83Chapter 4

Kidder, Hooton, Pecos, and the Birth of Bioarchaeology

Lane Anderson Beck

I. INTRODUCTION

Bioarchaeology, put simply, is the contextual analysis of human populations from archaeological sites (Buikstra, 1977). It uses skeletal biology and archaeology in combination to ask questions not about how people died, but about how they lived. It does this through focusing on the osteobiography of individuals and the biocultural adaptations of populations as viewed through the lens of archaeological context. Although use of the term bioarchaeology is relatively recent, the precepts of the field have deep roots in American archaeology.

In 1930 a report by E. A. Hooton on the people of Pecos Pueblo was published. Hooton’s emphasis on the analysis of human remains in reference to their archaeological context emerged, not as a tentative step, but as mature, integrative form of analysis. This project, when examined in detail, reveals a partnership between A. V. Kidder and E. A. Hooton as pioneers in developing an integrated, interdisciplinary perspective on the past. As Schwartz (2000:19) emphasizes, “Hooton’s work on the human remains was significant, for Kidder was realizing that the only way he was going to obtain the essence of the settlement’s cultural development was by using insights from a wide range of other social, natural, and environmental disciplines. This multidisciplinary approach to his archaeology became a centerpiece of Kidder’s research design.”

II. THE PECOS EXCAVATIONS

In 1915 the Department of Archaeology of Phillips Academy, Andover, began excavations at Pecos Pueblo, carried out under the direction of A. V. Kidder. 84As the site had been occupied continuously for a period of several hundred years, one of Kidder’s primary objectives was to identify temporally sequential cultural units through the analysis of ceramic and stratigraphic data (Kidder and Kidder, 1917; Kidder, 1924; Hooton, 1930). His success in this endeavor remains a landmark in the history of American archaeology. Using his chronological sequence, Kidder assigned temporal associations to over 2000 Pecos burials, enabling Hooton to investigate changing patterns in demography and disease over time. This is perhaps the largest series ever recovered from a scientifically excavated, stratified site in the New World (Kidder, 1924). Early in his excavations of the middens at Pecos, Kidder actively sought burials:

Some human bones had been found on the surface, and a few had come from the digging. We were most anxious to discover burials; so a reward of twenty-five cents was offered to the workmen for every skeleton uncovered. The next day one appeared, the following day six; the reward was reduced to ten cents; this brought fifteen more, and in the course of a week or so we were forced to discontinue the bonus or go into bankruptcy. The higher we got uphill the deeper grew the rubbish and the more crowded became the skeletons. (Kidder, 1924:94)

As the second season at Pecos began, Kidder discovered that burials were not limited to the midden areas on the sides of the mesa, but were also located throughout the mesa top, amid all the structures. As the number of interments expanded, Kidder recognized that he needed a physical anthropologist to step in and assume responsibility for burial analyses. Kidder believed that an osteologist must begin analysis in the field and not merely wait in the laboratory for burials to arrive (Kidder, 1924). As a result, he arranged to have Earnest Hooton, the physical anthropologist for this project, assist in excavation as well as in laboratory analysis. Hooton joined the field crew for 2 months during the 1920 season (Hooton, 1930).

Skeletons were subsequently shipped from Pecos to Boston. One of the first shipments was mistakenly delivered to the Peabody Museum rather than to the warehouse. During delivery, the crates were tossed from the truck, down the steps, and into the museum’s basement. Many of the crates broke open. Perhaps this circumstance led to Hooton’s complaints to Kidder about recent damage to the bones. Kidder reports that Hooton felt:

… I had been kicking about my skeletons. He said they had an awful lot of fresh breaks on the bones. In the Southwest, a bone will often crack and not come apart. When you take it out, it comes in two pieces and, it looks like a fresh break but, it isn’t a fresh break. I tried to explain that to Earnest, but that didn’t do any good, so I said, “You come out and dig some skeletons yourself.” So he did. Then I discovered that Earnest had done practically no excavation at all. He had worked a little in a long barrow and then he had been to the Canaries and worked in a cave but as far as digging skeletons, in bad conditions, he knew very little about it. He would clean out a long bone and put his knife under it and pry and the damn thing would break. It was very interesting having him there, because he gave us a lot of information about 85the age of children, the dentition, and he made out a whole lot of tables for us, of one sort or another. (Givens, 1992:141)

The tables referred to here were slates of standards for the determination of age and sex. Both Hooton and Kidder state that the in-field assessments for the later years at Pecos became very close to matching Hooton’s analysis in the laboratory (Hooton, 1930).

Just as Kidder’s excavations at Pecos are of major significance in the development of American archaeology, so is Hooton’s analysis a landmark in American physical anthropology. Hooton was among the first to explicitly use archaeological context as a guide to the questions he asked. This enabled him to raise intrasite research inquiries rather than being limited to total sample as the unit of analysis. Speaking in 1935, Hooton described the Pecos project as a turning point in physical anthropology that allowed research to go beyond a “mere description of bones” and facilitated studies of change within a population. He went on to speak of “the necessity of an intimate cooperation of the archaeologist with the physical anthropologist” (Hooton, 1935:503):

In the pre-war period the first research efforts of a physical anthropologist attached to a museum were likely to be studies of skeletal remains deposited by archaeologists as a result of their excavations. The job of the physical anthropologist was to describe these remains and to make some sort of a racial diagnosis. Usually the archaeologist prepared and published his report without any reference to the skeletal finds. Most were so conscious of their virtue in preserving the bones that they considered their scientific responsibilities fully discharged when the skeletons had been dumped in a museum. The present writer undertook several such tasks, mostly relating to the bones of American Indians. From them he learned the folly of dissociating excavation reports from the study of the skeletal material which they produce. One example will suffice. The Peabody Museum excavated a large Indian cemetery at Madisonville, Ohio, in spasmodic efforts beginning in 1882 and ending in 1911. It devolved upon this unfortunate to study the bones. In order to make such a study intelligible, he was forced to spend an entire summer struggling with the field notes and records of three generations of archaeologists who had worked the site. He had to patch together by collation and speculation some sort of consecutive account of the excavations. All evidence as to the relative ages of the different portions of the cemeteries had been lost, and stratigraphy was absent or unrecorded.

The physical anthropologist had to content himself with a consideration of the remains as of one period. Apart from the mere description of the bones, the only advance in anthropological method resulting from this effort was a fairly successful attempt to deduce the size of the population and its probable annual death rate from an examination of the proportions of each age and sex represented in the skeletal material. (Hooton, 1935:501)

Pecos was the first major archaeological sample to be so fully studied. The quality of the field notes combined with the good preservation of the human remains enabled Hooton to apply contemporary, standard approaches to skeletal analysis and to pioneer new methods. He combined demography, pathology, 86morphological, and metric data to examine changes in a community over time — time as defined by Kidder’s work on the archaeological context.

In his analysis of the people from Pecos, Hooton departed from the mainstream of skeletal biology. Although he did measure skulls and generate typologies, he did not stop there. Working closely with data generated by Kidder and his own observations made in the field, Hooton was among the first to seriously examine questions from the perspective of the archaeological context. Instead of focusing on the site as the generalized, single unit of analysis, Hooton subdivided the sample, utilizing Kidder’s chronology, and asked questions about how a human community had changed over time. In the 1920s this was not a routine procedure but instead a highly innovative approach.

III. LABORATORY ANALYSIS OF THE PECOS COLLECTION

Rather than simply saying that Hooton’s publication on Pecos Pueblo is a landmark study, one should take a detailed look at just what this report includes and how that relates to the scientific foundations of its time. All laboratory observations of the Pecos skeletons were made by Hooton, with the exception of cranial capacity estimates, which he assigned to two of his assistants.

Hooton begins the Pecos volume with a report of the excavations that summarizes their extent at the time of publication and provides an overview of the significance of the site and the work being done there (Chapter 1, pp. 3–13). He explicitly states that the work is still ongoing and that this analysis includes only the burials excavated by the end of the 1924 field season.

The first analytical portion of the Pecos report deals with post-depositional changes, “state of preservation” (pp. 14–15), which we would today term taphonomy. Hooton clearly recognized that the demographic pattern of the recovered burials was somewhat skewed and discussed the possibilities that such factors as age and sex of the deceased, as well as microenvironmental factors, could have led to differential bone preservation. He emphasized that the bones of infants and young children were less fully ossified than those of adults and that the bones of the elderly may be relatively thin and more porous than those of other adults. Presaging more recent studies of taphonomy and demographic bias (e.g., Walker et al., 1988), he also noted that women’s bones are generally smaller and lighter than those of men. He further reported that certain features of specific graves may alter the patterns of preservation and that the drainage of different types of soils and microenvironments created by burial associations can create situations of better or worse preservation. This is a very early report of taphonomy and its implications for burial analysis (see also Wilder, 1923).

87Following these cautionary notes on potential biases, Hooton proceeded to reconstruct demographic patterning (pp. 16–32). He began with the estimation of a mortality profile. At the date of this research, the first detailed standards for assessment of age-at-death from skeletal remains were just beginning to emerge. For example, T. Wingate Todd at Washington University in St. Louis was assembling an anatomical collection that included skeletal remains with documented sex and age-at-death. From this work, Todd proposed a series of pubic symphyseal phases that could be used to estimate age-at-death (Todd, 1920). Hooton contacted Todd to assist with the Pecos Pueblo analysis. Todd provided Hooton with a series of photographs that illustrated the age changes in the pubic symphysis (Hooton, 1930:21).

In addition to his own assessment of age-at-death and sex, Hooton also arranged for Todd to personally assess age and sex for the remains from Pecos (Hooton, 1930:18; Todd, 1927:494). Field estimates had also been recorded on each burial feature form. In analyzing age and sex data for Pecos, Hooton compared the three profiles, thus explicitly addressing the issue of interobserver error, a pioneering effort (Hooton, 1930:18).

Following his overview of paleodemography, Hooton compared the Pecos mortality profile to national death rates from various countries (pp. 24–25). Although it may seem odd that Hooton chose to compare Pecos to European data, it must be remembered that this was the first study of a large sample from a single archaeological site. There were no other well-provenienced North American series with which Hooton could compare Pecos. As a result of these and other analyses, Hooton emphasized juvenile underenumeration, which he interpreted as an artifact of the archaeological context rather than as a measure of community health (p. 24).

Next Hooton assessed cranial deformation patterns (pp. 33–39). He first categorized cranial deformation by form and degree and then examined temporal sequences for systematic changes over time. Patterning was interpreted in terms of ethnographic reports on cradleboarding as well as studies of infant behavior. He concluded that variability in form and degree of cranial deformation resulted from an interaction of infant behavior and skull shape. In his scenario, doliocephalic infants tended to rotate their head slightly toward one side, whereas brachyocephalic infants are more likely to lie flat. Hooton also proposed that the greater tendency toward flattening of the right side of the back of the skull in doliocephalic infants was related to handedness and a tendency of the infant to face toward its dominant side (p. 38).

Having evaluated the degree and form of cranial deformation, Hooton was able to begin his analysis of craniometric data (pp. 38–78). In order to include as many skulls as possible, Hooton had one of his graduate students, Harry Shapiro, statistically evaluate measurements of the deformed and undeformed crania from Pecos. Through correlation and regression analysis, Shapiro developed 88a correction formula that facilitated analysis of all measurable crania. Hooton generally reported measurements for both deformed and undeformed separately due to what he viewed as the probable imprecision in the correction formula (pp. 38–39).

Following 19th-century traditions, Hooton reported copious amounts of craniometric data. While a century earlier Morton (1839) had developed a suite of 10 cranial measurements for his research, Hooton used an expanded list of 29 measurements and 9 calculated indices. For each sex, means and standard deviations are reported for the total series and for the four temporal groups defined by Kidder’s field records. Within each temporal division, the measurements were subdivided into deformed and undeformed categories, resulting in a total of 61 tables and 35 plates that illustrated typical patterns. Hooton reported only summary statistics because he planned to publish a supplement containing all raw data. He concluded that the earliest population samples at Pecos were more variable than the later occupants of the pueblo.

The volume of craniometric calculations is remarkable, with the total number of summary statistics involving nearly 1000 data sets. In 1935, 2 years before the first computers became available, Hooton reported that he had just purchased several electronic calculators to facilitate his biostatisitics work at Harvard (Hooton, 1935). Thus, the Pecos statistics, generated during the 1920s, must have been calculated by hand. Furthermore, the remaining chapters in the book often contained large numbers of calculations. The table of contents describes 362 statistical tables, in addition to 26 figures and 97 photographic plates. While there was no attempt to apply multivariate approaches, such as the newly available coefficient of racial likeness (Pearson, 1926), Hooton’s emphasis upon making his data available to others set a high standard for future studies.

Following his section on skull measurement, Hooton presented information on cranial morphology (pp. 80–132). He observed 31 morphological characteristics, scored by form and degree of expression. These were then sorted by sex and time. Dental variants were also included: molar cusp variation for both arcades, degree of incisor shoveling, and a variety of other dental anomalies. Also within this section were data on dental wear, dental eruption, caries, abscesses, and antemortem tooth loss.

Some morphological attributes included features we now use for other purposes, such as the degree of dental wear. Others are morphological variables standardized today for estimation of sex, such as size of the mastoid process, while some are among those included in listings of discrete traits generally collected today for biological distance analyses. As with the measurements, Hooton provides detailed statistical examination of each feature as reported in 1 or more of 66 tables. He also noted that there is a “certain spurious correlation” in certain traits due to the traits sometimes being those used for identification of sex (p. 131), an issue we take seriously in statistical analyses today (e.g., Konigsberg and Buikstra, 1995).

89Continuing his detailed analyses, Hooton next turned from the skull to the postcranial skeleton (Chapter 5, pp. 133–184). He presented data on a bone-by-bone basis, with metric data and morphological observations both reported. For example, Hooton reported seven femur measurements and two indices, sides reported separately and partitioned by sex and temporal assignments. He reported changes over time, with the earliest groups being larger than the later groups. He also commented on patterns of bilateral asymmetry overall and in comparisons of male and females.

Morphological features of the postcranial skeleton are also reported. For example, six sets of morphological data characterizing the femur are presented and partitioned. Data range from description of the size of the linea aspera to observations on squatting facets. He noted that squatting facets are more common among males than among females and that the frequency in females diminishes over time while that for males is stable. After completing his report on the femur, including 15 summary tables, Hooton proceeded to offer similar analysis for the tibia, fibula, humerus, radius, ulna, clavicle, scapula, pelvis, and lumbar portion of the spinal column.

Calculations of stature followed Hooton’s discussion of morphology (p. 178). Employing a weighted combination of femur and tibia lengths, Hooton used Pearson’s formulae as the basis for his estimates, with males and females considered separately. Recognizing the limitations of these formulae when applied to Native American contexts, he argued that, even so, they were the best method available at that time. Results are reported for the pooled sample, as well as for the four temporal divisions partitioned by sex. While noting a slight decrease in stature for the more recent samples, Hooton remarked that the change was slight and may have been an artifact of sampling bias. He then compared the Pecos stature estimates to those for a variety of living populations and observed that the people of Jemez Pueblo were very close to estimates for Pecos (pp. 178–180).

Following this analysis of osteometric and morphological data, Hooton was still dissatisfied with the high degree of heterogeneity he found at Pecos and his inability to identify consistent temporal trends. At this point he shifted his reference point from the archaeological record to the morphological (Chapters 6–8, pp. 185–288).

Nevertheless it is apparent to the craniologist that the skeletal population of Pecos was at no time markedly homogeneous in type. On the contrary the handling and measurement of the Glaze subgroups leaves the impression of a number of markedly diverse cranial types, found in varying proportions at all periods. Therefore upon the conclusions of the study of period groups it was decided to reanalyze the material, relying upon morphological rather than upon archaeological criteria for the differentiation of groups. (Hooton, 1930:185)

In developing this approach, Hooton attempted to move beyond the comparisons typically made at that time and to generate a more complex method, one approaching multivariate analysis.

90To establish his morphological types, Hooton chose 129 skulls of adult males. He placed them on a laboratory table and then grouped them visually into homogeneous subsets. His final sorting resulted in eight morphological types, which he named according to what he saw as gross patterns of distinctive features (pp. 185–186). If he had simply referred to these as morphological groups rather than types and had numbered them rather than naming them, this aspect of his work would be less vulnerable to recent attributions of racism (Armelagos et al., 1982; Armelagos and Van Gerven, 2003).

While Hooton’s language concerning race and morphology is very much a product of its time, he emphasized that the names he assigned his groups were somewhat arbitrary. In naming his first, “Basket Maker” type, he notes “a general resemblance, perhaps fancied, to the veritable Basket Makers of the Arizona caves” (Hooton, 1930:185). He then emphasized that the “second type was styled ‘Pseudo-Negroid,’ not because of any theory of the observer as to the presence of a Negroid strain in the American Indian” but rather because of certain features reminiscent of those typically attributed in those days to individuals of African descent (Hooton, 1930:185). Following statistical validation of his types and in the context of alternative models for peopling of the New World, Hooton does, however, link type resemblances to heredity. He argues, for example, in a discussion of the “sequence of immigrant types, especially at Pecos” that the presence of the “Pseudo-Negroid Type” was due to an ancient admixture prior to the migrations from Asia (p. 356). He further states that the craniometric validation of morphological types reflects “the segregation of features in occasional individuals,” not the migration of distinctive groups from homeland regions (p. 357). His was clearly a typological approach uninformed by population genetics (see Chapter 2, by Cook, this volume).

After creating these types through visual sorting, Hooton then attempted to test their validity through observations of cranial deformation and craniometric comparisons. He thus concluded that six of his eight groups are internally more homogeneous than any of the temporal segments described in earlier chapters. One of the remaining two groups included some burials from the Pecos church, which are probably Spanish rather than Indian. He also compared stature across these types and again found homogeneity greater than that for the temporal divisions. As always, Hooton attempted to make his data fully available to the reader and included 31 photographic plates to illustrate his types. From these photographs, it is clear that several of the crania he chose to illustrate appear to be female rather than male, e.g., plates VI-4 and VI-6. Given that these fall primarily within a single type, coupled with the fact that stature sorted well across types, it would appear that a key variable in Hooton’s typology was relative size — a visual parallel to the first component of principal components analysis.

91Having confirmed these types through craniometric analysis, Hooton then considered variation in cranial morphology and chronology. Three of his types dominated the earliest time periods and were absent or relatively underrepresented in the later phases. Two types reverse this trend and are dominant in the later periods. One group is most common in the middle periods. The final two groups are consistent in frequency over time. In other words, Hooton’s typological analysis did identify temporal patterning that was not otherwise evident.

After a comparison of Pecos craniometric data with that from other sites, which proved inconclusive, Hooton turned to intrasite correlations of metric and morphological data (Chapter 9, pp. 289–305). He reported, for example, that dental wear and abscesses were highly correlated. Such close examination of correlation among variables did not become prominent in osteological studies until after the rise of computer-aided multivariate analysis during the second half of the 20th century. Hooton was again decades ahead of the profession.

The next chapter, which focuses on paleopathology (Chapter 10, pp. 306–330), is most commonly cited as Hooton’s landmark contribution to bioarchaeology (Jarcho, 1966a; Ubelaker, 1982). Prior to 1930, paleopathology tended to focus on diagnoses of obviously deformed remains, often divorced from archaeological contexts. Hooton’s study was both sensitive to population dynamics and the archaeological context. Both Jarcho and Ubelaker argue, for example, that Hooton’s Pecos Pueblo report is the precursor to later paleoepidemiological approaches (Jarcho, 1966a:22; Ubelaker, 1982:342). For each category of disease that he or his medical collaborators identified, Hooton discussed the frequency with which it was observed and also discussed the pattern of its distribution in the Pecos sample in terms of age, sex, and temporal horizon. He also consulted on individual cases with at least eight medical doctors, including both radiologists and pathologists. Table X-11, which summarizes the pathological observations, was in fact compiled by Dr. G D. Williams, with supplemental notes from other physicians (Hooton, 1930:305).

On page 305, Hooton indicates that he took notes on diseased bones as he conducted his morphological and metric surveys. He indicates that these notes will appear in Appendix III, but unfortunately there is no such appendix in the volume.

Hooton divided his discussion of pathology into four gross categories: arthritis, inflammatory lesions, trauma, and miscellaneous. Within the category “arthritis,” he separated spinal osteophytosis from degenerative joint disease. He noted that the frequency for both forms was somewhat lower in the earliest time periods and that the majority of the cases involve older adults. He also pointed out that many of the cases of arthritis in the long bones are associated with fractures.

92Under the heading “inflammatory lesions,” Hooton discussed periostitis and osteomyelitis in both the cranial and the postcranial skeleton. Three cases of possible syphilis were submitted to two physicians, James Ewing of Cornell Medical College and H. U. Williams of the University of Buffalo. Ewing said none of the cases were syphilis, while Williams said all three cases were probably syphilis. Hooton concluded that the evidence was inconclusive. One of the cases is from an apparent historic context.

Hooton reported that postcranial fractures were most common in the femur and humerus. The frequency was greater in older people and occurred on the right side more often than on the left. He found the frequency of cranial trauma highest in mature to older adult males. Fractures of any sort occurred more frequently in the most recent time period. Hooton concluded that the pattern suggested injuries to males in conflict and that the recent intervals might reflect an increase in violence (p. 315).

Hooton’s classic discussion of the condition he called “osteoporosis symmetrica” appeared under the heading of miscellaneous pathology. He also included cribra orbitalia as an early and milder stage of the disorder and claimed that it was clear that whatever produced these lesions was limited to childhood and adolescence. His description of this condition is worth repeating.

The typically honeycombed condition which involves a hyperostosis of the diploe and a destruction of the external table of compact bone, frequently extends to the parietals, where it may be seen in symmetrical patches sometimes extending over the greater portion of both bones and causing a thickening of ten to fifteen millimeters in the middle portion of the area affected. On the base of the skull it may be observed in the form of numerous small pits on the palatine roof and on the wings of the sphenoid. Traces of the condition may also be observed in some crania on the temporal bones just above the auditory meatus. The most pronounced osteoporotic conditions, in my experience, are found in the crania of immature subjects from ancient Peru and from the Sacred Cenote of Chichen Itza in Yucatan. (Hooton, 1930:316)

He also provided further details on active versus inactive conditions, their general appearance, and their radiographic attributes. One point that is often overlooked is that most of Hooton’s discussion of osteoporosis symmetrica does not focus upon remains from Pecos Pueblo. The crania he submitted to the doctors for radiographic analysis and morphological examination were instead from Chichen Itza, a Maya site in the Yucatan Peninsula. At Pecos Hooton felt that he saw traces of this condition, which appears in a more extreme expression in the Cenote collection (p. 316).

Within his discussioin of osteoporosis symmetrica appears information seldom acknowledged by subsequent scholars. He reported that Dr. Percy Howe of Forsyth Dental Infirmary showed him crania from monkeys fed on a scorbutic diet and that the crania of these monkeys exhibited a pattern highly similar to that in the Cenote collection (p. 317). He noted that the condition is restricted 93to bones that form intramembranously rather than cartilaginous bones. He also pointed out that the lesions were restricted to the diploe and outer table of the skull in regions outside areas of muscle attachment. His medical consultants both suggested that, if dietary, the condition looks most like scurvy or rickets but that it most closely resembles congenital anemias that we today would call thalassemia. On returning the discussion from Chichen Itza to Pecos, Hooton noted that the frequency there was approximately 3%.

Other lesions noted in the Pecos collection included Pott’s disease (tuberculosis) in one historic and one pre-contact set of remains. A case of cancer from the earlier part of the archaeological sequence is also noted in a young adult female.

The final analytical chapter in the main body of the Pecos report reconstructs changes in population dynamics over time (Chapter 11, pp. 331–343). In so doing, he explicitly divided his discussion into known, deduced, and unknown factors. Known factors included population estimates recorded in the historic literature and the number of burials excavated for each ceramic period. The two deduced variables involved chronology: the founding date for Pecos and the chronology for the ceramic sequence. The final group of variables Hooton described as “unknown factors.” These included the annual death rate and the percentage of total burial area that was excavated. He then presented various calculations for estimating population size at different times in the pueblo history that most closely match his known factors. His critical review of this work was once again expressed with total candor.

In this chapter I have built up a house of cards. The assumptions made and the methods employed are all questionable, perhaps erroneous. The reader need not attach much importance to this effort, nor rely at all upon its conclusions. I have merely attempted to reach a plausible solution of an impossible problem. (Hooton, 1930:340)

At the end of the complex chapter on population estimates, Hooton and Kidder each inserted an addendum (pp. 342–343). Tree ring dates by A. E. Douglass had just arrived, which reduced the duration of occupation at Pecos by 300 years. Hooton found this reduction astounding. His original estimates had focused on historic reports of a community of over 2000 at Pecos and he felt that the reduced time frame elevated this number to an unreasonable figure. He acknowledged that part of the problem was finding a way to deal more precisely with issues of growth.

The summary chapter for the Pecos book provided an encompassing overview (Chapter 12, pp. 344–363). Hooton cautioned the reader about potential imprecision and then turned toward the larger issues of what Pecos revealed about peopling of the New World. Here he drew on contemporary data about early man in the New World, as well as recent studies of blood group distributions. He concluded that humans arrived in the Americans not by one migration but by several, with the Eskimo being the most recent arrival before Euro-Americans.

94IV. HOOTON AND PECOS TODAY

Far beyond the case studies or summary reports of that era, Hooton asked questions about populations and how they change over time. He used every method available to him and presented innovative approaches to the study of population health and demography. The labels Hooton used for his morphological types have received considerable recent criticism, even though Hooton himself regarded these labels as somewhat arbitrary.

Hooton’s analysis of Pecos remains today one of the most comprehensive presentations of bioarchaeology ever generated. The breadth of data reported, the statistical comparisons both internally and externally, and the variety of experts consulted all form a model for analysis that has seldom been achieved. That his interpretations would be modified by the application of more refined observations or multivariate analysis of his original data does nothing to diminish the tremendous contribution of this landmark research.