NINE

The Mousterian, Present and Future of a Concept

(A PERSONAL VIEW)

INTRODUCTION

The “Mousterian” is a stone artifact industrial complex restricted mostly to Europe (and in most characteristic form to Western Europe) and parts of Western Asia and North Africa. If we disregard the difficulty of differentiating it from the latest Acheulean and affiliated industries, the Mousterian seems first to appear during the Last Interglacial, more than 130,000 years ago, and to be replaced by Upper Paleolithic industries some 40,000 years ago. The term Mousterian was first applied by Gabriel de Mortillet (1869, 1872) to collections from the site of Le Moustier (Dordogne, France), made by Edouard Lartet in 1864 (which the discoverer assigned to the Epoch of the Mammoth). The first Mousterian artifacts excavated in Spain were those from Covalejos (1872) and Fuente del Francés (1880), found by E. de la Pedraja, who understandably did not give them that newly minted designation. For many years, the definition of Mousterian assemblages was complicated by the use of systems of artifact classification that included overlapping categories and such classificatory paradoxes as “round points.” We owe to the late François Bordes and his colleagues the systematization of artifact type definitions and the elimination of such obvious absurdities.

The complex is for the most part distinctive, although at the early end of its range assemblages intergrade so thoroughly with the latest Acheulean industries and such oddities as the “proto-Quina Tayacian” that any boundary between them is blurred, and some final Mousterian industries have so many backed knives and other supposedly “Upper Paleolithic” tool types that they, too, seemingly intergrade with such early Upper Paleolithic complexes as the Chatelperronian. At its inception, it is quite difficult to draw any clear-cut distinction between Mousterian and earlier industrial complexes, and perhaps this situation should lead us to reexamine the traditional definition of the Mousterian as a distinctive complex. However, at the recent end of its trajectory, careful attention to the choice of raw materials and flaking techniques, and the nature of retouch and its by-products, should help to differentiate the Mousterian from true Upper Paleolithic complexes. Not just backed knives but several other tool types ordinarily thought characteristic of Upper Paleolithic industries also occur in perfectly ordinary Mousterian assemblages, though not usually in the relative numbers they later assume. Such tools are burins, endscrapers, truncations, backed elements other than “knives,” blades, and even the occasional bladelet. While these precocious types occur in the Mousterian, and Chatelperronian levels in general contain a few types usually associated with the Mousterian, there is little room for confusion of the Cantabrian Mousterian with either Chatelperronian or Aurignacian industries. The Chatelperronian levels at Morín and el Pendo (the latter atop an Early Aurugnacian level) are quite different in lithic contents—tool type proportions, choice of raw materials, flaking techniques—from the Mousterian occupations that precede them.

There are Mousterian flakes that show apparent bladelet removals, often grouped at the flake butt near projecting ridges between flake scars. Some authorities interpret them as “bladelet cores,” but the removals may instead have simply been intended to thin the thick ridges, rather than aiming to produce bladelets. (The infrequency of finds of the tiny bladelets themselves is likely due to the fact that the ultra-fine screens needed to recover them were not used by the excavators.) Some assemblages contain high proportions of flakes from preformed cores (so-called Levallois flakes, points, or blades), while others have rare Levallois flakes or none at all. Since Levallois technique is relatively wasteful of raw material (a good deal of material may be lost through the process of core preparation), although it is conservative of the effort needed to bring the finished flakes to final form, after preparation intended to facilitate the sequential removal of several similarly shaped flakes, it is most likely to abound where (and when) there are natural exposures that provide ready access to suitable raw material in large sizes. Bifacial tools may be present, and in some assemblages, even abundant.

A major difficulty in studying Mousterian assemblages has been establishing their respective age. Suitable radiometric techniques that can be used to determine the actual age of Mousterian materials found in terrestrial sediments, within tolerable limits of accuracy, are deplorably almost nonexistent. For earlier periods such techniques as potassium/argon dating and uranium series dating are available, and though the dates they provide have relatively large margins of error, those margins are acceptable when the dated materials themselves are very old. For later periods, the radiocarbon technique is highly satisfactory, particularly when performed with accelerator mass spectrometry, although ages beyond 50,000 years estimated with this technique are suspect, and are probably best regarded as minimum estimates. AMS radiocarbon dating has demonstrated that in Cantabrian Spain early Upper Paleolithic assemblages make their appearance about 40,000 years ago: as early as in Central Europe and several thousand years earlier than was originally thought. (The evidence that the first Upper Paleolithic industries in Cantabria date back that far is incontrovertible, as is proved by the series of good dates on carefully excavated Aurignacian Level 18 at the cave of Castillo.) In the intervening period, such relatively unproven or questionable techniques as amino acid racemization, hydration, fission-track, and thermoluminescence dating have been applied, yielding what are at best “consensus dates”: i.e., they seem not to disagree with the preconceived ideas of age held by many specialists, or at worst are so wildly unreasonable that they are dismissed by all scholars. Mousterian stone tools may be large and crude in appearance. For that reason, assemblages from quarry/workshop sites where flawed and abandoned roughouts abound, or others with a large proportion of large, “heavy-duty” pieces that may probably have been expedient tools, are often mistakenly attributed to the Mousterian solely on the basis of their primitive appearance. This problem is not unique to the Mousterian; collections of Acheulean tools have been inappropriately assigned early “dates” on the basis of their relatively crude appearance, and to a lesser extent this erroneous practice extends to some Upper Paleolithic assemblages as well.

Internally, the Mousterian complex is heterogeneous. In Western Europe, François Bordes distinguished four assemblage types or “facies” within it, largely based on high percentages of sidescrapers at one extreme, denticulates and notches at the other, and the presence or absence of certain diagnostic implements. The idea of the “facies” was derived from Bordes’s geological training, but strangely, although true geological facies intergrade, Bordes insisted that the Mousterian facies were mutually exclusive, non-overlapping entities. To define them, one suspects that Bordes had to ignore or dismiss as “mixed” some well-excavated, intergrading assemblages. He called the four facies the Charentian (with sidescraper-rich subtypes Quina and Ferrassie, differentiated on the proportional representation of Levallois technique), the Denticulate Mousterian (rich in notched and serrate-edged tools), the Mousterian of Acheulean Tradition (subtypes A with handaxes and B in which handaxes were replaced by backed knives), and the “Typical Mousterian” (both of the last-mentioned facies as originally defined have only moderate quantities of sidescrapers or denticulates). Our own research on Spanish Mousterian assemblages and that of H. de Lumley in Provence indicated years ago that these “polar opposites” actually intergrade quite completely, and so designations for what Bordes would have thought to be “anomalous” assemblages, such as the “sidescraper-rich Typical,” have had to be invented. (This seems unsatisfactory, since it makes of the Typical facies a catchall into which everything that will not fit one of the other facies is crammed.) However, I believe that the practice we shall have to follow to further Mousterian studies will involve the eventual abandonment of the facies concept and a concentration on the development of new ways of classifying artifacts, toolkits, and individual assemblages.

PROBLEMS OF CLASSIFICATION

Until recently, so little has been known from Spanish sites about the relationship between Mousterian tool types or assemblages and paleoenvironmental conditions that most of what I can say about the complex from personal experience comes from the study of the stone artifact assemblages.

Anyone who has experienced the problems of classifying stone artifacts from both Middle Paleolithic and Upper Paleolithic sites cannot fail to have noted that the latter are far and away the easier tools to classify. Partly, that is due to the number of multiple-edged Mousterian artifacts that seem to show no significant tendency for particular types to combine. Partly it is due to the fact that some working edges are ambiguous, so that it would be as easy to call them denticulates as sidescrapers. Bordes dealt with such pieces by assigning them to the type that appeared “better made” or “less common” in the context of the assemblage being classified. But these are unsatisfactory solutions, being both subjective and hard for others to replicate. Probably a more objective classification of such artifacts can be facilitated by the use of techniques of fuzzy logic/neural networks (von Altrock 1995, 1997).

In fact, the overall shapes of Mousterian flake tools are simply not as tightly standardized as they become later, in the Upper Paleolithic. In fact, in Mousterian assemblages the shape and size of retouched working edges are better standardized than are whole tools. The great variability of artifact size and form that was tolerated in Mousterian industries is clearly shown by the variability of their sizes. Measuring individual tools in samples of a single well-defined tool type collected from different Paleolithic sites, or other sites occupied by modern Homo sapiens sapiens, one discovers that most often the difference between the means or medians of the samples is statistically significant: when tools made by fully modern people are compared, the only exceptions seem to be those categories of such coarse stone implements as hammerstones or mortars and pestles, whose size is determined by constraints imposed by considerations of manageability and the requirements of the physical activity that employs them. In such cases, adding samples from different sites together increases the variability and standard deviation of the collection, a clear indication of dissimilarity of the individual samples.

But with products of Neandertals, the situation seems to me to be quite different. Samples of well-defined types from any single site are very variable in their dimensions and other characteristics, but when discrete samples are added, it is as though they all came from the same parent population. One Cantabrian Mousterian type offers an informative example. At several sites, including Cueva Morín, el Pendo, and most importantly el Castillo, Mousterian assemblages that contain characteristic cleavers made on flakes have been recovered. While cleaver flakes seem reminiscent of earlier Acheulean industries (so much so that in 1961 Pierre Biberson called what may be a cleaver flake–bearing Mousterian from the “brecchioid limestones” at Cap Chatelier at Sidi Abderrahman “Acheulean Stage 8”), the assemblages with cleaver flakes from Cantabria do not seem to be very ancient within the Mousterian complex. (Bordes misreadingly baptized these Cantabrian assemblages “Vasconian”; in addition to the name, which wrongly suggests that they are common in Basque country, the term suggested inappropriately that the assemblages were alike in many ways, when in fact they are internally heterogeneous.) When we consider the length measurements of (unbroken) cleaver flakes from the four Cantabrian collections discussed below, the lack of standardization of measurements and the similarity of collections is obvious.

TABLE 9.1. Descriptive statistics, Cantabrian cleaver flakes: Lengths (cm)*

* (Elpl = el Pendo; Morl = Morín; Castl = el Castillo; Alcl = Alcedo).

These calculations were made using the Minitab© statistical software program. Measurements are maximum lengths of specimens parallel to the “axis of symmetry,” based on artifacts in museum collections from older excavations, except in the case of Cueva Morín, where specimens excavated during the 1960s have been added to the museum pieces.

As the table (Table 9.1) of descriptive statistics shows, the means, medians, and maximal and minimal values for lengths of cleaver-flakes from four Cantabrian sites all seem quite similar to one another, despite the fact that the smallest collection, that from Alcedo, contains only five pieces, and the largest, that from Castillo, has 156. Since so many of the pieces are from older collections, and since it could not even be assumed that the underlying distribution of lengths was normal and unimodal, I chose to compare collection medians rather than means as measures of central tendency, using the non-parametric Mann-Whitney test instead of the more usual “Student’s t.” (The use of the “t”-test would probably have been justifiable, as the distribution of lengths in the large Castillo collection shows, but the results would not be appreciably different.) The tests (Table 9.2) show that none of the collections differs from any of the rest as much as they would be expected to differ by chance alone (at the 0.95 level). The two largest collections, that from Morín (62 pieces) and that from Castillo, are no more different than one would expect samples from the same population to be nearly 15 percent of the time.

I have also identified what I believe to be a series of trimmed tools in bone from Mousterian levels at Morín and el Pendo (Freeman 1978, 1980; González Echegaray, Freeman et al. 1971, 1973). The fact that the bone supports for the more or less standardized retouched edges are not regularly patterned has led others who think in terms of more regular bone and antler types of the Upper Paleolithic (where crude bone tools also exist, but are often ignored) to challenge that identification. But (1) the pieces in question are not simply crushed by trampling; (2) the only creature capable of mouthing the larger bones would be the spotted hyena, which is absent entirely from the levels with “bone tools” at el Pendo, while 2 premolars and a coprolite are its only remains in one of the two Morín levels with such pieces—convincing evidence that hyenas denned in any of these occupation levels is lacking; (3) some Upper Paleolithic levels, such as Level 5 at Cueva Morín, also have hyena remains, but they lack these kinds of “worked bones” entirely; (4) the pieces identified as deliberately fabricated tools closely parallel in frequency the proportional representation of analogous stone tool types from the same levels; and most conclusive of all, in my opinion, (5) the identification of deliberate retouch on these pieces was verified by a faunal expert, Jesús Altuna. It seems likely that the variability of patterning of overall shapes and sizes of these pieces has been the major—perhaps the only—obstacle to their acceptance as tools.

TABLE 9.2. Mann-Whitney U test for difference between medians*

* These calculations were made using the Minitab^© statistical software program. Measurements are maximum lengths of specimens parallel to the “axis of symmetry,” based on artifacts in museum collections from older excavations, except in the case of Cueva Morín, where specimens excavated during the 1960s have been added to the museum pieces. ETA is the symbol used for the median.

The “looseness of patterning” I have noted in connection with artifact design and production is also observable in other aspects of Neandertal behavior. Spatial differentiation of activities is not as well developed as it became in the Upper Paleolithic. At Cueva Morín, the demarcation of different areas within the occupation of Mousterian Upper Level 17 is not nearly as crisply marked as it became later, in Earlier Aurignacian Level 8a (Freeman 1993). Despite the presence of remnants of buildings in both levels, the structure of the Aurignacian level is more regular, and simpler, than it was in the Mousterian level, and the areal differentiation of the distribution of different “toolkits” is much better demarcated.

Since their definition, most of our interest has been focused on Bordes’s facies and a diligent search for the causes of facies difference. In the process of learning about those causes, we have found that we were chasing an illusion. I believe that to progress in Mousterian studies we must shift our attention away from the supposed Mousterian “facies,” and even away from the whole assemblage as the unit of study (though we must still take into account each and every artifact from every level). Instead, in future our attention should better be focused on identification and study of the toolkits that make up each whole assemblage, on the relations between them and their relationships to contextual information such as sediments and biotic remains, and on the positions (activity areas) that the toolkits occupy within an occupation. The idea that if whole excavated assemblages are studied they will document all the activities undertaken within a level is fallacious. Even if tool types were homogenously distributed over a living floor (and they never are), we could still not recover those perishable parts of the inventory that are almost certainly important to the interpretation of represented activities, and in any case archeologists virtually never excavate an occupation level in its entirety.

There are two reasons that I believe that Bordes’s facies concept will have to be abandoned. First of all, the facies are now known to be arbitrary constructs of the classifier, imposed on assemblages that actually seem to intergrade completely, as least in my own experience with Cantabrian sites (Freeman 1994). Secondly, it is now known that well-excavated assemblages actually occupy somewhat different positions along a continuum that has denticulate-rich assemblages at one extreme and sidescraper-rich assemblages on the other, and that Bordes’s interpretation of the causes of facies difference as stylistic rather than functional was fallacious. Paradoxically, while in his artifact classification Bordes virtually eliminated most possibly stylistic attributes of stone tools from consideration, insisting, so to speak, that a knife from any period or part of the world should be called by the simple typological designation “knife,” and regarding the attributes that set knives from one part of the world off from those used in another as “accidental,” he nevertheless regarded the differences between his “facies,” which are after all just groups of tool types, as being the stylistic manifestations of group identity used by different “tribes,” who signaled their uniqueness by the proportions of the different types of tools that they made and used. Such a cumbersome means of marking intertribal difference is unknown to modern anthropologists who have studied living groups, who signal their identity by signs such as body-marking, clothing, headdress, adornment, or banners or the marks they put on easily decorated items such as animal hides, textiles, ceramics, bone, and wooden artifacts, etc. Some of these signs can easily be seen at a distance. If further proof that the facies cannot be the tool groups distinctive of different “tribes” were needed, it is provided by Level 16 at Cueva Morín, where large samples of tool types that Bordes would have had to assign to different facies (307 “essential” flake tools from the Denticulate and 222 from the Typical Mousterian) were recovered from a single contemporaneous occupation in areas far too small (3 square meters as opposed to 7 square meters) to have been possibly inhabited at the same time by two different tribes. In such cases, the likely explanation of assemblage differences is functional, not stylistic in Bordes’s sense.

THE AUTHORS OF THE MOUSTERIAN

So much for the nature of the artifactual evidence. What about the physical type of the people who made Mousterian tools? The authors of Mousterian assemblages are almost all assumed to have been Neandertals. There have been indications at some sites of a supposed “co-existence,” and perhaps even a genetic continuity of Neandertals and fully modern Homo sapiens sapiens. However, more recent studies seem to show that Neandertals are not in fact ancestral to modern humans, but a specialized side branch of the human family tree. Despite the opinions of some authorities, the nature of the artifacts they made tells us very little about genetic continuity between Neandertals and modern people. This is in spite of the fact that (at least late) Neandertals buried their dead, cared for the incapacitated members of society, and occasionally engraved or pecked geometric patterns into hard surfaces and used “pencils” of coloring material to decorate something, perhaps their own skins. Some years ago, we found indications that a real gap divided the behavior of Mousterian peoples from that of their Upper Paleolithic successors at sites such as Cueva Morín, despite the fact that some of the tools made by both groups looked superficially quite similar. However, anyone who mistakenly believes that the difference between the two is due to some such single, oversimplistic causal factor as the lack of a gene to permit articulate speech betrays his or her ignorance of the complexities of Neandertal behavior and human evolution in general. Whatever the case, any real proof of the relationship of Neandertals to modern Homo sapiens sapiens will have to come from studies of skeletal materials and genetics; it cannot be based on an examination of stone tools. Biological and cultural evolution are not the same, and then as now, there was no necessary correspondence between hominid physical type and culture, no matter what some scholars may erroneously think.

VARIABILITY AND INNOVATION

During the 2004 Neandertal Workshop in Santillana del Mar, Francesco d’Errico asked the assembly whether it is possible to equate behavioral variety with innovation. Clearly, the answer is no! Innovation extends the behavioral repertoire by the addition of novelty. Variability is simply the exercise of different parts of a behavioral repertoire that may already preexist the choice. The distinction is important. Within any industrial complex, such as the Solutrean, the degree of variability of stone tools is really not much greater than that marking the difference between the Mousterian “facies.” And the number of tool types represented in a well-excavated assemblage of given size does not really distinguish Mousterian and Upper Paleolithic assemblages at all well. (This fact is not generally appreciated, because the de Sonneville–Bordes/Perrot Upper Paleolithic type list contains so many more types than the Middle Paleolithic type list of Bordes.) However, that is far from the case for rates of innovation. Whether manifest in the addition of several new tool types, including bone and antler artifacts, to the general inventory of Upper Paleolithic tools, or the specialization of sites and occupations, or the comparatively rapid turnover of industrial complexes, or the addition or “popularization” of personal adornment and wall art, innovations took place very much more rapidly during the Upper Paleolithic than they did during the Mousterian.

There are certainly cases where Neandertals have apparently behaved in novel ways: burying the dead and the use of adornment and body embellishment may be two examples of Neandertal behavioral innovation. But one “innovation” that caused considerable excitement when it was announced was the use of flowers as grave offerings at Shanidar cave, found during the 1960s (Solecki 1971, 1975; Leroi-Gourhan 1975). In this case, the excitement may have been premature. Instead of occurring as usual in isolated grains, the pollen from eight of the represented species apparently associated with Burial IV at Shanidar was in many cases found in clusters of “from 2 to more than 100 pollen grains. Certain of these clusters have retained the form of the anther of the plant” (Leroi-Gourhan 1968, 1975: 562). This suggests that the pollen was not windborne, but came from whole flower heads deposited in the sediments. The flowers identified were from the genera Achillea (a daisy), Centaurea (the corn flower group), Senecio (ragwort), Muscari (grape hyacinth), Ephedra (horsetail), and Althaea (hollyhock). The pollen analyst believes that the flowers were probably collected from May through July, which is in approximate agreement with the date of discovery of the Shanidar IV burial during excavation in early August 1960. All the represented genera are known to have medicinal properties, a point which was stressed at the time (Solecki 1975), but that is nothing extraordinary. Virtually all plants grown as ornamentals have been claimed to have medicinal properties at one time or another in the past. The genera identified all have members that the local Kurds grow in their gardens, and Solecki tells us that his workmen were in the habit of bringing whole flowers into the site tucked into their cummerbunds or the handles of their wheelbarrows (1971: 93–94, 176). I have always suspected that the flower pollen from Shanidar IV could be the result of accidental modern contamination. After the samples were subjected to acidic treatment to remove pollen from the surrounding sediment, it would have been difficult, to say the least, to tell ancient from modern pollen. The discovery of the delicate scale from a butterfly’s wing in one of the pollen samples does not help the case for contemporaneity of the pollen and the burials. Last, pollen samples from Shanidar also contained vestiges of tobacco pollen and that from cultivated date palm, which were certainly not plants known to the Neandertals. Were innovative Neandertals at Shanidar the “First Flower People,” as Solecki would have it? Perhaps, but the case is still not proven.

INFERENCE AND SPECULATION

Variability and imprecise standardization (looseness of patterning) are hallmarks of Mousterian tools. The nature of this variability is such that the processes of teaching each succeeding generation of Neandertals how to make tools must have been very different from the socialization processes of modern Homo sapiens sapiens. In general, the fact that given artifact types from a single site are internally quite variable, while the ranges of variation within a given type from different sites in a wide region overlap to a large degree, shows that there is little or no deliberate stylistic information encoded in the types as Bordes defined them, nor as the measurements show is there the sort of unconscious stylistic load that is often incorporated in the products of different modern identity-conscious socio-cultural groups. In part that may be an accidental consequence of the fact that it is hard to alter lithic artifacts without affecting their function, but it may also reflect an important adaptive reality. It may well be that intergroup boundaries were not as purposefully maintained, signaled, and defended as they are among most modern societies. The distinction of “we” from “they” may have been adaptively dysfunctional as long as human groups were small and resources abundant. Appropriate mates would have been hard for an adult to find in a co-resident group whose maximum numbers were only as large as a few score people of all ages. Permeable group boundaries would have eased intergroup movements of personnel, and could have been advantageous to survival.

As a student, I was taught that the processes of natural selection would in the long run result in enhancing the formal differentiation of tool types, as a result of their increasing functional specificity, or suitability for different tasks. But this process may take either (or both) of two directions. Tools may either be improved by making them better suited to the performance of different elementary tasks, such as hacking, slicing, or perforating (regardless of the substrate on which they are used), or they may be designed to improve their efficiency in performing a limited set of tasks on a selected substrate (the apple corer and peeler is one example, the nutcracker another, and the whaler’s harpoon head a third), or on a material that only exists in a particular world region such as the tropics (manioc shredders or the ice knives used by the Inuit, are examples).

Through the Early Würm/Weichsel, the evolution of artifact morphology generally reflects continued design improvements that made working edges and whole tools increasingly efficient for use in a small number of primary operations. I have called this kind of adaptation of the tool inventory “technique-oriented.” Tools for chopping become more effective choppers (and less efficient hammers or slicers), slicing implements become more efficient as slicers, and in general each implement type becomes more differentiated and better adapted to a specific kind of manipulation. But those primary operations may be performed on a variety of materials: skins, vegetal materials, and meat can all be sliced with the same cutting edge. This, it seems to me, is the most appropriate interpretation of the saying that some Mousterian tools are “general purpose tools”: they can perform the same limited set of operations on a variety of resources in vastly divergent environments. There is little evidence that any tool type was specifically tailored to work on one material alone.

There are a few stone tool types, such as the Szeletian knives of Central/Eastern Europe and the cleaver flakes of North Africa and parts of Spain, that apparently signal some degree of regional diversity, seeming already to reflect the beginnings of regionally distinctive adaptations, although the artifact classifications currently in use ignore the most obvious “stylistic” attributes of tools. While there do seem to be some interregional boundaries across which notable differences in artifact inventory can be discerned, the size of the areas in which artifact series are homologous at any time seems surprisingly large compared to the Upper Paleolithic condition. When idiosyncratic attributes of artifact assemblages confined to more restricted geographic regions do at last begin to appear during the Middle Paleolithic, they do seem to be the result of conscious, deliberate stylistic differentiation.

This is in marked contrast to the development of some of the tool types in the Upper Paleolithic artifact inventory which are regionally restricted or must have been used on a specific small number of locally available resources. I have called this “regional-and-resource-oriented” adaptation. In addition to the specialization of tools for use on specific resources in particular regions, enigmatic decorative devices appear, such as geometric patterns incised in bone artifacts, or painted symbols such as those accompanying some of the large painted animal figures at Lascaux. These may have signaled both individual and group identity, as it became increasingly necessary to demarcate regions and to stress one’s claim to rights and privileges within them. Another, non-artifactual, kind of evidence for the appearance of this new adaptive orientation is the discovery in increased numbers of bones of small, nocturnal, burrow-dwelling fur-bearers in Upper Paleolithic occupations. Such creatures, ordinarily invisible to hominids living in their vicinity, can only be taken in numbers with traps that have been specially designed to take advantage of the animals’ size and behavior. Such traps have not yet been identified in most Upper Paleolithic sites, but the bones of their prey attest to their former existence.

At one time animal bones from Mousterian levels as a rule suggested to me an opportunistic exploitation of all mammals available in the environment that were “easy enough to see and easy enough to hunt.” I contrasted this with the intentional selective harvesting of wild resources attested for some Upper Paleolithic sites, where those few resources that were most productive were deliberately chosen. Since that time, some Mousterian levels have been excavated that indicate that hunters occasionally focused on the procurement of one or a small number of species, as they did more commonly during the Upper Paleolithic. However, it is my impression that when resources are present in a Mousterian site, they are usually (though not always) those that were readily available in the near vicinity of the site itself, and that they were not transported in any quantity for any great distance: large numbers of shellfish do not travel far from the coast, and alpine animals are only abundant in alpine sites, etc. Now, as any human group becomes more familiar with its natural surroundings, in a process the late Robert Braidwood called “settling-in,” it is almost inevitable that it will eventually learn to exploit those resources that are locally most available and easiest to take, either seasonally or throughout the year. (That, it seems to me, is the explanation for the accumulation of shellfish remains in the Gibraltar caves.) Mousterian groups could very well have learned to move periodically from site to site to take advantage of resource availability, but there seems to me to be little evidence for the movement of materials between regions.

During the Upper Paleolithic, humans settled areas much further into the inhospitable northerly latitudes than they ever had done earlier. Apparently this movement was facilitated by the development and extensive use of storage devices such as pits, some apparently used for cooking. Storage facilities would of course help tide one over a lean season when resources were scarce, and would facilitate the protection of resources from the actions of competitors, and add to the available battery of food-preservation techniques. It is striking that while storage facilities are not entirely absent, they seem to be very rare in Mousterian sites: this certainly speaks to the tenuous nature of Mousterian life in seasonally inhospitable regions. It is evidence for the more limited nature of food preparation and preservation techniques (open-fire roasting, stone boiling in skins [?], food drying and chilling?) available to Mousterian peoples, in comparison to the wider range (including baking or boiling in pits, the storage of roots and tubers in dark, humid conditions, reheating, and probably controlled fermentation or pickling, as well as better defense against carnivores or destructive small organisms) that were available during the Upper Paleolithic.

In sum, I do not believe that selective exploitation by Mousterian peoples ever reached what may be called the final, perfected stage of “primary hunting/gathering efficiency.” That would have involved designing specialized tools for use in a specific region of operation or on a specific set of resources. How can one differentiate Mousterian behavior from that characteristic of the “wild-harvesting” adaptations of the Upper Paleolithic? The latter must leave distinctive traces in the archeological record: artifact inventories that are well differentiated on a microregional scale, including differentiation of attributes of what de Sonneville–Bordes considered single tool types (an example is the presence of different “styles” of Solutrean points, seen in the comparison of Smith’s 1966 study of French Solutrean pieces, the concave-based points common in Cantabria, or the stemmed points from Parpalló and Ambrosio in the Valencian region); accumulations of selected raw materials in sites; the nature of the selected prey; evidence for the harvesting of healthy individuals of all ages rather than just the young, the old, and the feeble; and evidence for the relatively long-distance transport of resources. One might also expect to find that different parts of the habitat were specifically chosen for the extraction of the different resources available in each: the mountain slopes for alpine animals, the coast for fish and shellfish, and specific quarry/workshop areas used for the acquisition of stone raw materials, and that this pattern of exploitation was combined with evidence for the long-distance transport, by a single human group, of quantities of materials from their sources to their areas of utilization, where they were not readily available (or with evidence of interregional exchange). This would seem to coincide with the pattern of movement of resources like marine mollusks or alpine animals documented for the Cantabrian Upper Paleolithic or at least its later manifestations. (It may be that Upper Paleolithic stone tools had a more limited, regionally restricted trajectory than did raw materials for tool manufacture, items for adornment, or foodstuffs.) While no single kind of evidence may prove conclusive of such an adaptation alone, in combination their presence is most suggestive. On the contrary, I believe that Mousterian sites provide much less evidence for the movement of goods between regions.

A POSSIBLE DIRECTION FOR FUTURE RESEARCH

In the past (1977, 2005) I have suggested some changes in our procedures that I believe might prove fruitful in future studies of the Mousterian complex. They involve the combination of a viewpoint that tends to be more anthropological than geological with new techniques for artifact classification (taking into account core reduction sequences, the staging of implement manufacture and resharpening, and the effects of continued use on working edges), and using proven quantitative procedures for the definition of stone artifact types from individual sites, the classification of toolkits, excavated assemblages, and the areas where different activities were performed. My perspective borrows from B. Malinowski’s (1960) concept of institutions, and adds a modification of F. Gearing’s (1962) concept of the “structural pose.”

We know that every living society apportions some special part of its complex set of behavioral inventories to individuals defined as especially suited to the requisite behavior patterns (and there is no reason to think that Neandertals were very different from modern humans in this respect). It is possible to study this apportionment in two ways. When one is interested in the learned behavioral patterns assigned to the several positions in a society that an individual may occupy, one should study the behavior patterns as the “rôles” of individuals. But when one focuses on the purposes of behavior, the individual performers and their positions are less pertinent than the patterns themselves, and the behavioral categories of greatest relevance are sets of responses that are culturally defined as appropriate to identifiable and recurrent situations. I have called these sets of responses the “functional modes” of a social group: modes such as dancing, curing, hunting, mourning, toolmaking, clothing manufacture, and fighting are examples. They are loosely defined and may even intergrade, for no human behavior (present or past) is ever packaged in minimal, contrastive, non-overlapping sets. Some of these functional modes, several of which may be operative in the behavior of individuals at the same time, are manifest in the behavior of single individuals, while others may require the cooperation of several individuals, organized into loosely constituted temporary groups or rigidly structured, long-enduring “corporate” bodies.

Each functional mode employs a cultural apparatus, consisting in the range of permissible behavioral alternatives open to the performers, a set of attributes and values that serve to guide performance, and sometimes, a set of physical equipment used by the performers, which may be called the matériel. Even in cases where the functional mode of behavior requires no matériel, its activities often alter the natural surroundings in recognizable ways.

Excavating relatively unmixed and largely “undisturbed” archeological occupations, the archeologist (or, if you will, the prehistorian or paleoanthropologist) can recover durable artifacts in association with patterned contextual materials such as fungal spores, chemical traces, remains of plants and animals, and sediments, as well as information about the location and the relative position and abundance of each of these categories of evidence. A quantitative search for significant patterned relationships between artifactual and contextual data can optimally evaluate the contribution of random effects to these relationships, and define related constellations of matériel which vary together, and independent of other sets. These, if correctly isolated, can represent the matériel and by-products of activities associated with distinct functional modes of behavior: some are the toolkits and by-products of activities associated with distinct functional modes of behavior: some are the toolkits and by-products of economic/technological activities; others may reflect organizational or ideological elements. Each different individual’s behavior is to some extent idiosyncratic, and so the matériel and by-products of a set of activities undertaken by one group of performers can be expected to vary “stylistically” from those of another set of performers doing the same or similar things. This is a potential route to the recognition of “team membership,” and the eventual definition of “regional style zones,” the delineation of different socio-cultural systems, and an understanding of their location and duration and of the stylistic changes they have undergone over time.

The program I have outlined is at present admittedly a visionary ideal, but it is not at all unrealistic: it can be realized as evidence from future well-controlled excavations becomes increasingly available. I believe that we are now poised at the brink of great developments in Mousterian research, and that investigations in Cantabria will soon add substantially to our understanding of this (in many respects) enigmatic industrial complex.

But the complex is likely to remain an enigma for some time, if we cannot overcome the traditional reluctance of some of our colleagues to criticize the conclusions of their teachers, even when those conclusions have been shown to be in error. That attitude is entirely mistaken, as I am sure those same teachers would agree. All honor and respect are due those pioneers in our field whose brilliant contributions brought us to our present state of understanding of the Mousterian industrial complex. But that does not mean that we can never be more than their carbon copies. It is of course much easier for today’s investigators to follow the path of least resistance, slavishly repeating the ideas of their teachers. But that would be a crippling error. As they become outmoded, older approaches must be supplanted by the newer, more productive perspectives that we have gained as a result of seeing where the perspectives of the past fail to fit the data now in hand. That is the essential prerequisite of future progress.

The readiness of Cantabrian scholars to challenge theories that have outlived their usefulness is one of their strengths, “uncivilized” though that attitude may seem to some. I do not mean to imply that our field is advanced in the least by the snide critical attitude of a very few of our colleagues from other countries, who have gained their reputations by criticizing the work of others with “data” or “reinterpretations” that they have fabricated just for that purpose, or by mastering the art of damning with faint praise. But a positive and well-intentioned critical approach to investigation is a hallmark of good science, and that is what the best younger scholars in Spain and the Cantabrian region do so well.

Spanish investigators, among whom I think particularly of those working with renewed intensity in Mousterian studies, are already taking their place as recognized leaders in prehistoric research. I hope that work at the cave of Sidrón in Asturias may add immeasurably to our knowledge of regional Neandertals. I am confident that ongoing excavations at the cave of Castillo, following in the tradition established by the late Victoria Cabrera and her husband, Federico Bernaldo de Quirós, will continue to provide invaluable information about the development of the Mousterian in Cantabrian Spain, as well as about the nature of the transition to the Upper Paleolithic, not just in the region, but in Western Europe as a whole. We are at the threshold of a new era in Paleolithic studies. Let us all hope that its results will be as bright as its promise.

REFERENCES

Biberson, P. 1961. Le Paléolithique Inférieur du Maroc Atlantique. Rabat, Service des Antiquités, Fasc. 17.

Freeman, L. G. 1975. By Their Works You Shall Know Them: Cultural Developments in the Paleolithic. In Hominisation und Verhalten, ed. G. Kurth and I. Eibl-Eibesfeldt, 234–61. Stuttgart, Gustav Fischer Verlag.

———. 1977. Anthropology without Informants. In Horizons of Anthropology, 2nd ed., ed. S. Tax and L. G. Freeman, 132–50. Chicago, Aldine.

———. 1978. Clasificación de los útiles de hueso del nivel 17. In Vida y muerte en Cueva Morín, ed. J. G. Echegaray and L. G. Freeman, 253–62. Colección del Bolsillo 7. Santander, Institución Cultural de Cantabria.

———. 1980. Hueso trabajado de el Pendo. In El yacimiento de la Cueva de “el Pendo” (Excavaciones 1953–1957), ed. J. González Echegaray et al., 65–67. Madrid, Bibliotheca Praehistorica Hispana 17.

———. 1991. What Mean These Stones? Remarks on Raw Material Use in the Spanish Paleolithic. In Raw Material Economies among Prehistoric Hunter-Gatherers, ed. A. Montet-White and S. Holen, 74–125. University of Kansas Publications in Anthropology 19. Lawrence.

———. 1993. La “transición” en Cantabria: La importáncia de Cueva Morín y sus vecinos en el debate actual. In El origen del hombre moderno en el suroeste de Europa, ed. V. Cabrera, 171–93. Madrid, Universidad Nacional de Educación a Distancia.

———. 1994. Kaleidoscope or Tarnished Mirror? Thirty Years of Mousterian Investigations in Cantabria. In Homenaje al Dr. Joaquín González Echegaray, ed. J. A. Lasheras, 37–54. Monografía 17. Santillana, Museo y Centro de Investigación de Altamira.

———. 2005. La investigación sobre el Paleolítico Medio en la region cantábrica: ¿De dónde venimos? ¿Dónde estamos ahora? In Neandertales cantábricos: Estado de la cuestión, ed. R. Montes and J. A. Lastheras, 21–38. Santillana del Mar, Museo de Altamira.

Gearing, F. 1962. Priests and Warriors. Memoir 93. Washington, DC, American Anthropological Association.

González Echegaray, J., L. G. Freeman, et al. 1971. Cueva Morín, excavaciones 1966–1968. Patronato de las Cuevas Prehistóricas de la Provincia de Santander 6: 135–61. Santander.

González Echegaray, J., L. G. Freeman, et al. 1973. Cueva Morín, excavaciones 1969. Patronato de las Cuevas Prehistóricas de la Provincia de Santander 10: 58–64. Santander.

Leroi-Gourhan, Arl. 1968. Le Néanderthalien IV de Shanidar. Bulletin de la Société Préhistorique Française 65(3): 79–83.

———. 1975. The Flowers Found with Shanidar IV, a Neanderthal Burial in Iraq. Science 190: 562–64.

Malinowski, B. 1960 (1944). A Scientific Theory of Culture and Other Essays. New York, Galaxy Books, Oxford University Press.

Smith, P. 1966. Le Solutréen en France. Bordeaux, Delmas.

Solecki, R. 1971. Shanidar, the First Flower People. New York, Knopf.

———. 1975. Shanidar IV, a Neanderthal Flower Burial in Northern Iraq. Science 190: 880–81.

Von Altrock, C. 1995. Fuzzy Logic and Neurofuzzy Applications Explained. Upper Saddle River, NJ, Prentice Hall.

———. 1997. Fuzzy Logic and Neurofuzzy Applications in Business and Finance. Upper Saddle River, NJ, Prentice Hall.