Cantabria and Vascongadas, 21,000-17,000 B.P.: Toward a Solutrean Settlement Pattern

195

MUNIBE Sociedad de Ciencias Naturales ARANZADI San Sebastián Año XXXI - Numere 3-4- 1979 Páginas 195-202

Cantabria and vascongadas, 21,000-17,000 B.P.: Toward a solutrean settlement pattern.

LAWRENCE GUY STRAUS*

Introduction been «site-catchment analysis», developed by Higgs, Vita-Finzi and others to attempt Our knowledge of Paleolithic adaptations assessment of the role of particular sites has been based principally on the study of with regard to potentially exploitable resour- artifacts and, increasingly, of faunal remains ces located within fixed radii from the sites recovered from archeological deposits. The- (see for example, Vita-Finzi and Higgs 1970). re is, however, a third class of evidence which The model has been criticized for its rigidity, can provide valuable insights into the opera- although provisions have been added to ac- tion of past cultural systems: site location. count for walking times in different types of In a little-known, pioneering work published terrain. More seriously, the model implies in 1908, R. P. Lorenzo Sierra presented «No- rather simple, single-minded procurement tas para el mapa paletnográfico de la Provin- strategies on the part of mobile hunter-ga- cia de Santander» to the 1.er Congreso de therer societies, and does not take into ac- Naturalistas Españoles in Zaragoza. Since count the very likely possibility of elaborate those early days, Paleolithic prehistory tur- «logistical strategies» with multi-purpose task ned increasing toward a focal emphasis on «embedding» under certain evironmental cir- typology as the sine qua non of what was be- cumstances (see Binford 1978; n.d.). Despi- lieved to be a developing scientific discipline. te its mechanistic weaknesses, however, the The meritorious practices of increasingly site catchment notion is useful because it complete artifact retrieval, classification and focuses attention on the roles of sites as loci statistical comparison have often had, howe- of human activities and occupation in rela- ver, the unfortunate effect of equating sites tionship to surrounding terrain and potential with artifact mines. The treatment (theoreti- resources of use in the maintenance of so- cally, if not practically) of archeological si- ciety and individual lives. tes like paleontological localities in this sen- Not controlling their food supplies, Paleo- se utilizes only part of the information which lithic groups could assure their existence sites can provide, no matter how meticulous through the development of strategies for and complete the recovery and recording me- the exploitation of regional resources. Given thods used. finite resources and a growing human popu- To balance the influence of the French lation (see Straus 1977; Cohen 1977), Upper school of «typological prehistory», recent Paleolithic groups, such as those of the So- years have seen the development of a British lutrean period (c.21,000-17,000B.P.) in north- school of economic prehistory which places coastal Spain (Cantabria - Vascongadas), de- emphasis on faunal remains (and other evi- veloped elaborate compound technologies of dence of subsistence activities), and on the stone, antler (? cordage and wood) for the geographical locations of sites within lands- procurement of an increasingly diversified capes. A major technique of this school has spectrum of animals (including riverine. estuarine, littoral and marine species). Many * University of New Mexico. Albuquerque. N. M. sites contain indications of the use of mass (U.S.A.) hunting techniques (drives, surrounds) in 196 L. G. STRAUS the procurement of red deer (Cervus elaphus) Analyses at the sites of La Riera (Posada de and ibex (Capra pyrenaica). Small but con- Llanes, Asturias) conducted by Dr. J. Altuna sistent numbers of choppers and grinding sto- and Dr. N. J. Shackleton, and El Juyo (Igollo nes in Solutrean collections may attest to de Camargo, Santander) by Dr. R. G. Klein, increasing utilization of plant foods such as for example, are directed at attempting to de- nuts, berries and roots. In addition to the termine the seasonality of long series of oc- technological developments, changes in the cupations of caves to test ideas concerning division of labor and in social organization the yearly rounds of Upper Paleolithic-Meso- are implied from the direct subsistence evi- lithic groups under varying environmental dence. Under conditions of relatively high po- (and demographic?) conditions. At the mo- pulation density and highly variegated terrain, ment, however, indications of Solutrean sea- territorial organization and sophisticated ac- sonality are most scanty and unreliable. What tivity scheduling would seem inevitable in or- is required is a large number of similarly der to assure adequate wild food supplies analyzed occupations from penecontempora- throughout the year. Hunting success de- neous sites located at different elevations pends on accurate information on resource and in different topographic settings. location and condition; such information can One set of data which is however availa- be maximized and game can be successfully ble concerns the geographical location and «managed» through a territorially-based dis- topographic setting of 33 known Solutrean tribution of hunters. Within territories indi- sites in the region, all but 2 of which have vidual families or bands may have foraged been visited by the author. Typical Solutrean independently part of the year and yet con- artifacts from all but 7 can be found in Spa- gregated at established or opportune times nish museums and other institutions. Solu- to best take advantage of aggregated resour- trean points are said, in various early publi- ces (by means of communal drives of animal cations, to have been found in the caves of herds, for example), to conduct ceremonies Caranceja, Fuente del Francés, Bona, Haza, (initiation, fertility and hunting magic, etc.), Mirón and Sel (=Cueva del Agua? =Cueva to exchange mates, etc. The survival of Last de la Peña?), and although the collections Glacial hunter-gatherer groups could well ha- could not be located, these descriptions are, ve depended on their territorially-based coo- to varying degrees, taken as credible since perative abilities to exploit the full wild re- Solutrean points are such distinctive diag- sources of well-known local environments. nostics (see Straus [1975a] for details). A The precise placement of sites would have Solutrean point found in disturbed circums- been part of the repertoire of adaptive skills tances during the 1974 excavations at Ras- possessed by these groups. caño (Mirones, Santander) may actually ha- Given the mobility of the main Solutrean ve come from the adjacent cave of Bona, food resources —medium and large-size un- which was apparently being dug at the same gulates— (and variability in the condition time as Rascaño by R. P. Sierra (see Straus and availability of these and other resources n.d.). Various reports of Solutrean materials such as shellfish, fish, plants, etc.), hunter- at Atxurra, Lezetxiki and Lumentxa have been gatherers had to respond through patterns positively demonstrated to be erroneous of mobility of their own, both seasonally and (Straus 1975a). Distributed over an area so- circumstantially. In recent years various stu- me 350 km. long by about 30 km. wide, the dents of prehistoric Cantabria have sugges- 33 Solutrean sites, pertaining to a 4000 year ted patterns of strategic site placement and period (Straus et al. 1978), can provide va- transhumant movements between the coast luable locational information, adding to our and mountainous interior as key elements in knowledge of Solutrean settlement-subsisten- the subsistence strategies of Upper Paleoli- ce sytems. thic - Mesolithic hunter - gatherers (Freeman 1973; Bailey 1973, n. d.; Straus 1975, 1976a, Problems and Qualifications 1976b, 1977; Davidson 1976). (Notable con- cern for site location is also expressed, for Various problems must first be conside- example, by Altuna [1972], in his thesis). red. First of all -aswe are trying to con- CANTABRIA AND VASCONGADAS, 21.000 - 17.000 B. P. 197

trol for at La Riera by detailed paleoclimatic ques can liberate us from the dangers reconstruction-is the problem of variable and circularity of dating by artifacts. In any climatic conditions, since the Cantabrian So- event, the notion of «archeological cultures» lutrean spanned at least the «Laugerie» In- is at best only a heuristic classificatory de- terstadial and the Würm IV stadial phase just vice. prior to «Laugerie», as well as probably part All caveats considered, the information of the last Würm III cold phase and the very presented in Table 1 nonetheless gives us an beginning of the «Laugerie» Interstadial (La- impression of where Solutrean hunter-gathe- ville and Leroi-Gourhan personal communica- rers chose to locate at least some of their tions and m. s.). In short, environmental con- sites. ditions and, therefore, behavioral responses must have varied somewhat during even this relatively short period. Patterns of settlement Elements of the Solutrean Settlement and movement established under interstadial Pattern conditions would no doubt have been modi- The sites have been grouped by river sys- fied during stadials. Secondly, all the sites tems, as it is our belief that, logically, the we know of are caves; all the open air Solu- major rivers of Cantabria provided the prin- trean sites which no doubt once existed ha- cipal means of access between the coastal ve been either deeply buried or destroyed by plain and interior (as they do today). The erosion. Over one hundred years of prospec- alignment of Solutrean sites along such sys- tion in Cantabria have failed to locate a sin- tems is striking (see Figure 1). Such linear gle one -notsurprising in light of the steep coast-Cordillera arrangements of sites are gradients between the Cordillera and coast, most clear in the cases of the Rios Nalón, substantial rainfall and vegetation. Thus we Sella (plus Ríos Cabras and Gueña), Saja- are obviously lacking a major source of in- Besaya, Pas and Miera-Bahía de Santander. formation about Solutrean (and other) sys- The four paired central Vizcayan sites (San- tems. This bias must be kept in mind, but timamiñe and Atxeta, Atxuri and Bolinkoba), should not stop us from making use of tho- while not in the same drainage, are associa- se data from caves -excellent sediment ted with valleys (Ria de Guernica - Rio Oca traps-which we do possess (see Straus and Ibaizabal system) separated by a low 1979a). Clearly with only 33 sites, we are (c. 200 m.) pass above Amorebieta. Given at any rate dealing with a rather small sam- this patterning, it is easy to imagine other ple of what was once a large population of such alignments along the Rios Deva (Astu- sites. It should be noted, however, that so- rias-Santander) , Nansa, Asón, Deva (Guipúz- me of the sites have thick Solutrean depo- coa) and Urumea, where in each case the sits, indicating multiple deposits (at least 20 present sample of sites only includes 1-2, lo- at La Riera, for example), whereas others cated either near the river mouth or well seem to have had only one or a few brief oc- upstream. cupations. It is difficult to speak of precise distan- Thirdly, there is the serious theoretical ces for various reasons, linear (air) distan- problem of dealing with sites with no sure ces are deceptive due to the variably acci- «fossil director» artifact types. In the absen- dented terrain; so-called walking distances ce of Solutrean points or other certain con- are subjective approximations, based on our tradicting bits of evidence (other kinds of careful yet personal scrutiny of 20th century 14 «fossil directors» such as harpoons, C da- 1:50,000 maps and knowledge of the lands- tes, etc.), it is impossible oftentimes to know cape; hall distances to the coast (and eleva- if mid-late Upper Paleolithic assemblages tions) would have varied somewhat depen- might be «Solutrean», such is the degree of ding on the extent (and fluctuation) of sea- coinpositional variability among collections of level regression. Present estimates (assu- individual culture - stratigraphic units (see ming tectonic stability) made in other areas Straus 1975b, 1979b). Only the widespread run the gamut from 85 m. to 130 m. or more application of chronometric dating techni- for mull glacial conditions c. 18,000 B. P. (see 198 L. G. STRAUS

FIGURE 1. 1. Peña de Candamo. 9. La Riera. 18. El Pendo. 27. La Haza. 2. Las Caldas. 10. Tres Calabres. 19. Castillo. 28. Santimamiñe. 3. Cueva Oscura 11. Balmori. 20. La Pasiega. 29. Atxeta. (Perlora). 12. Cueva del Sel. 21. Camargo. 30. Atxuri. 4. Cova Rosa. 13. Cueva Chufín. 22. Cueva Morín. 31. Bolinkoba. 5. El Cierro. 14. Altamira. 23. Fuente del Francés. 32. Ermittia. 6. El Buxu. 15. Peña de Caranceja. 24. El Salitre. 33. Aitzbitarte. 7. Coberizas. 16. Hornos de la Peña. 25. Bona. 8. Cueto de la Mina 17. Cobalejos. 26. El Mirón.

CLIMAP 1976). We have somewhat arbitra- tabria was close in distance (5-12 km.) and rily chosen the-100 m. isobath as a reasona- outline to that of today, unlike the case of ble compromise, partly because it is marked the coast of Aquitaine c. 18,000 B. P. (see on Spanish naval charts. Some sites have no charts 127 & 128, Instituto Hidrográfico de la doubt been lost to sea-level transgression, but Marina, Cádiz). in any event the Last Glacial coastline off Can- CANTABRIA AND VASCONGADAS, 21.000 - 17.000 B. P. 199

Despite these interpretive difficulties, it the small valley-side woodlands (when pre- is clear that Solutrean occupations were si- sent). In addition, these sites would have tuated somewhat bimodally: near the present been well-placed for the exploitation of litto- coast and near the footslopes of the Cordille- ral and estuarine resources, at most an hour’s ra or Picos de Europa, both of which were lo- walk to the north, or for hunting caprids on cally glaciated at the time. There are, howe- nearby steep coastal ranges (especially in ver, some sites which seem intermediate in the case of Posada, with the Sierra de Criera location. These include, for example, Aitzbi- beginning at less than 2 km. from the sites). tarte, Cueva Morín and El Pendo. Generally, Shellfish gathering or mountain hunting ex- the near-coastal sites are at and surrounded peditions from these central spots could easi- by the lowest elevations, whereas the sites ly monitor the location of other useful plant farthest from the shore (Bolinkoba, Atxuri, and animal resources, and gather raw mate- Haza, Mirón, Salitre, Bona, Castillo, Pasiega, rials (eg. flint and quartzite) en route. Fina- Hornos, Chufín, Sel) are obviously in very lly, the coastal plain sites all have good ave- high, mountainous areas. However the distan- nues of access to the hinterlands via major ces are so relatively short in all cases as to river valleys or, at least, via low passes (as not require major treks in order to exploit in the case of Posada, with the Cabras-Gueña very different habitats. There is no evidence pass at 390 m.). Insofar as it is possible to of Solutrean occupation on the Meseta or in measure relative importance of sites (length, the Ebro Basin to the south of the Cordillera repetitiveness, multiplicity of a c t i v i t i es (with the dubious exception of Coscobilo, Na- among occupations, etc.) by such inadequate varra), so movements seem to have been con- indicators as thickness and horizontal extent fined to the narrow strip represented by our of deposits, size of extant artifact and faunal sample. Limited contacts between this region collections, etc., it seems safe to say that and Pyrenean France are suggested by the the major Solutrean occupations of Cantabria presence of a few Cantabrian concave-base were all at near-coastal sites (Cueto de la points in Basque and Pyrenean sites (see Mina, La Riera, Altamira, Aitzbitarte). Even Straus 1978). otherwise important interior sites such as Castillo had very thin Solutrean deposits and The coastal plain sites are usually located relatively small collections, generally spea- in very sheltered spots (dolinas, enclosed va- king. This may suggest that the upland sites lleys or south-facing slopes of ridges, etc.). might have been used for specialized and/or Orientations are almost uniformly toward the short-term human occupations. A possible ex- south and/or west. They are generally not ception to this might be the unpublished site exposed to sea winds. Many are low vis à vis of Las Caldas, which is however in low coun- local relief (eg. El Pendo, Cobalejos, El Cie- try. There are, of course, sites in the coastal rro, the Posada sites, Fuente del Francés, area which also suggest short, specialized oc- etcétera), and are best lit in winter when the cupations, perhaps as auxiliary stations to sun is low on the SW horizon (see Carballo nearby sites of greater scale (eg. Caranceja 1960:23). A notable exception to these norms near Altamira, Atxeta near Santimamiñe, Tres is Altamira: but in this case neither the form Calabres, Coberizas and Balmori near Cueto nor the precise orientation of the original ca- de la Mina and La Riera, El Cierro near Cova ve mouth is known. The Llera sites around Rosa). Without very precise dating, natura- Posada de Llanes, the El Carmen sites near lly, these speculations will remain just that. Ribadesella and the Santander Plain sites in The other basic group of sites -genera- particular, were chosen for their excellent lly in the interior not far from the Cordillera sheltered locations and orientation. They are and at usually fairly high elevations-bear also located centrally on what would have another set of characteristics. They do not then been even more extensive, largely open, seem to be as closely oriented with regard rolling grasslands or heaths (with localized to favorable solar exposure, but they are usua- woods and thickets) -ideally situated for the lly situated so as to have excellent dominant exploitation of ungulates grazing on these views of valleys. Most of them are in high, pastures or taking shelter and browsing in exposed positions, surrendering comfort and 200 L. G. STRAUS

ease of access for strategic placement over- depending on climatic and «economic» con- looking gorges, passes, box canyons, etc. So- ditions, on season and on anticipated objec- me combine a number of strategic features tives of each move, is suspected. such as long-distance visibility along a major A final observation may be made prelimi- valley, dominance of a defile and proximity narily concerning the clear patterning of So- to a pass between two valley systems lutrean-age sites in Cantabria. Most of the (eg. Castillo group, Atxuri, Bolinkoba). Sites alignments or clusters of Solutrean sites in- with remarkable strategic views over wide clude examples of cave art. The following ca- areas include Peña de Candamo and Altamira ves with art representations are known to con- (although the latter is near the coast). Many tain only Solutrean archeological deposits: sites are associated with box canyonsor small, steepsided enclosed valleys off major Buxu, Candamo, Chufín, Haza and Pasiega (see Straus 1975a, 1974). The following art valleys: Buxu, Las Caldas, Hornos de la Pe- caves have Solutrean and other Upper Paleo- ña, Mirón, La Haza, Aitzbitarte. Others domi- lithic deposits: Hornos de la Peña, Pendo. nate deep, narrow gorges, sometimes at con- Morín (Oso), Altamira, Castillo, Salitre, Rie- fluences of two or more valleys: Chufín, Ca- ra, Balmori, Cueto de la Mina, Coberizas, Co- ranceja, Bona, Salitre, Ermittia (the latter in va Rosa, Santimamiñe and Atxuri (see Gon- steep, rugged country, but near the usually zález Echegaray 1978). Many other art loca- abrupt Guipúzcoan coast). All of these fea- lities without any known archeological depo- tures would have been useful in spotting and sits are situated very Solutrean sites (eg. Les hunting herd animals, perhaps by such tech- Pedroses, Herrerias, Quintanal, Cullalvera, niques as communal drives and surrounds, Covalanas, Santián, Las Monedas, Las Chime- making good use of natural traps and cons- neas, La Clotilde, Las Aguas, Micolón, etc.). trictions to the movement of game. Animals The place of the Solutrean in rupestral art, moving between high and low pasture, for minimized by Breuil, has long been upheld by example, could be driven from main valleys Jordá (most recently in Jordá 1977), and stri- into side valleys, trapped (using moveable king similarities between engraved scapulae fences, nets, positioned hunters and other from the Solutrean deposit at Altamira (and persons, etc.) and killed. Or they could be the «Lower Magdalenian» of Castillo) and ru- intercepted at gorges, especially in associa- pestral engravings in several caves have long tion with river crossings. With the views most been noted, most recently by Almagro (1976) of these sites afford, the human groups would in a detailed study. Whatever artistic, symbo- have had advance warning of herd approach lic, magical or other significance the cave either by direct sighting or by signalling from art may have, the sanctuaries may well have monitoring parties. had territorial significance, as meeting pia- Despite these clear strategic advantages, ces for bands pertaining to larger endoga- most of these sites are high, windy, cold and mous, dialectic units (?) . Fission and fusion generally less protected than the coastal plain among contemporary hunter-gatherer groups cites. They would have been far less tenable is a wellknown phenomenon related to envi- during Las Glacial winters than the low, shel- ronmental, subsistence and social factors. Ma- tered sites, and may have been used for spe- te exchange, initiation and group hunting are cial purposes (base camps for hunding par- three possible major reasons for aggregation ties or actual hunting stands?) at various ti- in Last Glacial Cantabria. The art sanctuaries mes by «expeditions» from the coast, or at -inwhich the act of art was probably more least just seasonally (presumably in summer) important than the result-may have played by transhumant groups. While, because of the key roles on both pragmatic and supernatural shortness of coast-mountains distances, the planes: teaching/initiation, game manage- hypothesis of universal winter-summer trans- ment-hunt planning/magical control of animal humance may be overly simple, it deserves fertility, movements and procurement. That testing with biological indicators of seasona- no major cluster or alignment of Solutrean lity. A combination of movements to the inte- sites is without its sanctury (s) , is certainly rior by full groups and by special task groups, significant. CANTABRIA AND VASCONGADAS, 21.000 - 17.000 B. P. 201 202 L. G. STRAUS

Conclusions BINFORD. L. (1978).— Nunamiut Ethnoarchaeology The Solutrean-age inhabitants of Cantabria (New York). . n. d. Organization and formation processes: loo- utilized their nacrow, zonally diverse environ- king at curated technologies. In Lithic Analysis in ment in full, systematic fashion. Each major Archaeology, R. Tringham. ed. (New York) (in coastal plain sector (especially when near press). probable inlets or estuaries and entry points CARBALLO. J. (1960).— Investigaciones Prehistóricas II to the interior) and each major river valley (Santander). system contains sites. The coastal sites seem CLIMAP (1976).— The surface of the Ice-Age Earth. usually to have been chosen for their shelte- Science 191: 1131 -37. red and central locations. The hinterland si- COHEN, M,. (1977).— The food Crisis in Prehistory tes (and certain particular near-coastal sites (New Haven). where a true plain is lacking) were clearly DAVIDSON. I. (1976) .—Seasonality in Spain. Zephyrus selected for strategic reasons related proba- 26-27:167 -73. FREEMAN,L. (1973) .—The significance of mammalian bly to the hunt, with apparently less regard faunas from Paleolithic occupations in Cantabrian for comfort. It should be said that caves are Spain. American Antiquity 38:3-44. so numerous in heavily karstic, limestone GONZALEZ ECHEGARAY, J. (1978).—Cuevas con arte Cantabria, that we are dealing with a real rupestre en la Región Cantábrica. In Curso de Ar- case of choice among mony possible alterna- te Rupestre Paleolítico (Santander). tive occupation loci on the part of Solutrean JORDA. F. (1977).— Historia de Asturias I: Prehistoria and other Upper Paleolithic groups. (Oviedo). LAVILLE. H. n. d .—Preliminary analysis of the sedi- The problems which need to be resolved ments from La Riera Cave. Unpublished manuscript. now involve the theoretical scale (area per LEROI-GOURHAN, Arl. n. d .—Preliminary analysis of group) of hunting territories for the exploi- the pollen from La Riera Cave. Unpublished ma- tation of mobile red deer and ibex (plus hor- nuscript. se and bison), the development of a precise STRAUS, L. G. (1974-5) .—Posible atribución al Solu- Solutrean chronology based on detailed re- trense del yacimiento de La Pasiega. Ampurias construction of climatic oscillations and on 36-37:217-23. radiocarbon dating, and the ascertainment of . (1975a) .—A Study of The Solutrean in Vasco-Can- the seasonality of individual occupations. tabrian Spain. Unpublished doctoral dissertation, When this is done, then we will begin to University of Chicago. approximate an understanding of settlement . (1975b) .—¿Solutrense o Magdaleniense inferior cantábrico? Significado de las diferencias*. Bo- patterns in this interesting period. letín del Instituto de Estudios Asturianos 86:781-90. Acknowledgements . (1977) .—Of deerslayers and mountain men: Pa- leolithic faunal exploitation in Cantabrian Spain. Thanks are owed to my American, Basque In For Theory Building in Archaeology, L. Binford. and Spanish colleagues for exchanges of ideas ed. (New York). over the past 7 years of my involvement in . (1978) .—Thoughts on Solutrean concave base of which has been funded by the National point distribution. Lithic Technology 6:23-25. Science Foundation (USA). . (1979a) .—Caves: a paleoanthropological resource. World Archaeology 10 (3): 331-339. . (1979b) .—Variabilité dans les industries solu- REFERENCES tréennes de l’Espagne cantabrique. Bulletin de la Seciélté préhistorique française 76: ALMAGRO BASCH. M. (1976) .—Los Omoplatos deco- . (n.d.) .—Early excavations in the cave of El Ras. rados de la Cueva de «El Castillo». Museo Arqueo- cario. Unpublished chapter of the Rascaño site lógico Nacional, Monografías Arqueológicas 2 (Ma- report, J. González Echegaray, ed. drid). STRAUS, L., F. BERNALDO DE QUIROS, V. CABRERA & BAILEY, G. (1973) .—Concheros del norte de España: G. CLARK (1978) .—Solutrean chronology and lithic una hipótesis preliminar. Actas del XII Congreso variability . Zephyrus 28-29: 109 -1 2. Nacional de Arqueología, pp. 73-84 (Zaragoza). VITA-FINZI, C. & E. HIGGS (1970).-Prehistoriceco - . (1979) .—The Early Postglacial Settlement of nomy in the Mount Carmel area of Palestine. Site Northern Europe. In The Mesolithic of Northwest catchment analysis. Proceedings of the Prehistoric Europe: Social Europe, P. Mellars. ed. (Cambridge). Society 36:1-137.