Internet Archaeol. 5. Walker & Gibert

3.2 Untoward methodological consequences

Untoward methodological consequences for sampling strategies stem from steps taken to meet technical challenges at our sites. They strongly affect how our archaeological findings should be interpreted and therefore must be outlined before results are presented, lest misunderstandings arise.

Dry-sieving many tonnes of miners' rubble in loose soil at Sima de las Palomas has produced a disproportionately high number of unstratified palaeolithic, palaeoanthropological, and palaeontological finds (>10 mm), compared with the results of 50 days of scientific excavation of 3m³ at the top of the breccia column, using wet-sieving retrieval of small skeletal items and flint chips (>2mm) and saving all struck limestone flakes. In undisturbed sediment, these imply Pleistocene knapping, whereas irregular struck limestone flakes from disturbed rubble (without adhering breccia or retouch, and not Levallois/pseudo-Levallois points) could well result from iron-miners' picks and hammers swung at dislodged limestone blocks or at the cave and mine-level walls; hence limestone may be under-represented in the total collection (Figure 7). All chips, fragments, nodules and flakes of rocks brought to the site by the Pleistocene inhabitants are kept: flint, chert, rock crystal, quartz, and reddish-brown quartzite (seen occasionally in local gravels); oddly, the hill's ubiquitous greyish-white micaceous quartzite was ignored by Pleistocene knappers.

Figure 7: Numbers of items of different rock types at Cueva Negra and Sima de las Palomas.

Cueva Negra presents a different sampling problem. It was formed by karstic processes: erosive water seeps in through solution holes in the roof and drains away through a solution fissure beside the west wall. It lies in Upper Miocene (Vindobonian) sandy limestone containing very fragmented bioclasts - too crumbly for knapping (unlike Cabezo Gordo where metamorphism toughened its Triassic limestone, dolomite and marble).

Cueva Negra's knappers brought pebbles of frangible, calcareous chert, or hard, fine-grained, metamorphosed siliceous dolomite and limestone, or sometimes flint or quartzite. These all occur 800m away on the same hillside (along with complete Pectinid and Ostroid shells), in a 50m diameter conglomerate outcrop in Vindobonian (Tortonian) calcarenite immediately beneath the strata wherein Cueva Negra was formed. A post-orogenic, late Tertiary pebble beach formed here when Tethys Sea waves washed metamorphous nodules out of a former sea cliff of Jurassic Lias dolomitic limestone (cf. Baena et al. 1973).

A recurrent centripetal disc core and other knapping débris come from the conglomerate outcrop. Here, Pleistocene folk took nodules, blanks, and worked artifacts, back to Cueva Negra. They never took any pebbles of ultrabasic and basic rocks, schists or slates, washed down by the Quípar from its distant headwaters among mountains characterized by igneous and metamorphous rocks. Such pebbles, together with limestone cobbles, appear below Cueva Negra in the eroded sides of glacis-terrace B. However, because during the early Pleistocene its surface formed a flood-plain across the valley, continuous with the cave floor, those gravels then lay buried, several metres deep, unknown and therefore unsought. Aggradation occurred when more loess and silt were deposited than swampy rivers could remove from ice-age flood-plains (cf. Cuenca and Walker 1986). Corroboration of nearby swamps comes from several species of wildfowl among sixty bird species excavated.

Minute chips of frangible calcareous chert and hard fine-grained limestone at Cueva Negra are collected in the sure knowledge that they originated from stones that were carried there. We have excavated intact and split pebbles, and tests on unstratified pebbles from the cave and cobbles from the conglomerate outcrop reveals far more of both are of hard fine-grained limestone than of chert, flint or quartzite. Many pebbles were bashed open for one-off use of sharp fragments of limestone or chert: few have obvious striking platforms or bulbs of percussion of struck flakes, and retouched pieces are rare because great skill is needed to knap these intransigent minerals (as experiments by both ourselves and Derek Roe have shown).

Figure 8: Cueva Negra: numbers of classifiable types from units 2 and 3 at the end of the 1995 season (only 6m² of unit 3 have been excavated).

Because these limestone and chert cobbles shatter into innumerable bits, the task of interpreting our findings is further complicated when raw data from both our caves are reviewed (Figure 7, Figure 8). Even though different amounts have been excavated of the undisturbed sedimentary units (2 and 3) at Cueva Negra (more of 2, less of 3), nevertheless their relative incidences show similar rank order, whether of unclassifiable to classifiable pieces, or of separately classifiable artifact types (differences with regard to incidences of palaeolithic finds from disturbed superficial soil, unit 1, reflect greater use of coarse dry-sieving for clearing unstratified dirt here). In units 2 and 3, proportions are about the same of classifiable (around 10%) to unclassifiable items (around 90%), whereas in unit 1 the ratio is roughly 35%:65%. For similar reasons, the total Sima de las Palomas collection shows a comparable, if even closer, ratio - where lithics excavated at the top of the breccia column are still too few for separate statistical treatment.

Lithic incidences at Cueva Negra (vis-à-vis Sima de las Palomas) were also influenced by: (a) the greater proportion of items recovered to date by systematic trowelling and wet-sieving, (b) a longer period of excavation, (c) the presence of a larger and deeper area of undisturbed sediments, (d) the lack of heavy rubble, allowing use of 8mm or 6mm meshes when dry-sieving unit 1, and (e) the presence in its loose soil of pieces thrown out when the Civil-War hideaway and pits were dug into units 2 and 3.

Awkward consequences must be faced of having, on the one hand, a largely dismantled, 18m high column with barely 2m² accessible to scientific excavation at Sima de las Palomas (and only 3m³ removed), and on the other an excavated area of 25m² whose lower levels remain a closed book at Cueva Negra. For good or ill, the contrast is more important than it might otherwise have been because of particularly archaic hominid fossils found at Sima de las Palomas that seem likely to have come from very deep levels, which excavation of the column will not reach for many years to come.

This goes beyond a matter merely of theoretical implications setting a practical challenge where the only response possible is condemned to unfold in frustratingly slow motion. There are, alas, more immediate practical consequences which are indeed awkward. They stem from an invidious situation in which we find ourselves that is manifestly not of our making: namely that the iron miners left behind more unstratified hominid fossils, whose intriguing morphology demands publication, than the stratified ones we have so far excavated systematically. Meanwhile, a few palaeoanthropologists are on the prowl whose squeamishness about putting unstratified hominids under a spotlight before those next 25 years are up which will see our excavation down to the lower levels, brings to mind Æsop's beady-eyed fox, here peering downwards rather than upwards!