, the measurement of 10Be, ICP-MS, and petrographic analysis) of chert samples from outcrops and archaeological sites (Julig et al. 2007; Boaretto et al. 2009; Ekshtain et al. 2014, 2017, 2019; Ekshtain and Tryon, 2019; Agam, 2020; Agam et al. 2020; Bellar et al. 2020; Finkel et al. 2020a; Ekshtain and Zaidner, 2022; Bellar, 2023). In addition, Minimal Distance Analysis was performed by using ArcGIS (Parow-Souchon and Purschwitz, 2020). Some studies mentioned the raw material selectivity through the classification of chert’s macroscopic aspects and the analyses of the quality and use-wear of chert artifacts (Delage, 2007; Hovers, 2009; Wilson et al. 2016; Agam, 2020; Agam and Zupancich, 2020; Assaf, 2021; Agam et al. 2022). The blind test about the macroscopic classification of chert types was also performed (Agam and Wilson, 2018). A few studies identified the sources of Paleolithic obsidian artifacts (Frahm and Hauck, 2017; Frahm and Tryon, 2019). From a wider perspective, some studies discussed lithic provisioning strategies at the Upper Paleolithic (Kuhn, 2004) and the Middle Paleolithic (Henry, 2011; Varoner et al. 2022). In this way, lithic raw material studies are progressing in various themes, focusing on the identification of raw material sources. The “quality” of chert raw material is usually evaluated on the basis of macroscopic characteristics. More specifically, archaeologists often evaluated that fine-grained (textured) chert is high quality, and coarse-grained (textured) is low quality (Goring-Morris and Davidson, 2006; Delage, 2007; Neeley, 2007; Marder and Goring-Morris, 2020; Parow-Souchon and Purschwitz, 2020).
In our previous study, we analyzed chert artifacts in some Middle and Upper Paleolithic assemblages from the Jebel Qalkha area in southern Jordan (Suga et al. 2022). We classified the lithic artifacts into several chert types based on the quantitative criteria with reference to previous studies (Henry, 1995; Henry et al. 2014; Henry and Mraz, 2020) and presented the variety of chert. No significant difference was observed in the frequency of chert types between the Late Middle Paleolithic (LMP) and the Initial Upper Paleolithic (IUP), but the relative frequency of fine-grained chert increased in the Early Upper Paleolithic (EUP). This change coincided with the increase in bladelets, and we considered the possibility that the selection of chert types changed according to lithic production technology and forms of the products. However, the exact reasons for the chert raw material change remained unclear.
This study aims to quantify the “quality” of chert used as lithic raw material. We examine how the differences in macroscopic appearances that are conventionally used as criteria for the chert “quality”, are related to some quantitative attributes in hardness involved in flaking lithics. Based on the result, we discuss the reasons for the changes in raw material use from LMP to EUP in southern Jordan.
Background of fracture predictability
In archaeology, the “quality” of lithic raw material is often evaluated on the degree of its easiness to be flaked. Several studies use the term “fracture predictability” based on the properties of lithic raw material that is prone to be flaked as predicted (Braun et al. 2009; Eren et al. 2014; Moník and Hadraba, 2016; Caruana and Mtshali, 2018; Egeland et al. 2019). The flakes knapped from lithic raw materials with the high fracture predictability (high quality) tend to have sharp edges suitable for knives or scrapers (Harmand, 2009; Terradillos-Bernal and Rodríguez-Álvarez, 2017). Some previous studies from the viewpoint of fracture predictability examined several physical attributes including brittleness, elasticity, isotropy, homogeneity, continuity, and granularity (Cotterell and Kamminga, 1987; Luedtke, 1992; Whittaker, 1994; Inizan et al. 1999; Sharon, 2008; Eren et al. 2011; Goldman-Neuman and Hovers, 2012; Garvey, 2015; Monik and Hadraba, 2016; Rodríguez-Rellán, 2016; Namen et al. 2022). These attributes can be classified into two group: a group of structural properties (isotropy, homogeneity, continuity and granularity) and a group of mechanical properties (brittleness and elasticity).
As for the structural properties, the high fracture predictability was linked to some petrographic characteristics, such as little or no crystalline macrostructure, few impurities, and an overall small average crystal size (Luedtke, 1992; Domanski et al. 1994; Braun et al. 2009; Eren et al. 2014; Egeland et al. 2019). These characteristics are considered to facilitate fracture propagation inside the rock with little interference and thus increase the predictability of flaking, resulting in the feeling of “easiness” in lithic production (Cotterell et al. 1985; Cotterell and Kamminga, 1987; Brantingham et al. 2000). Moreover, some studies approached the fracture predictability through petrographic analyses by observing macroscopic attributes and microscopic features in thin section (Brantingham et al. 2000; Stout et al. 2005; Sherwood et al. 2018). These studies aimed to quantify the fracture predictability by calculating the average size of phenocrysts and the ratio of impurities in flaking removals.
As for the mechanical properties, there are many attempts to quantify the fracture predictability of heat-treated raw material by measuring mechanical properties (Domanski and Webb, 1992; Domanski et al. 1994; Domanski et al. 2009; Yonekura, 2010; Schmidt et al. 2012, 2019; Zhou et al. 2014; Mraz et al. 2019; Moník et al. 2021). Some studies explained that the heat-treatment of silica rocks increase their quality due to the formation of new Si–O–Si bonds that make the fracture path less meandering (Schmidt et al. 2012, 2019). In addition to heat treatment, there are some studies comparing mechanical properties between different kinds of lithic raw materials (Webb and Domanski, 2008; Tsobgou and Dabard, 2010; McPherron et al. 2014; Moník and Hadraba, 2016; Rodríguez-Rellán, 2016; Namen et al. 2022). These studies mainly measured strength, elasticity and toughness as key indicators. These mechanical properties are used not only in petrology but also in materials science. However, the discussion of how these mechanical properties influence the lithic production is still fragmentary. The following section summarizes current archaeological views on strength, elasticity, and toughness and their application to lithic raw materials.
Strength Strength is the absolute value of fracture stress when objects break down. Lithic raw materials with high strength can be useful in its exploitation for producing flakes as it prevents shattering of flakes’ striking platforms (Doelman et al. 2001). On the other hand, a previous study pointed out potential fluctuations in the strength value of brittle materials with internal cracks and indicated an importance of using appropriate kinds of tests and sufficient numbers of rock samples (Tsirk, 2014). More specifically, many studies employed Compressive or Tensile strength as parameters of lithic raw material strength (Domanski et al. 1994; Webb and Domanski, 2008; Yonekura and Suzuki, 2009; Zhou et al. 2014).
Elasticity Elasticity is the characteristic of objects returning to the original form after their plastic deformation. Lithic raw materials behave elastically at a macroscopic scale, and elasticity is a useful index that represents flexibility. For example, less elastic materials are stiffer while more elastic ones can be called more flexible. Previous studies quantified elasticity by measuring Young’s modulus (Domanski et al. 1994; Schmidt et al. 2019; Moník et al. 2021; Namen et al. 2022). In addition, the higher Young`s modulus is, the more resistant to deformation a rock is (Luedtke, 1992; Braun et al. 2009).
Toughness Toughness is the degree of resistance against external forces and breakage, and this characteristics is used for brittle materials with internal cracks. Previous studies quantified toughness by measuring Fracture toughness (KIC) (Domanski and Webb, 1992; Domanski et al. 1994; Webb and Domanski, 2008; Domanski et al. 2009; Schmidt et al. 2019; Moník et al. 2021; Namen et al. 2022). Raw materials of flaked stone tools, such as obsidian and chert, are similar to glass, and they often have microscopic cracks inside the rocks. When these internal cracks receive certain load (fracture stress), they induce a sudden breakage of the material. Fracture toughness is considered a useful parameter to evaluate raw materials of flaked stone tools (Tsirk, 2014).
Geological studies suggest that the above three mechanical properties are correlated mutually in general (Sachpazis, 1990; Zhang, 2002; Yasar and Erdogan, 2004; Aydin and Basu, 2005). In addition, the degree of mechanical properties can be discussed in relation to the petrographic characteristics (Namen et al. 2022). Here we summarize some previous studies’ predictions about how the mechanical properties of rocks influence lithic flaking. Firstly, high strength and elasticity make lithic raw materials resistant and hard to strain or deform. This facilitates the propagation of fracture forces and thus allows for predictable flaking (Braun et al. 2009; Egeland et al. 2019). Secondly, low fracture toughness makes rocks to be flaked with less fracture forces, facilitating the preparations of core ridges and platforms (Doelman et al. 2001). On the other hand, when fracture toughness (KIC) is too large, the raw material is too tough to work (Tsirk, 2014). These statements are relevant to the properties of fracture predictability mentioned above. The lithic grade scale proposed by Callahan (1979) suggests that lithic raw materials of better workability (high fracture predictability) have large values of elasticity and modest values of toughness (KIC).
Based on the previous studies, we estimate that lithic raw materials with high fracture predictability are rocks with high strength and elasticity but low toughness. Such rocks rebound without deformation when they are impacted under their fracture strength, but they suddenly break when they receive a force larger than their fracture strength. These characteristics indicate that brittle rocks are suited for the production of flaked stone tools. To evaluate these mechanical properties quantitatively, this study focuses on two types hardness, which will be explained below.