Levallois method
The Levallois method is a complex flake procedure that, by means of a special preparation of the upper face of the core (and, optionally, its periphery and its percussion platform) it is possible to preconceive or predetermine, totally or partially, the shape and size of the flake, before it has been extracted. In this way, lithic carving follows specific steps that give rise to very specific formal results.
History
This lithic procedure gets its name from the fact that the first discoveries took place in the 19th century in the surroundings of Levallois-Perret, a French city in the department of Hauts-de-Seine. It must be emphasized that the Levallois concept has varied greatly from then until recent times: at first the default extraction of the discoid was not distinguished, for which reason both were generically referred to as “Mousterian nuclei”. Henri Breuil was the first to use the term Levalloisien, although the emphasis was still placed on the morphological aspects of both the flakes and the cores, which were described from a merely formal point of view (silhouette, faceted heel...). Despite the important studies by V. Commont, they were interpreted biasedly, which led to discriminating various cultures of a Levallois character or, more properly, "Mustero-Levalloisians" in the Middle Paleolithic of the Mediterranean basin and the Near East based on the faceted character of the heel and not in the very essence of the Levallois idea.
François Bordes focused his interest on the concept of flake predetermination by means of a prior specific preparation, insisting on using the noun Levallois instead of the adjective levalloisiense, as he wanted to emphasize that it was a method to produce flakes, and not a product in itself, nor a specific culture. He likewise already separated some variants and relegated to the background certain accessory or incidental aspects that were not technologically decisive:
«The default form of a lasca (it's him) only valid criterion of the Levallois technique»F. Bordes, 1953, page 226
Variants
Within the Levallois concept, specialists distinguish various models depending on both the product to be obtained and the strategy followed for it. Thus, on the one hand, Levallois nuclei for flakes are distinguished from those intended to obtain blades or those that produce points. On the other hand, a difference is made between the nuclei of repetitive extractions, that is to say, those that allow obtaining several lacquered products with a single preparation, from those of preferential extractions, that is, those in which a single piece is obtained, requiring a new preparation every time you want to obtain another piece. Finally, the nuclei in which the exploitation is centripetal are also distinguished from those that have it parallel or polarized.
| STRATEGIES | Preferential extraction | Repeated withdrawal |
|---|---|---|
| Centripetal extract | Preferential Levallois Extraction | Repetitive centripetal extract |
| Parallel extraction | Levallois tip extraction | Levallois leaves extract |
However, differences persist among specialists in the way of seeing the Levallois concept or idea, both regarding the variants that can be included within the base method, as well as the attributes that the products obtained in this way must have. In fact, experiments in carving and reassembly of archaeological pieces are showing that there are very different carving strategies with which practically equivalent products can be obtained.
Obtaining Levallois flakes
The Levallois flake, in the strict sense, can be obtained by different methods that, basically, are separated into two categoriesː one that consists of preparing the core to obtain a large predetermined flake, making the most of its exploitation surface and achieving a stricter control of the product at the cost of a higher energy expenditure; and the obtaining of a more or less numerous series of predetermined flakes with a single preparation, achieving greater energy savings, but losing some control over the product and obtaining smaller pieces in relation to the core-support.
The classical method: preferential extraction
The Levallois nucleus for a preferential flake is the classical prototype. The procedure followed begins, in this case, by choosing an edge-support: a slightly elongated and flattened rounded pebble on which a plane of percussion that will serve to peel the main face of the nucleus by means of centripetal lifting. If necessary, a percussion platform is also prepared, generally located at one of the narrow ends of the support, to be able to strike in the same direction as its elongation axis. From this percussion plane, a large Levallois flake is extracted with a single blow, which necessarily has to be accurate and directed towards the center of the nucleus, the shape of which has been preconceived in advance, thanks to its more or less domed shape. and regular on the main face of the nucleus. This flake will normally be longer than it is wide and with several centripetal negatives on its upper face (at least four) which are the typical scars from the core preparation. Some Levallois cores are unusual in that they are sometimes struck from the side, rather than the end, so that the flakes are as wide as they are long, if not more.
Once exploited and abandoned, these nuclei have a very characteristic shape in which the peripheral bifacial preparation with a planoconvex section and a large chipped negative occupying practically the entire main face of the nucleus stand out; on certain occasions they are designated as "turtle shell nuclei". In Europe and, more specifically, in Spain, the Levallois nuclei appear in the Acheulean, but most of the time they have the blow on the side, and not longitudinal; similarly, Lower Paleolithic Levallois flakes are wider than they are long. Instead, the Middle Paleolithic Levallois is the typical one with the elongated and better controlled extraction.
 Levallois nucle with lateral percussion for a short lasca. |  Lasca Levallois short preferential. |
On the other hand, C. van Riet Lowe tried to establish in 1945 an evolutionary scheme that would explain the appearance and development of the preferential Levallois method in Africa, identifying several initial phases "pre-Levallois" in Stellenbosch (South Africa), another “proto-Levallois” phase named more specifically depending on its characteristics and regions, as Victoria-West in Uganda and Kerzaz type in Algeria, up to the Levallois proper, which is already linked to the evidence known in Europe during the Acheulean period.
An alternative method: repetitive centripetal extraction
At the end of the 1980s, the French researcher Éric Boëda carefully studied a series of Mousterian pieces from the Biache-Saint-Vaast site (Pas de Calais). discovering that, during the Lower and Middle Paleolithic, a Levallois variant had existed that had gone unnoticed or that had been identified with a common flake technique in those prehistoric periods, discoid centripetal extraction (which does not yield flakes). Shortly after, he carried out an in-depth revision of the Levallois method through the study of these and other archaeological objects and carving experiences. Boëda dedicates a large part of his work to establishing the differences between centripetal reiterative Levallois nuclei and discoid nuclei centripetal extraction. He also set out to establish the relationship between the classical Levallois method, for a preferential flake, and this repetitive procedure.
 Nucleus centrípeto discoide Not Levallois. |  Nucleus Levallois repetitive centripetal. |
Both the reiterative Levallois nucleus and the Discoid share a very profitable centripetal exploitation strategy; but, while the Levallois has a main mining surface with very little buckling, almost flat, the discoid core has them very curved, practically (bi)pyramidal or, at least, clearly (bi) convex. This is because the volumetric conception of the Levallois method is peculiar: the reiterative centripetal Levallois nuclei have a surrounding percussion plane, prepared to obtain a good batch of predetermined flakes, hitting peripherally; In addition, the exploitation face is quite flat, with a preparation very similar to that of a classic Levallois core, which makes it possible to make the most of its surface and obtain relatively large and invasive flakes, repeatedly. In other words, while in a non bifacial centripetal Levallois nucleus, both faces are equivalent, since both are exploited to extract flakes (hence its biconvex morphology, seen in profile); a reiterative Levallois nucleus, on the other hand, is planoconvex because each face has a different function: one acts as a peripheral percussion platform and the other as a centripetal exploitation zone.
 Lasca Levallois of the centripetal repetitive method. |  Nucleus Levallois repetitive centripetal. |
On the other hand, both the Levallois nuclei to obtain a single preferential flake and the reiterative centripetal Levallois nuclei have a preparation phase that is very similar, if not identical. But in the former, the preparation only leads to a flake with a balanced morphology, perfectly predetermined in which all the exploitation surface is used (a lot of energy is expended for a single extraction, but the control is very high, if there are no errors or accidents). On the other hand, in the latter, the energy profitability is greater, since a considerable series of predetermined flakes is obtained, of standardized size and shape (although with less control over each piece). The flakes are also invasive, although not so much, and they have the added characteristic that their negative serves to prepare the extraction of the next flake, for this reason the repetitive method produces pieces whose upper face has several negatives of flake, one of which may be larger than the others, being the result of some other flake obtained previously. By contrast, the flakes obtained by the classical method (preferential), have a series of centripetal negatives, all of very similar size, on their upper face.
Obtaining Levallois tips and leaves
Apart from flakes in the strict sense, the Levallois method is also used to obtain more specialized flake products, especially lithic tips and sheets, which, in both cases, require a different preparation than that described above.
Obtaining Levallois points: a parallel and preferential method
Levallois cores for spikes are prepared in various ways, all very specific, seeking to create a longitudinal rib on the exploitation side of the core (the so-called Levallois rib); There are various means to obtain this nerve, as it has been possible to demonstrate in carving experiments and in the reassembly of pieces from archaeological sites. This nerve is prominent enough to direct the fracture of the flake, when hitting a platform specifically prepared for it. The pieces thus obtained have a symmetrical triangular morphology, with a wide proximal zone (in which a dihedral or faceted heel is preserved) and a very pointed terminal zone divided in two by the aforementioned nerve as a bisector, in addition to being seen in profile, very fine and balanced. When the extracted point does not reach the limits of the exploitation face of the core, it is possible to take advantage of the same preparation to obtain a second Levallois point that will carry on its upper face the negative of the piece previously extracted. Although this rarely occurs, the procedure could be reproduced experimentally by François Bordes, who therefore speaks of first and second order Levallois points. However, before using the expressions proposed by Bordes it is necessary to examine the stigmata of the piece (either the point or the core) or carry out, if possible, remounting, and determine if two or a single point were extracted with the same preparation.
Obtaining Levallois leaves: a parallel and iterative method
Since the Middle Paleolithic, humans have been able to obtain flakes long enough to be classified as sheets obtained consciously, that is, not by chance, but by predetermining them in the core by by means of one of the variants of the Levallois method. The Levallois cores for leaves combine the choice of a very suitable morphological edge-support, that is, more elongated, with a particular Levallois preparation that forms a domed face tending to be cylindrical. Hitting from one or both ends of the support (poles), with parallel percussions (polarized), the flakes obtained are elongated (leaves). Since the negatives of the sheets already extracted, being longitudinal and almost parallel, serve to prepare the extraction of the following sheets (directing the fracture waves), it is not necessary to re-arrange the surface of the core to continue exploiting it; that is why it is said that this is an repetitive (or recursive) extraction method.
In fact, the Levallois method for leaves, more evolved, is the one that will crystallize, in the Upper Paleolithic, in the different laminar extraction methods.
Context and extension of the concept
Chronologically, it is one of the technical innovations of the full Acheulean, although it developed, above all, in the Mousterian (between isotopic stages 10 and 8), representing an important technological advance in the lithic industry. Geographically, it takes place in Africa, where there are very specific variants, also in Europe and South Asia. In fact, the Levallois method did not completely disappear in the Upper Paleolithic, reappearing in the Neolithic and in Australian Late Prehistory. We also have news that the Levallois method was reinvented in America, which justifies the following quote:
"It is interesting to note that this technique was developed independently, and in various ways, at various points of the Globe"François Bordes, 1971, page 19
As indicated above, certain scholars on the subject, such as Luis Benito del Rey, from the University of Salamanca, based on the observation of pieces from numerous Spanish sites (Cave of El Castillo, Cueva Morín and Cueva del Pendo in Cantabria; Cueva de Las Grajas, in Cueva Horá, in Andalusia...) propose to consider the Levallois idea from an open point of view, speaking of a Levallois substratum sensu lato which would include any flake method in which the total or partial predetermination of the product to be obtained is sought, even though the preparation of the core is slight, or even nil. Always, taking into account the existing difficulty for distinguish this type of pieces, many of which have been interpreted to date as fortuitous:
"...it is to be assumed that, if with a preparation of the nucleus or, sometimes, without any, they could already extract the preconceived lasca, they would not walk it "carefully" if they did not need it. On the other hand, it is the same thing that they did with the percussion plane of the nucleus, which was not confronted by them "necessarily" in the extraction of the Levallois product, if, as they had the nucleus with an adequate plain percussion plane, for example, they could already extract the predetermined lasca (hoe, tip)"Benito del Rey, 1984, page 24, note 44.
Opposite are those who defend a purist vision of the concept, pointing out as Levallois those stone objects in which there is no doubt of the specific preparation, even according to the classical ideas of the 60s, such is the case of Philip van Peer, which is based on his observations of deposits in North Africa, concluding that iterative Levallois methods simply do not exist, or have been misinterpreted. For Van Peer, the preparation of the core in order to obtain a preferential flake is a sine qua non condition to establish the true Levallois method (although he accepts that, once the piece is extracted, it is possible to prepare it again). the core for a new extraction); likewise, he sees no reason to distinguish Levallois flakes from Levallois blades, since both are technologically equivalent, and the only difference between them is morphology, which he considers secondary.
In recent times, therefore, there has been an undeniable open controversy about the limits of the Levallois Method, with positions going from one extreme to the other (as was revealed in a specialized publication of the year 2003 that is cited, by authors, in the references): thus, supporting Boëda's ideas about the variability and existence of repetitive Levallois methods, we find the previous proposals of Benito del Rey, later consolidated and supported by new data from both Acheulean and Mousterian sites; Jacques Tixier and his team also strongly support them. We find more nuanced positions in some members of the Atapuerca team and other French researchers, also with profound knowledge of the problem both in Europe and in Africa, such as Vincent Mourre. Perhaps the most resounding opposition to accepting the existence of a Reiterative Levallois Method for flakes (completely different from discoid centripetal techniques) can be found in certain authors such as Peresani or Slimak, for whom both procedures belong to the same concept, although we also find Spanish or French authors in this vein. To give just a few examples.