During the Quijos Valley regional survey a total of 2,256 collections were carried out within the 137 km² survey region. Of these, 792 contained at least one obsidian artifact, resulting in an obsidian assemblage of 2,476 artifacts. In addition to obsidian, local cobbles of fine-grained basalt and rhyolite were used for similar tasks as obsidian. In total, 42 artifacts made from these materials were recovered from 30 collections.

The survey obsidian assemblage was classified into three color categories (Table). In general, artifact types vary little between black and clear obsidian. Artifact dorsal surfaces indicate that the majority, if not all, of obsidian utilized within the survey zone came in the form of river cobbles, some with a heavily battered cortex and others with lightly scuffed cortex. Expedient reduction flakes dominate the assemblage and comprise all of the grey obsidian recovered. Flake tools, however, make up a much larger proportion of black obsidian artifacts than clear, which likely relates to the larger mean size of black obsidian cobbles available in the survey zone, as discussed below. Overall, clear obsidian made up 62.8% (n=1,556) of all obsidian artifacts, while black obsidian comprised 36.8% (n=910) and grey obsidian .4% (n=10). Although clear obsidian comprises a much larger proportion of artifacts, the overall weight of these artifacts is much less than that of black obsidian (Table). In fact, black obsidian comprised 62.5% of the cumulative weight of all obsidian recovered, while clear obsidian only 37.2%. The greater weight for black obsidian likely reflects the abundance of black obsidian cobbles, 10-30 cm in size, from the Aliso-Pumayacu obsidian source. Although the black obsidian from this secondary source is not a high quality material, it is serviceable for the production of informal flake tools. Clear obsidian samples, on the other hand, are of a higher quality, but were only observed as small cobbles during the source survey. It is likely that clear obsidian was only available to consumers in small sizes, thus the correspondence in small artifact size.

Density maps of the survey zone showing the distribution of raw and mean obsidian weight based on collection units were generated. High concentrations of obsidian artifacts are common in the southern arm of the survey zone. The largest concentration, from Lote 1166, appears to have been a secondary refuse pit or cache, since high quantities of broken ground stone artifacts, especially axes, also were recovered there (see the Raw Obsidian Weight Density Map). Outside of the southern arm, Lote 121, located on a northern terrace of the Río Quijos across from the modern town of Baeza, and Lote 966 located on a terrace of the northern branch of the Río Quijos close to the town of Borja, also contained high concentrations of obsidian artifacts. When obsidian quantity is standardized and the mean artifact weight per collection is investigated, a different visual distribution is evident. The distribution of mean artifact weight per collection indicates that only those high density areas in the southern branch of the survey zone correspond to the areas of high obsidian artifact count. This suggests that in the north, west and central portions of the survey zone high raw weight counts represent concentrations of smaller, thinner flakes, whereas high mean artifact weight counts represent concentrations of larger artifacts. Incidentally, these are the same areas that yielded the majority of non-obsidian groundstone and chipped artifacts (See map of Collection Units with Non-obsidian Chipped and Groundstone Artifacts).

To investigate this pattern further, a simple stem-and-leaf plot of mean obsidian artifact weight per collection unit across the survey zone was produced. All outliers from this batch (cases ≥ 6.3 g, n=75) were plotted, and the results revealed a concentration of the heaviest, and thus largest, obsidian artifacts in the southern arm. Black obsidian comprised 63% (n=138) of these outliers, while 37% (n=80) were clear. To investigate artifact distribution by weight from another angle, the location of all collection units containing obsidian artifacts with a mean weight in the lower quartile (cases ≥.1 and ≤ .7g, n=196) was plotted. Lower quartile collection units concentrate in the northern and western arms of the survey zone, farthest from the southern source area. Black obsidian comprised 22% (n=100) of the lower quartile counts, while clear comprised 78% (n=363), and less than 1% (n=1) was grey. Therefore the largest artifacts are primarily of black obsidian and they are located closest to the Aliso-Pumayacu source, while the smallest artifacts are primarily clear and located the farthest from the source area.

The correlation between color, size and proximity to the Aliso-Pumayacu obsidian source also explains the large proportion of black flake tools. Black casual cores make up almost half of all flake tools (Table), whereas flakes used as cores comprise the largest proportion of clear flake tools. The availability of larger black obsidian cobbles across the survey area, but especially in the southern arm, explains the greater proportion, and mean weight, of black casual cores. As discussed above, there comes a point when the removal of flakes from a casual core by direct hard hammer percussion is no longer viable due to the size limits of the core. At this point, the core is either reduced further with bipolar reduction or discarded. Since black obsidian was relatively plentiful, there was less pressure to reduce black casual cores completely via bipolar reduction. Clear obsidian, being less common and of a higher quality, would have been more valuable and therefore, it was reduced beyond the limits of a casual core. Relatively few clear casual cores were found because they were being reduced into smaller artifacts via bipolar reduction. The relatively high proportion of clear flakes used as cores also reflects this material maximization behavior. Large clear flakes were being reduced for additional tools, even though they were, on average, smaller than black obsidian flakes. As shown on the flake tool type by color breakdown, clear flakes used as casual cores were, on average, 3.5 times lighter, and thus smaller, than the same artifacts made from black obsidian.

The variety of artifact use wear in the heaviest and lightest obsidian artifacts attests to the functional differences between them. The difference between the types of use wear on the heaviest artifacts (the batch outliers from the stem-and-leaf plot) and those from the collection units containing obsidian artifacts with a mean weight in the lower quartile are very significant (Χ² = 25.74, .99 > p > .999). The observed deviations from expected values were greatest for use wear types #3 (truncated flaking) and #10 (rounded and dull). Scraping actions on hard surfaces like wood, shell, bone, and antler can produce truncated flaking. Rounding and dulling of the edge can be produced by slicing activities on extremely hard surfaces like ceramics and stone. Slicing of soft materials, like flesh or hide could not be accurately identified on the tools using the macroscopic method employed. Additionally, the specific types of scraping activities that result in either truncated flaking or overlapping edge flaking is unclear. Nonetheless, small clear obsidian flakes were favored for the specific scraping action on hard surfaces that resulted in truncated edge flaking, while large black obsidian artifacts were favored for the cutting and slicing of very hard materials like ceramics, thick shell, and stone.

The breakdown of use wear types on black and clear obsidian is also very telling. Overall, use wear was identified on 31.6% (n=782) of all obsidian artifacts: 32% (n=293) of all black obsidian artifacts, 31% (n=484) of clear obsidian artifacts, and 50% (n=5) of all grey obsidian artifacts. These represent 923 use wear events, since 140 artifacts contained more than one type of use wear. Four types of use wear comprised the majority of the forms observed: rounded and dull, crushing, overlapping microflaking, and truncated microflaking. Crushing of the artifact edge and edge abrasion are associated with a cutting or slicing action on a hard surface, while microflaking is associated with scraping. Approximately 52% (n=475) of all use wear was produced by scraping activities, while 18% was produced by cutting on a hard surface. Artifacts with scraping use wear were recovered throughout the survey zone (Map). In addition, scraping was the activity most likely to occur with artifacts that also were used for other activities. Looking at the obsidian assemblage as a whole, the difference between the types of use wear on all clear and black obsidian artifacts is significant (Χ² = 22.04, .95 > p > .98). The observed deviation from expected values is greatest for clear obsidian with use wear type #9, continuous lunate flake removals, which is produced by continuous slicing.

To summarize the obsidian assemblage taken as a synchronic whole, two main patterns emerge. The first pattern is that, not surprisingly, the largest obsidian artifacts were recovered from collection units closest to the Aliso-Pumayacu source. These artifacts were primarily made of black obsidian, which corresponds in color to the larger cobbles of obsidian identified throughout the source area. Conversely, the smallest obsidian artifacts tended to be located farthest from the source area, and were primarily made from clear obsidian. This likely corresponds to the smaller average cobble size of clear obsidian. Beyond this simple difference, there was little to no noticeable pattern of obsidian material or artifact clustering at any one location at the expense of other locations, except for a potential dump or cache centered on Lote 1166 in the southern arm of the survey zone. In other words, no discernable differentiation in access to obsidian material or artifact type that could be attributed to sociopolitical hierarchy was evident. Differential access to obsidian appears to have been a product of relative proximity to the source.

Secondly, scraping was the most common activity carried out with all artifacts, while cutting and slicing the second most common. Variations also were observed in wear patterns within these two general activity categories, corresponding to artifact size. Larger black obsidian artifacts were preferred for the cutting of very hard materials, while small clear obsidian flakes were preferred for specific, as yet unidentified, scraping activities.

Return to "The Quijos Valley Chipped Stone Assemblage" Main Page