History of prepared-core processing
Although there had been different types of tools created prior to this technique, mostly that of handaxes and cleavers, this new technique was a great advancement for early man. Believed roughly to have been first used 200,000 years ago this technique involved a removal of flakes from a piece of stone to achieve desired shape and thickness. The use of this technique is believed to be a significant change in culture and shows an increasing growth of cognitive ability, as one that is using this method must be able to imagine the end product and maintain that image while conditioning the stone to the desired shape and end result tool.
Methods for reducing cores
Centripetal core reduction
Centripetal or radial core reduction technology encompasses a substantial range of archaeological variability, using pieces of raw material with natural convexities just as well as heavily prepared centripetal cores. The technique is defined by the characteristic direction of percussion from the perimeter of the core towards the centre, hence the term "centripetal". The perimeter of the circular to oval-shaped core served as a platform for striking off flakes, blades and points, which further shaped the face of detachment.
Several technological criteria characterise the definition of the centripetal cores.
- The primary reduction surface is shaped such that the morphology of the product is a function of the lateral and distal convexities of the surface, which serve to guide the shock wave of each flake.
- Centripetally worked cores may possess either one or two faces of detachment and striking platforms can vary from unprepared, cortical to faceted chapeau de gendarme ("cocked hat") shapes.
- Centripetally-reduced cores were made via hard hammer percussion, which was probably necessary when attempting to work steep platform angles.
Centripetal core reduction techniques are found in most Middle Paleolithic assemblages known from Africa and Eurasia. There are a number of different methods of technological actions involving centripetal core reduction, the most extensively documented of which is the so-called "Levallois" technology.
Levallois Core Reduction
The recognition of Levallois as a distinct core reduction strategy dates to the late 19th Century. The term was used to describe specific flakes with certain surface attributes that were recovered during that period in northern France. These early descriptions were purely typological and based on the morphology of the flake products themselves. However, there was never a great deal of consensus among scholars, which typological attributes could be used to identify Levallois products. Gradually, more and more emphasis was put on the idea that Levallois flakes were the products of a particular method or process of production. Indeed, F. Bordes emphasised that Levallois was essentially a method and not a particular product. However, the shape and character of a Levallois blank is also thought to be "predetermined" by the elaborate Levallois core preparation process . While a shape control system undoubtedly exists for the Levallois cores, there remain a number of significant problems. Indeed, how applied force will propagate through a specific core is determined by a number of variables and not only by the will or the desire of the Middle Palaeolithic flintknapper. Fracture mechanic variables include size, shape and internal structure of a particular flint nodule, but also the mass and resilience of the hammer stone and finally the angle and force of the blow and the shape of the core's striking platform. Given the imprecision of hand-eye co-ordination , a rather high probability for only partial core reduction success is very real. Not only are there a number of significant problems with defining Levallois on the basis of predetermined blanks, but there is also considerable disagreement over what set of attributes should be used to characterise a Levallois product. Furthermore, it has also been demonstrated that very different core reduction strategies can produce seemingly diagnostic Levallois blanks.
The generally accepted criteria for a prepared-core to be categorized as a Levallois core are as follows: (1) core is organized in terms of two intersecting flaking surfaces; (2) the flaking surfaces have a hierarchical relationship, striking platform and primary reduction surface; (3) the shape of the primary reduction surface is such that the flake morphology is predetermined; (4) the fracture plane is sub-parallel to the intersection of the two previously mentioned surfaces; and (5) the striking platform is adjusted for removal of flakes that are parallel to the fracture plane, this is usually done through retouch and faceting. 
The fact of the matter is that the Levallois Method is a term, which has a different meaning according to context. The Levallois Method concerns the productivity of a Levallois surface, which can be exploited according to a "lineal" or "preferential" method with the production of a single Levallois product, or which can be exploited according to a recurrent method with the production of several Levallois products. The lineal or preferential Levallois method corresponds best to the classic definition of Levallois (see the right illustration). The recurrent Levallois method can be unipolar, with only one striking platform, bipolar, with opposed striking platforms, or centripetal, with two or more adjacent striking platforms. The unipolar, bipolar or centripetal recurrent Levallois technique is marked by the detachment of a series of large Levallois flakes, such that the preceding removals ready the surface for the subsequent ones, thus eliminating the need for extensive repreparation. The centripetal recurrent Levallois technique also includes pseudo-Levallois points and sometimes side-struck pieces as well. In some contexts, on the other hand, the Levallois Method denotes the specific organisation of scars and ridges on a Levallois surface, with one method focusing on the production of flakes and the other method focussing on the production of points. Contrary to Boëda , Van Peer further concludes that the recurrent bipolar and centripetal Levallois methods do not exist. Only the notion of one preferential striking platform is the most essential characteristic of the true Levallois reduction strategy. Van Peer also claims that a separate Levallois method for blade reduction does not exist either.
There are thus serious problems and implications of the now widely accepted processual definition of the Levallois method. The presence of the Levallois criteria on a blank or a core do no longer necessarily guarantee the real Levallois character of a specific core reduction. So, we should not only study flakes and cores, since blanks and cores cannot be identified as Levallois by their own morphology. We should thus also try to reconstruct the dynamic reduction process by refitting. However, many Middle Palaeolithic assemblages do not allow these extensive reconstructions. So, analysis based on discarded cores and flakes can be problematic, since most of the flaking patterns that preceded a core's discard are simply not observable on them. But, if we cannot be sure of the status of a seemingly diagnostic "Levallois" product, can we then be sure of the Levallois or non-Levallois intent of the core reduction process? The centripetal or radial core reduction scheme, which is used in this study, functions independently of the other characteristics, which define the Levallois method. In Levallois and non-Levallois centripetal core reduction, flakes are struck from the perimeter of the core towards the centre. This is the only attribute, which both reduction strategies have in common. Three major variants within the centripetal core reduction strategy are distinguished: (1) unipolar [one striking platform], (2) bipolar [two opposed striking platforms] and (3) centripetal [two or more adjacent striking platforms]. The preparation of these cores is thus always centripetal. However, the removal of the major flakes can be unipolar, bipolar or centripetal. The term Levallois only refers to specimens, which clearly bear evidence of the production of a limited number of large central-positioned flakes on the core, which have been struck off from individual proximal striking platforms, and which were usually prepared several times. Levallois thus only refers to preferential or linear (recurrent or not) centripetal core reduction. Only a very small percentage of the centripetal cores and flakes will thus be classified as true Levallois products, by virtue of the presence of a single flake scar covering more than 60% of the reduction surface . Core reduction strategies are by necessity dynamic processes, since they must manage an ever-decreasing amount of raw material. Variability in centripetal core reduction strategies can be explained by dynamic adjustments made during the process of core reduction. Indeed, flintknappers always had to respond to the often-unpredictable nature of stone fracture.
- Bringmans, P.M.M.A., Vermeersch, P.M., Gullentops, F., Groenendijk, A.J., Meijs, E.P.M., de Warrimont, J.-P. & Cordy, J.-M. 2003. Preliminary Excavation Report on the Middle Palaeolithic Valley Settlements at Veldwezelt-Hezerwater (prov. of Limburg). Archeologie in Vlaanderen - Archaeology in Flanders 1999/2000 VII: 9-30.
- Bringmans, P.M.M.A., Vermeersch, P.M., Groenendijk, A.J., Meijs, E.P.M., de Warrimont, J.-P. & Gullentops, F. 2004. The Late Saalian Middle Palaeolithic "Lower-Sites" at Veldwezelt-Hezerwater (Limburg - Belgium). In: Le Secrétariat du Congrès (eds), Acts of the XIVth UISPP Congress, University of Liège, Belgium. September 2-8, 2001. Section 5: The Middle Palaeolithic. Oxford. British Archaeological Reports (BAR) International Series 1239: 187-195.