Although certain problems were encountered while this study was being conducted, their impact does not significantly alter the conclusions presented here. Three main difficulties can be outlined:
Sites were assigned to specific 'eras' solely on the basis of recorded pottery. Often, these assignations were given by the original survey directors, and these were taken at face value and incorporated into the database. For the remainder of sites, however, era determination was based on the relative frequency of black-on-white, red-on-buff, corrugated and plain pottery. The major problem with this is that, very often, while the proportions or counts of each type were reported for each site, descriptions of specific motifs were rarely included. This meant that it was often hard to precisely categorize sites, since it was impossible to examine the original material itself. Nonetheless, since the majority of assignations were done by the original directors, the labels used here are probably still representative of distinct temporal phases.
For those cases where no pottery data were available, the label 'Unknown' was used and those sites were considered along sites for each era, in the eventuality that they outlined interesting patterns.
Different databases were used in compiling the sites used in this study, (1) the CARP survey data; (2) the more detailed SARG survey data, and; (3) the AZSite database compiled by the US Forest Service. All these databases contained sites included in at least one of the other two, and to minimize the potential impact of duplication on the total number of sites used here, site numbers and their location were cross-examined between the databases. When sites numbers could be matched, the more up-to-date AZSite locational data was retained.
If it was impossible to establish a match between databases on the basis of site numbers, sites belonging to the SARG or CARP databases that fell within 50 meters of an AZSite record were considered to be one and the same. This practice was supported by the fact that, in many cases, the distribution pattern of CARP and SARG sites slightly off the AZSite coordinates often demonstrated the same configuration. For instance, if the triangle formed by three sites in the CARP database was matched by that formed by three similarly disposed (although slightly 'off' in metrical terms) AZSite sites, this was taken as evidence that the same sites were included in both databases. Again, when this duplication was observed, the ASZite coordinates were retained.
In all instances of site duplication, all of the data available were combined in the 'theme table' used here. It can be supplied to any interested parties by the author, upon request and provided that acknowledgment of sources is made.
Because of these problems, the interpretation derived from patterns observable in the dataset used here have to remain tentative. Some interesting patterns nonetheless emerged and warrant further discussion.
| Introduction | Methods | Analysis & Interpretation | About me! |