Content uploaded by Thomas Jennings
Author content
All content in this area was uploaded by Thomas Jennings on Jan 27, 2019
Content may be subject to copyright.
11
Here we present analyses of serrated projectile-point blades to identify evidence of evolu-
tionary convergence in stone tools using an empirical case study from the prehistoric record
of the Southern Plains–Woodland border in central North America. Point-edge serration
was introduced in this region during the Late Paleoindian period (12,850–11,700 B.P.),
variably used throughout the Archaic period, abandoned by the Middle Woodland period
(2,100–1,500 B.P.), and introduced again in the Late Woodland to Early Mississippian/Late
Prehistoric period (1,300–650 B.P.). We focus on the evolutionary relationship between
the earliest serrated Late Paleoindian Dalton and the Late Prehistoric Scallorn point types
and explore how a cultural-evolutionary approach can help identify convergent evolution
in the stone-tool record and consider the behavioral situations in which convergence arose.
First, we use cladistic analysis to generate a phylogenetic tree that shows hypotheses of
relatedness, or, in this case, change within lineages that does not reflect ancestry. Second,
we evaluate morphological similarities and differences using geometric morphometric
analysis. Third, we consider point evolutionary trajectory and morphology as a response
to adaptive challenges that caused two populations with at best dim ancestral connections
to converge on the same tool design.
Background
Serrations are a series of isolated tooth-like projections on a tool margin (figure 11.1). In
prehistoric stone-tool production, serrations were applied during the final steps, requir-
ing the flintknapper to intentionally modify the edge to isolate portions of the margin.
Because serration is an obvious feature of artifact form, it is often used as a diagnostic
trait to define tool types.
The Study Area
The Southern Plains–Woodland border encompasses several physiographic provinces
(Fenneman and Johnson 1946; see figure 11.2). The area was occupied repeatedly
The Convergent Evolution of Serrated Points on the Southern
Plains–Woodland Border of Central North America
Ashley M. Smallwood, Heather L. Smith, Charlotte D. Pevny,
and Thomas A. Jennings
© Massachusetts Institute of TechnologyAll Rights Reserved
11554_011b.indd 203 11/17/2017 6:14:23 PM
204 Ashley M. Smallwood and colleagues
a
bc
Figure 11.1
Examples of serrated-point types compared in this study: (a) Dalton, (b) Kimberley, and (c) Scallorn.
Harrell
Tortuga Flats
Graham-Applegate Unprovenienced
points
Brand
0 240 480120
Nmi
Figure 11.2
Map of the study area showing the location of sites that yielded serrated points used in this analysis.
11554_011b.indd 204 11/17/2017 6:14:24 PM
The Convergent Evolution of Serrated Points 205
throughout prehistory, from ca. 15,500 B.P. on, but most extensively after about 13,000
B.P., beginning with the well-documented Clovis complex (Waters and Stafford 2007;
Waters et al. 2011). The region includes Dalton and Scallorn “heartlands,” where the
earliest and latest dated serrated-point types have been found, and it encompasses an area
where Dalton and Scallorn points geographically co-occur.
The Evolution of Serrated Point Technology: Late Paleoindian Dalton Points
In North America, serrated bifacial points first occur in the portion of the archeological
record associated with the Late Paleoindian Dalton period (Anderson and Sassaman 2012).
Dalton is associated with a date range of ca. 12,500–11,300 B.P. (Goodyear 1982). Com-
ponents with Dalton points have been radiocarbon dated at sites throughout the Eastern
Woodlands (e.g., Dust Cave and Stanfield-Worley, both in Alabama), and dates consis-
tently demonstrate that Dalton postdates the Clovis complex and predates or co-occurs
with Early Archaic side-notched complexes (Anderson, Smallwood, and Miller 2015;
Miller and Gingerich 2013; Sherwood, Driskell, Randall, and Meeks 2004).
The Dalton complex spans the latter part of the Younger Dryas, a climatic episode
dated to ca. 12,850–11,700 B.P. and marked by a reversal of general warming trends
and a return to glacial-like conditions (Anderson et al. 2015). During the Younger Dryas,
biotic communities underwent a major reorganization, but the impacts of such changes
on Paleoindian populations are regular topics of debate (e.g., Anderson and Bissett 2015;
Eren 2012; Meltzer and Holliday 2010). In a continental-scale extinction event, more
than 30 genera of mammals disappeared by the first part of the Younger Dryas (Grayson
and Meltzer 2015; Haynes 2009), including many species of megafauna on which Clovis
hunters relied. The extinction of these large herbivores allowed for the competitive release
of medium-bodied mammals (e.g., deer), requiring an adaptive shift in hunting strategies
among later populations, including Dalton (Koldehoff and Walthall 2009). It is in this
context that serrations first entered the archeological record.
Dalton flintknappers crafted lanceolate points with concave bases, often with basal
thinning (Bradley 1997; Morse 1971). Dalton-point blades are variable in form. Some are
serrated along the lateral margins above the hafted area; others are serrated and beveled;
and still others have no obvious modification beyond sharp edges. Less frequently, Dalton
points exhibit burins or are blunted and rounded at the distal end. Functional hypotheses
explain the variation in blade morphology as changes produced from knife use and pro-
gressive dulling, resharpening, and reworking (Goodyear 1974), or as forms deliberately
crafted for varied hunting needs (O’Brien and Wood 1998). Both approaches recognize
Dalton points functioned, either intermittently or exclusively, as dart points, likely thrown
with an atlatl (Goodyear 1974; O’Brien and Wood 1998).
Dalton points occur throughout the Eastern Woodlands, from the Atlantic Coast in the
East, south along the Gulf Coastal Plain, north to the Upper Mississippi River valley,
11554_011b.indd 205 11/17/2017 6:14:24 PM
206 Ashley M. Smallwood and colleagues
and west into parts of the Central Lowlands (Anderson et al. 2015). Some of the best-
documented Dalton sites are located along the Southern Plains–Woodland border. Because
of the density and nature of Dalton sites in northeastern Arkansas, that area is referred to
as the “Dalton Heartland” (Koldehoff and Walthall 2009). Variation in Dalton assemblages
from sites such as Sloan, identified by Morse (1997) as a Dalton cemetery, and Lace
Place, a possible long-term camp, have been used to model Dalton settlement (Ballenger
2001; Gillam 1996; Morse 1971; Schiffer 1975). Although these models disagree on the
degree of logistical versus residential mobility, as well as drainage versus cross-drainage
landscape use, they elude to underlying shifts in Paleoindian lifeways that emerged
during Dalton times. As post-Pleistocene hunters, Dalton populations exploited a variety
of resources, with an inferred emphasis on deer, and by the Late Paleoindian period,
Dalton territory size apparently decreased, which likely required more-intensive use of
local resources.
After the disappearance of Dalton, point types with serrations occurred variably through-
out the Archaic period in the study area. However, during the subsequent Woodland period
in the Southeast and the Late Archaic period in Texas, serrations decreased in frequency
and became essentially absent by the Middle Woodland period, 2,100–1,500 B.P. That
would soon change with the introduction of the bow and arrow.
Late Woodland and Mississippian Period Technological Transitions and
Scallorn Points
The emergence of small, narrow-stemmed triangular and triangular notched points is often
attributed to the technological shift to the bow and arrow (Thomas 1978). Although the
timing of the adoption and nature of the spread of the bow and arrow varied regionally
(see Nassaney and Pyle 1999 for a panregional overview), most researchers agree that by
the Late Woodland period, ca. 1,300 B.P., bow-and-arrow technology was the dominant
weapon system in the Eastern Woodlands (Blitz and Porth 2013; Nassaney and Pyle
1999).
Investigations of the transition from dart to arrow points consider changes in point
variation and point size as a means to understand the change in projectile technology.
Blitz and Porth (2013), who date the adoption of the bow in the Eastern Woodlands to
approximately 1,700–1,600 B.P., identify a significant size-reduction threshold associated
with Lowe-cluster points dated to the Middle Woodland–Late Woodland transition. This
size reduction is attributed to the alteration of dart-point design for initial bow technology.
Further, the authors conclude that a later reduction in thickness among smaller, lightweight
triangular arrow points—forms such as Jack’s Reef and Hamilton/Madison, which date
to the Late Woodland–Mississippian transition, ca. 1,400–1,000 B.P.—represents a tech-
nological refinement.
11554_011b.indd 206 11/17/2017 6:14:24 PM
The Convergent Evolution of Serrated Points 207
Roughly 11,000 years after the introduction of serrated Dalton points, a new type
of finely serrated projectile emerged coincident with the spread of the bow and arrow.
Originally typed on the basis of specimens from Texas and named “Scallorn Stemmed” by
Kelley (1947), Scallorn points are among the earliest dated arrow points in the region and
have been identified at sites throughout the Mississippi Valley and the Midwest (Anderson
and Smith 2003; McGahey 2000; O’Brien and Wood 1998). In Central Texas and along
the Texas coast, Scallorn points have been radiometrically dated to between 1,270 B.P.
and 650 B.P. (Lohse, Black, and Cholak 2014; Ricklis 2004a).
Scallorn points, considered temporally diagnostic of the Late Prehistoric Austin interval
(Collins 2004; Prewitt 1981), are small, triangular, corner-notched points with straight to
convex lateral edges and pronounced barbs (Turner, Hester, and McReynolds 2011). Stems
can be straight or expanding, and bases are straight, convex, or concave. Fine serrations
extend along the blade edges. Asphaltum has been noted on the stems of Scallorn points
(Huebner and Comuzzie 1992), and its presence is inferred to be an adhesive to haft the
points to shafts. Abraded and grooved “shaft straighteners” are also indications of this
new technology (Hester 2004).
Scallorn points have been recovered from archeological sites throughout the study
area. The eastern half shares environmental and cultural affinities with the rest of the
Southeastern Woodlands. In this area, Scallorn points were contemporaneous with the
rise of the Caddo culture ca. 1,200 B.P. and associated with the use of pottery, increased
sedentism, population growth, organizational complexity, intense reliance on agriculture,
and a panregional, ceremonially based interaction network (Anderson and Mainfort 2002).
The western portion of the study area is environmentally heterogeneous and includes
savannah, prairies, plains, and desert. Annual rainfall steadily decreases to the west. Agri-
culture, if present at all, came late; varied depending on precipitation; and was not very
important compared to what occurred in the Eastern Woodlands and on the High Plains.
In Central Texas, Late Prehistoric Scallorn-point makers continued to live a hunter-gather
lifestyle, and settlement and subsistence patterns were a continuation of adaptations of
the preceding Late Archaic period (Collins 2004). Thus, the distribution of Scallorn
points does not appear to coincide with a particular ecological niche or subsistence
strategy.
Materials
The samples we examined include Dalton and Scallorn points recovered from buried
contexts at sites that contributed to defining the characteristics of the Late Paleoindian
Dalton period and the Late Prehistoric Austin interval. These samples were limited to
points that have complete bases, serrated edges, and a minimum of 7 millimeters of blade
length.
11554_011b.indd 207 11/17/2017 6:14:24 PM
208 Ashley M. Smallwood and colleagues
Goodyear (1974) recovered a Dalton assemblage from the Brand site (30PO139) located
in northeastern Arkansas. A total of 305 Dalton points were reported, and 77 of these have
serrated blades. Goodyear explored variation in Dalton-point morphology from Brand and
the functionality that may have created the varied point forms. Based on the high incidence
of serrated points and the occurrence of other tools suggested to be used in bone and antler
working, Goodyear (1974) interpreted Brand as a hunting camp used for butchering and
processing deer. The large number of serrated points makes the Dalton assemblage from
Brand an excellent sample for the earliest evidence of serrated points. For this paper, we
reanalyzed and digitized 30 serrated points to compare blade-edge shape.
A sample of Scallorn points was analyzed from Graham-Applegate Rancheria
(41LL419), located in central Texas (Hixson 2003). The Austin-interval component con-
tains large burned-rock features and five houses, each of which has a large central hearth
and stone circles and pavements representing foundations and floors. Although the lithic
assemblage has not been fully analyzed, it contains dozens of Scallorn points (Hixson,
personal communication, 2016). For this study, we used 13 of the serrated points.
The Harrell site (NT-5), located in Young County in north-central Texas, was an Austin-
interval habitation site with a well-defined cemetery, suggesting an extended period of
occupation (Fox 1939; Hughes 1942). Numerous hearths were found in the upper few
feet of the refuse midden. Digging tools suggest Harrell inhabitants practiced farming, but
very few macrobotanical remains were recovered. Of the 555 points recovered, we used
six serrated Scallorn points.
Tortuga Flats (41ZV155) is located in south-central Texas (Hester and Hill 1975; Hill
and Hester 1973; Inman, Hill, and Hester 1998). Small mobile groups seasonally used
the site. Archaeological materials, including hammerstones, exhausted cores, biface thin-
ning flakes, other flakes and flake fragments, and preforms indicate tool production and
resharpening activities took place at the locale. Food processing occurred in at least one
area where a metate fragment was identified, and a refuse area contained the remains
of bison, antelope, deer, coyote, and rabbit as well as discarded tools. Thirteen Scallorn
points were collected from surface and excavation contexts, and we used a sample of
four points.
We supplemented our sample of point types for the cladistics analyses to ensure a robust
phylogenetic tree with point types representing temporal and geographic variation. The
additional points were found in published sources that covered sites throughout the study
area—Arkansas, Louisiana, Oklahoma, and Texas—and represent types ranging in time
from the Early Paleoindian to the Mississippian/Late Prehistoric periods.
The geometric morphometric analyses focused on a subsample of points, including those
from Brand, Graham-Applegate, Harrell, and Tortuga Flats. Points from these sites were
recorded, analyzed, and photographed firsthand by the authors. To highlight the similarities
and differences in the two point types, a sample of unequivocally unrelated serrated points
was also included for comparison—seven serrated Kimberley points collected in 1902
11554_011b.indd 208 11/17/2017 6:14:25 PM
The Convergent Evolution of Serrated Points 209
from the region East Kimberley, Australia. Kimberley points are found in archeological
and ethnographic contexts in northwest Australia and are suggested to date from 1,500 to
1,000 B.P., with production continuing into the nineteenth century (Akerman, Fullagar,
and van Gijn 2002). These pressure-flaked bifacial points have very fine serrations along
the blade margin (Akerman et al. 2002). The ethnographic record shows Kimberley points
tipped dart spears used for hunting (Akerman et al. 2002); residues on these points suggest
they were also used as knives (Lommel 1997). Akerman et al. (2002) suggest that they
were also important commodities in trade and exchange networks.
Methods I: Cladistic Analysis
The dataset included 349 points spread over 50 previously defined point types (chapter 12,
this volume). Rather than evaluating the classifications of all the point types in the dataset,
our focus is on understanding the evolutionary relationship between Dalton and Scallorn
points and demonstrating evolutionary distance between these serrated point types. We
incorporated 12 Dalton points from the Brand site, three Scallorn points from Tortuga
Flats, and five Scallorn points from Graham-Applegate. The remaining sample of point
measurements came from published sources collected from images using the program
ImageJ (Schneider, Rasband, and Eliceiri 2012).
The character set contained two descriptive traits and six continuous measures. Follow-
ing O’Brien and Lyman (2003), continuous data were condensed into categorical groups.
The divisions were arbitrary and relative to the data spread for each variable. Because size
variation can potentially distort cladistic comparisons, continuous data were size-adjusted
following Lycett, von Cramon-Taubadel, and Foley (2006). The geometric mean was
calculated for each point using four measures: maximum length, base length, maximum
base width, and maximum blade width. The geometric mean is the fourth root of the
product of these four measures. For each point, all six continuous character measurement
values were then divided by the points’ specific geometric mean to yield the size-adjusted
values (see chapter 12, this volume). These calculated values became characters 3–8 in
the cladistic analyses.
Table 11.1 and figure 11.3 show the eight characters and character states. Although
point-edge characteristics, including serrations, were likely influenced by evolutionary
pressures related to tool design and intended use or other factors, the presence or absence
of serrations was not included as a character in the cladistic analyses. This ensured a
degree of independence between the trait of interest—serrations—and the reconstructed
phylogenetic tree(s).
The dataset of 349 points and eight characters was analyzed using PAUP* 4.0 (Swofford
1998), following methods described by O’Brien et al. (2014). Clovis points, the oldest
dated points in the dataset, were selected as the outgroup (see arguments for this approach
11554_011b.indd 209 11/17/2017 6:14:25 PM
210 Ashley M. Smallwood and colleagues
Table 11.1
Point characters and character states or description of values used in the cladistic analysis
Variable type Character Character state or description of value calculation
Descriptive Base type Lanceolate; side notched; corner notched; straight stemmed;
contracting stemmed; expanding stemmed; basal notched
Proximal base shape Concave; at; convex
Continuous* SA base length Maximum length of the point base (i.e., the portion of the point
that would have been in the haft)
SA base width Maximum width of the point base
SA concavity Calculated using the depth of basal indentation from the proximal
towards the distal tip (at or convex based points received values
of zero)
SA length/blade
width
Calculated using the shape ratio of maximum point length divided
by the maximum blade width
SA basal constriction Calculated by dividing the maximum basal width by the minimum
basal width
SA blade width/base
width
Calculated using the shape ratio of maximum blade width divided
by maximum base width
*Continuous characters were size-adjusted (SA) using the geometric mean of each individual point in the
database.
Side-notched Corner-notched
Straight-stemmed
Contracting-stemmed Expanding-stemmed
Lanceolate
Basal-notched
A
Aʹ
B
Bʹ
Cʹ
C
DʹD
Eʹ
E
Fʹ
F
A-Aʹ = Maximum length D-Dʹ = Minimum base width
B-Bʹ = Maximum base length E-Eʹ = Maximum base width
C-Cʹ = Maximum blade width F-Fʹ = Concavity
ab
Figure 11.3
Schematics showing point characters and character states or description of values used in the cladistic analysis:
(a) diagram illustrating base-type character states and (b) diagram illustrating how continuous variables were
measured.
11554_011b.indd 210 11/17/2017 6:14:25 PM
The Convergent Evolution of Serrated Points 211
in chapter 12, this volume). A heuristic search using the principle of parsimony was then
performed using 1,000 replicates. From these, a single, 50-percent majority-rule consensus
tree was produced. Because the consistency index can be affected by the number of taxa
(Sanderson and Donoghue 1989), we used the retention index (RI) as a goodness-of-fit
measure to assess tree support (Collard, Shennan, and Tehrani 2006).
Methods II: Geometric Morphometric Analyses
Geometric morphometric analyses were performed on two datasets combining Dalton,
Scallorn, and Kimberley points. First, an analysis was conducted on complete serrated
points (n = 50): 21 serrated Dalton points from Brand; 23 serrated Scallorn, includ-
ing points from Harrell, Tortuga Flats, and 15 unprovenienced points donated to Texas
A&M University; and six Kimberley points. Second, an analysis was performed on 60
fragmented serrated projectile points: 30 serrated Dalton points from Brand; 23 serrated
Scallorn points from Graham-Applegate, Harrell, and Tortuga Flats; and 7 Kimberley
points. The first dataset was analyzed to compare blade morphology, assessing the entire
serrated distal-end shape. The second dataset was used to analyze an isolated segment of
serrations along the blade margin, magnifying the serrations to compare the morphology
on a finer scale.
Dataset One
A landmark approach to geometric morphometric shape analysis was used to assess
variation of serrated-point blade morphology in plan view (figure 11.4a). The three point
types considered here were complete specimens; however, their base shapes varied and
prevented identification of landmarks that could represent uniform morphological features
across the sample. Such corresponding landmarks are often used as homologous land-
marks in Procrustes superimposition to align specimens horizontally along the X-axis in a
Cartesian coordinate system (Bookstein 1991; Rohlf and Slice 1990; Zelditch, Swiderski,
Sheets, and Fink 2004).
To compare three different point types with heterogeneous base shapes, a different
approach was used to align the data clouds along the X-axis, averaging the slopes of the
distal lateral margins around the X-axis (following Smith and DeWitt 2016). To digitize
artifact images, tools were positioned horizontally with basal margin to the left in digital
photographs, and a constellation of type II semilandmarks was placed along each artifact’s
perimeter in tpsDig2 (v. 2.12; Rohlf 2008a). Landmark constellations were reduced to a
suite of 400 semilandmarks that consisted of outlines made of 200 equidistant semila-
ndmarks assigned to each lateral margin, defined as the length of the blade edge from
the distal tip to its confluence with the basal portion. This position was identified by the
highest and lowest coordinate along the Y-axis, with points positioned horizontally along
11554_011b.indd 211 11/17/2017 6:14:25 PM
212 Ashley M. Smallwood and colleagues
the X-axis. Inadvertently, this position was also found to represent the location where ser-
rated edges ceased in most specimens. Outlines representing the proximal portion of the
artifacts—the proximal portion of each lateral margin as defined above—were deleted to
focus analyses on blade shape. The resulting semilandmark density was more than suf-
ficiently saturated to capture serrated-edge shape differences.
Generalized least-squares Procrustes superimposition (generalized Procrustes analysis)
was conducted in tpsRelw (v. 1.45; Rohlf 2008b) to superimpose the constellations of
corresponding semilandmarks (Rohlf and Slice 1990), translating each constellation to the
200 semilandmarks
7 mm from geometric center
200 semilandmarks
200 semilandmarks
1 cm
a
b
Figure 11.4
Illustration of landmark approach to geometric morphometric shape analysis in plan view, showing landmark
distribution for assessing variation in serrated-point blade morphology (a) and for assessing variation in isolated
serrated edges with 7 millimeters of lateral blade margins (b).
11554_011b.indd 212 11/17/2017 6:14:25 PM
The Convergent Evolution of Serrated Points 213
same centroid location, scaling each constellation to the same centroid size, and iteratively
rotating each constellation until the summed squared distances between the semiland-
marks and mean semilandmark position was minimized (Bookstein 1991; Mitteroecker,
Gunz, Windhager, and Schaefer 2013; Rohlf 1999). Superimposed semilandmark constel-
lations (Procrustes shape coordinates) were subjected to principal component (PC) analysis
(Adams, Rohlf, and Slice 2004, 2013; Bookstein 1991; Mitteroecker et al. 2013). Centroid
size, the square root of the summed squared distances between all semilandmarks to their
common centroid, served as an unbiased size variable (Bookstein 1991) in analyses of
shape and form (e.g., Smith, Smallwood, and DeWitt 2014).
Multivariate analysis of variance (MANOVA) was used to test models of morphologi-
cal homogeneity by testing variance in shape among artifacts organized by point type. In
MANOVA, shape analysis used size as a covariate to characterize and statistically control
for linear allometry (Smith et al. 2014). Principal components of shape variation were
also used to visualize shape characteristics that represent major factors of variability in
the samples. Statistical analyses were conducted using JMP software version 10 (SAS
Inst. Inc., Cary, NC).
Dataset Two
To identify variation in the shape of serrated edges, 7 millimeters of lateral blade margins
were isolated on 59 specimens of Dalton, Scallorn, and Kimberley points (figure 11.4b).
Isolating a section of the blade margins allowed fragmented specimens to be included; the
same semilandmark density (200 semilandmarks) was used to saturate the small section.
Specimens in the dataset were digitized using the same procedure described above except
that, before landmark constellations were reduced, points in excess of 7 millimeters
(determined by the length of the longest fragment) from the geometric center of each
artifact were deleted to standardize the length of the serrated edges and to isolate lateral-
edge shape from the morphology present at the transition from base to blade. Landmarks
placed along the lateral-edge segments were reduced to 200 type II semilandmarks, and
generalized least-squares Procrustes superimposition and PC analysis were conducted as
described above.
Results
Cladistic Analysis
The heuristic search returned 1,000 equally parsimonious trees. The 50-percent majority-
rule consensus tree has a retention index of 0.86, which falls well within the range of
values for cladograms produced using cultural datasets (Collard et al. 2006), indicating
strong tree support (chapter 1, this volume).
11554_011b.indd 213 11/17/2017 6:14:25 PM
214 Ashley M. Smallwood and colleagues
All Dalton points in the study sample fall within a single clade (figure 11.5), the ances-
tral node of which is characterized by lanceolate points with concave bases. Points in this
clade have relatively low values (character-state group values of 1 or 2) for all metric
size-adjusted shape characters.
Scallorn points occur in multiple clades. Eleven of the 18 Scallorn points fall in clades
that also contain Ellis points, which are transitional, terminal Archaic/Woodland forms
that date ca. 2,150–1,270 B.P. (Lohse et al. 2014) and are often classified as dart points.
Other Scallorn specimens fall occur in clades with Motley points, a dart point dated to
ca. 3,650–2750 B.P. (Anderson and Smith 2003), and other later arrow points, including
Alba, Bonham, Cuney, and Friley. No Scallorn points fall within clades that also contain
Dalton points, which we would expect, given the millennia between them.
Geometric Morphometric Analysis
Multivariate Shape Analysis for Dataset One: Complete Serrated Projectile Points
The first four PCs explain 95.09 percent of variability in the dataset (n = 50). Figure
11.6a illustrates shape characteristics expressed at the positive and negative ends of the
PC axes. Models were organized by point type, and tests of blade shape found strong dif-
ferences among Dalton, Scallorn, and Kimberley points (p = 0.001). Results illustrated as
canonical centroid plots show that Scallorn and Kimberley samples overlap in shape space,
demonstrating that blade shape in these two point types have more in common with each
other than either has with Dalton (figure 11.7). Dalton points are described by the negative
loading of PC1 and positive loadings of PC3 and PC4, which describe blade shapes that
are long relative to width and have incurvate lateral margins.
Multivariate Shape Analysis for Dataset Two: Isolated Serrated Margin
Although variability in blade shape supports a typological difference between Dalton and
Scallorn, or perhaps different blade functions, the shapes of serrations on the isolated blade
margins are indistinguishable. The first six PCs explain 92.64 percent of variability in the
dataset (n = 60). Figure 11.6b illustrates shape characteristics expressed at the positive
and negative ends of the PC axes, demonstrating variation in margin shape. The shapes
of serrated edges are not significantly different among point types (p = 0.22), suggesting
that the variability observable in each PC axis is present within each type.
Discussion
Results of the cladistic analysis provide support for the hypothesis that Dalton and Scal-
lorn point serrations reflect an example of convergence among separate projectile-point
lineages. The consensus tree shows that serrated points do not form a single evolutionary
11554_011b.indd 214 11/17/2017 6:14:25 PM
The Convergent Evolution of Serrated Points 215
Angostura.8
Angostura.10
Angostura.11
99.8
Clovis
Clovis.2
Clovis.4
Clovis.5
Clovis.8
Golondrina.3
Clovis.3
Clovis.7
Angostura.9
Clovis.10
Clovis.11
Clovis.13
Clovis.14
Clovis.15
100
Golondrina
Golondrina.2
Gower
SanPatriceHopeSJ.4
99.2
96.6
Clovis.9
Clovis.16
Clovis.17
100
Dalton
SanPatriceHopeSJ
99
Dalton.6
Dalton.10
Dalton.12
SanPatriceHopeSJ.5
99.4
Dalton.7
Dalton.8
Dalton.9
Dalton.11
Golondrina.4
98.3
Clovis.12
Dalton.2
Dalton.3
Pelican
SanPatriceHopeSJ.2
SanPatriceHopeSJ.3
98.6
Dalton.4
Dalton.5
Darl
94
Angostura.7
Bulverde.2
Bulverde.6
Bulverde.7
Delhi.2
Evans.4
Lange.3
Lange.4
98.8
Kent.4
Scottsblu.4
98.5
Kent.12
Lange
Lange.2
99.4
Lange.5
Yarbrough.6
Yarbrough.8
97.2
Evans
Godley
Godley.2
Godley.6
Yarbrough.4
99.7
Godley.3
Godley.4
Yarbrough.2
99.8
Godley.5
Godley.7
Godley.8
Godley.9
Godley.11
Palmillas
Palmillas.2
Yarbrough.3
97
Evans.2
Evans.3
Lange.6
Sinner
Sinner.2
Sinner.4
Sinner.9
Yarbrough
98.1
Sinner.3
Sinner.5
Sinner.6
99
Sinner.8
Yarbrough
.5
Yarbrough.7
Yarbrough.9
94
90.2
Angostura
Angostura.2
Angostura.5
Angostura.6
99.8
Angostura.3
Angostura.4
99.2
Godley.10
Godley.12
Palmillas.3
Palmillas.4
Palmillas.5
Perdiz
99
Sinner.7
Bassett
Bassett.2
Bassett.5
Bassett.10
Bassett.11
Perdiz.8
100
Bassett.12
100
98.8
90.9
Bassett.3
Bassett.4
Bassett.6
Bassett.7
Bassett.8
Bassett.9
98.8
Perdiz.6
Perdiz.7
90.2
92.5
Clifton
Clifton.2
100
84
Perdiz.3
Perdiz.4
100
90.7
Perdiz.2
Perdiz.5
88.6
93.3
Gary.47
Shumla
97.9
Gary.3
Gary.4
Gary.5
Gary.6
Gary.7
Gary.8
Gary.9
Gary.15
Gary.22
Gary.37
Gary.43
Wells
99.4
Gary.11
Gary.12
Gary.41
Gary.42
Gary.44
Kent.13
99.8
Gary.2
Gary.10
Gary.25
100
Bell
Scallorn
100
Williams.11
96
BirdsCreek.4
Ensor.6
97.6
BirdsCreek.8
97.3
DoolyBranch.3
Ellis.5
Scallorn.11
97.4
93.8
DoolyBranch.6
Ellis.4
Ellis.6
KirkCornerNotched
Marcos
Marcos.6
Marcos.7
Marcos.8
Motley
Williams
Williams.5
54.6
93.6
BigSandy
BigSandy.3
BirdsCreek
BirdsCreek.2
BirdsCreek.3
BirdsCreek.7
Ensor.4
Ensor.8
99.8
hcnarByloo
D
Ellis.2
Ensor.3
Ensor.7
Marcos.3
Marcos.5
Palmer
SanPatriceKeithville.4
100
SanPatriceKeithville.5
Scallorn.3
BigSandy.2
Edgewood
Ellis.11
100
Ellis.15
Ensor.5
Palmer.2
Palmer.4
Scallorn.2
Scallorn.7
97.5
Ellis
Ellis.18
Scallorn.4
Scallorn.5
Scallorn.9
97.9
Scallorn.8
97
Ellis.19
Ellis.20
Epps.3
Trinity
98.4
BirdsCreek.5
Ellis.3
Ellis.8
Ellis.9
Ellis.10
Ellis.14
Ellis.17
Ellis.23
Epps.2
Gary
Gary.17
Gary.29
Kent
Kent.6
Kent.11
100
Gary.13
Gary.14
Gary.18
Gary.23
Gary.27
Gary.30
Gary.31
Gary.32
Wells.2
100
Gary.35
Gary.36
Gary.39
Gary.40
Kent.3
Kent.5
100
Gary.19
Gary.24
Gary.26
Gary.28
Gary.33
Kent.2
100
Williams.3
100
Motley.2
Motley.3
Motley.4
Motley.5
Motley.6
Motley.7
Scallorn.6
Scallorn.12
Scallorn.14
Scallorn.15
100 100 98.3
Trinity.2
Williams.2
Williams.4
Williams.7
Williams.8
Williams.9
Williams.10
98
DoolyBranch.4
DoolyBranch.5
Ellis.7
Ellis.12
Ellis.13
Ellis.21
Ellis.22
Scallorn.10
Scallorn.17
95
BirdsCreek.6
DoolyBranch.2
Edwards
Ensor
Fairland.2
Fairland.3
Palmer.3
SanPatriceKeithville.2
100
SanPatriceDixon
SanPatriceDixon.2
100
SanPatriceHopeSJ.6
SanPatriceHopeSJ.7
SanPatriceHopeSJ.8
SanPatriceKeithville
96.2
Edwards.2
Edwards.3
Marshall.3
94.2
Edwards.4
Edwards.6
77.2
Edwards.5
97.8
73
Marshall.6
93.9
Fairland
Marshall.2
95.2
Fairland.4
Kirk
Marcos.4
Marshall.5
100
Ellis.16
Ensor.2
Marcos.2
Marshall
SanPatriceKeithville.3
SanPatriceKeithville.6
Scallorn.13
Williams.6
92.3
Epps
Marcos.9
98.2
Marshall.4
100
Gary.16
Gary.20
Gary.21
Gary.46
100
Gary.38
Kent.7
Kent.10
100
Gary.34
Gary.45
98.4
Kent.8
98.4
Bulverde
Bulverde.3
Bulverde.4
Darl.2
Darl.4
Darl.5
100
Darl.3
99.9
99.8
Bulverde.5
Delhi
Scottsblu.3
100
Delhi.3
Kent.9
Scottsblu
Scottsblu.2
Alba
Alba.2
Alba.3
Alba.4
100
Bonham.2
Scallorn.18
100
100
Bonham
Bonham.3
Cuney
Cuney.3
100
Cuney.2
Friley.4
100
Colbert.2
Friley.3
100
100
Scallorn.16
100
Friley.2
100
Colbert
100
Friley
100
100
100
0
97.4
98.8 94.2
90.1
99.9
93.9
97.5
Clovis.6
See figure 11.8 for an enlargement
of this section of the chart
See figure 11.9 for an enlargement
of this section of the chart
Figure 11.5
Majority-rule consensus tree, demonstrating all Scallorn clades (arrows with circles) are phylogenetically distinct
and distant from the Dalton clade (arrows with triangles). Note that this figure is included at this scale simply
to illustrate evolutionary distance between the point types.
11554_011b.indd 215 11/17/2017 6:14:26 PM
216 Ashley M. Smallwood and colleagues
Mean shape
–PC1
–PC2
–PC3
–PC4
+PC1
+PC2
+PC3
+PC4
a
Figure 11.6
Multivariate analysis of shape characteristics expressed at the positive and negative ends of the principal
component axes for (a) complete serrated projectile points and (b) isolated serrated margin.
11554_011b.indd 216 11/17/2017 6:14:26 PM
The Convergent Evolution of Serrated Points 217
Mean shape
–PC1
–PC2
–PC3
–PC4
–PC5
–PC6
+PC1
+PC2
+PC3
+PC4
+PC5
+PC6
b
Figure 11.6 (continued)
11554_011b.indd 217 11/17/2017 6:14:26 PM
218 Ashley M. Smallwood and colleagues
clade, demonstrating that the use of serrations was not directly tied to the evolution of
projectile-point shapes. In the study region, the application of serrations to blade margins
was adopted within multiple phylogenetically distant clades. As expected, the Dalton clade
is closely linked to Clovis and other Paleoindian lanceolates (see figure 11.8; O’Brien and
Lyman 2003). Surprisingly, this clade does not give rise to many later points in the study
region. Most Scallorn points are associated with Ellis points, which are found predomi-
nately in Texas but also grouped into the broader Lowe cluster, which has widespread
occurrence (Anderson and Smith 2003).
Ellis was originally classified as a dart point, but recent reanalysis suggests it and forms
like it could be evidence of experimentation in point designs for a new delivery technology.
Lyman, VanPool, and O’Brien (2008) found that the appearance of the bow was associ-
ated with an increase in dart-point variation, as flintknappers adjusted existing dart-point
forms through trial and error to find designs effective for tipping arrows. Similarly, Blitz
and Porth (2013) proposed that this morphological change was the product of arrow-point
refinement designed to function with bow technology. They show that Lowe-cluster points
from the Southeast represent transitional dart-arrow points. The association of Ellis points
with Scallorn, the earliest notched arrow point in Texas, lends further support to this
proposal (figure 11.9). Other Scallorn points in the cladogram are associated with Motley
dart points and other later arrow points. These other phylogenetic links may indicate that
multiple cultural-transmission processes (e.g., direct bias, guided variation) accompanied
the adoption of the bow and arrow in this region.
The geometric morphometric analyses provide insight into the context in which the
two distantly related populations converged on serrations. In the comparison of complete
–2 .0
–1 .5
–1 .0
–0 .5
0.0
0.5
1.0
1.5
2.0
PC2
PC3
PC4
Kimberley
Scallorn
–2 .0 –1 .5 –1 .0 –0 .5 0.00.51.01.52.
0
Canonical 2
Canonical 1
PC1Grand
Dalton
Figure 11.7
Canonical centroid plots, showing that Scallorn and Kimberley samples overlap in shape space.
11554_011b.indd 218 11/17/2017 6:14:27 PM
The Convergent Evolution of Serrated Points 219
Angostura.8
Angostura.10
Angostura.11
99.8
Clovis
Clovis.2
Clovis.4
Clovis.5
Clovis.8
Golondrina.3
Clovis.3
Clovis.7
Angostura.9
Clovis.10
Clovis.11
Clovis.13
Clovis.14
Clovis.15
Golondrina
Golondrina.2
Gower
SanPatriceHopeSJ.4
99.2
96.6
Clovis.9
Clovis.16
Clovis.17
100
Dalton
SanPatriceHopeSJ
99
Dalton.6
Dalton.10
Dalton.12
SanPatriceHopeSJ.5
99.4
Dalton.7
Dalton.8
Dalton.9
Dalton.11
Golondrina.4
98.3
Clovis.12
Dalton.2
Dalton.3
Pelican
SanPatriceHopeSJ.2
SanPatriceHopeSJ.3
98.6
Dalton.4
Dalton.5
Darl
94
Angostura.7
Bulverde.2
Bulverde.6
Bulverde.7
Delhi.2
Evans.4
Lange.3
Lange.4
98.8
Kent.4
Scottsblu.4
98.5
Kent.12
Lange
Lange.2
99.4
Lange.5
Yarbrough.6
Yarbrough.8
97.2
Evans
Godley
Godley.2
Godley.6
Yarbrough.4
99.7
Godley.3
Godley.4
Yarbrough.2
99.8
Godley.5
Godley.7
Godley.8
Godley.9
Godley.11
Palmillas
Palmillas.2
Yarbrough.
3
97
Evans.2
Evans.3
Lange.6
Sinner
Sinner.2
Sinner.4
Sinner.9
Yarbroug
h
98.1
Sinner.3
Sinner.5
Sinner.6
99
94
90.2
90.1
99.9
93.9
97.5
Figure 11.8
Majority-rule consensus tree, highlighting the Dalton clade closely linked to Clovis and other Paleoindian
lanceolates. Dalton points are marked with arrows with triangles.
11554_011b.indd 219 11/17/2017 6:14:27 PM
220 Ashley M. Smallwood and colleagues
serrated-point blade shape, Dalton, Scallorn, and Kimberley points are significantly differ-
ent; thus, the blades that Dalton and Scallorn flintknappers serrated are different in shape.
In fact, Scallorn point blades share more shape similarities with the Kimberley points
from Australia than with Dalton points. We suspect these differences are reflections of
differences in design for point function. However, in the analysis of the isolated serrated
margin, the shape of Dalton, Scallorn, and Kimberley serrations are not significantly dif-
ferent. The lack of significant difference in the shape suggests the serrations themselves
were designed for the same function. Thus, Dalton and Scallorn point makers converged
on the use of morphologically indistinguishable serrations.
BigSandy
BigSandy.3
BirdsCreek
BirdsCreek.2
BirdsCreek.3
BirdsCreek.7
Ensor.4
Ensor.8
99.8
h
c
na
r
Byloo
D
Ellis.2
Ensor.3
Ensor.7
Marcos.3
Marcos.5
Palmer
SanPatriceKeithville.4
100100
SanPatriceKeithville.5
Scallorn.3
BigSandy.2
Edgewood
Ellis.11
100
100
Ellis.15
Ensor.5
Palmer.2
Palmer.4
Scallorn.2
Scallorn.7
97.5
Ellis
Ellis.18
Scallorn.4
Scallorn.5
Scallorn.9
97.9
97.9
Scallorn.8
97
Ellis.19
Ellis.20
Epps.3
Trinity
98.4
BirdsCreek.5
Ellis.3
Ellis.8
Ellis.9
Ellis.10
Ellis.14
Ellis.17
Ellis.23
Epps.2
Gary
Gary.17
Gary.29
Kent
Kent.6
Kent.11
100
Gary.13
Gary.14
Gary.18
Gary.23
Gary.27
Gary.30
Gary.31
Gary.32
Wells.2
100
Gary.35
Gary.36
Gary.39
Gary.40
Kent.3
Kent.5
100
Gary.19
Gary.24
Gary.26
100
98
Figure 11.9
Majority-rule consensus tree, highlighting the association of Scallorn with Ellis points. Scallorn points are
marked with arrows with circles.
11554_011b.indd 220 11/17/2017 6:14:27 PM
The Convergent Evolution of Serrated Points 221
Why Serrate?
Why did these distantly related (if at all) populations, with distinct blade-shape designs,
converge on the application of serrations along their point margins? Experimental studies
help answer this question. Wilkins, Schoville, and Brown (2014) compared wound tracks
between untipped and stone-tipped spears and found no significant difference in pen-
etration depth. Similarly, Loendorf et al. (2015) found no difference in penetration for
serrated versus nonserrated stone tips. These results suggest that penetration depth was
likely not a key characteristic influencing the use of serrated projectiles. An alternative
functional effect might relate to wound size and associated internal shredding. Human
hunters are usually slower than their prey and pursue prey for long periods of time (chapter
6, this volume); poison or increased bleeding decreases capture time. A spear or arrow
point rarely kills large animals right away, but following a blood trail is an effective
technique to track animals as they weaken (Kelly 1995). Unfortunately, no experimental
studies of wound size have been conducted comparing serrated to nonserrated projectiles.
However, Wilkins et al. (2014) found that, compared to untipped spears, stone-tipped
spears produced significantly larger wounds, with a widening of the inner wound track.
Forensic studies comparing serrated to nonserrated-knife stab wounds show serrated
knives produce damage striations to skin (Pounder, Bhatt, Cormack, and Hunt 2011),
cartilage (Pounder, Cormack, Broadbent, and Millar 2011), and soft tissues, including
arteries (Jacques, Kogon, and Shkrum 2014) and that these striations are not produced by
nonserrated knives.
These experimental results show that the shape of the weapon tip has a significant
impact on wound size and shape and that serrations cause additional damage to multiple
internal tissues. Serrations increase damage caused by tearing, which could be produced
either through projectile or knife use. Explaining the adoption of serrations by Dalton and
Scallorn populations must take into account this functional advantage.
Serrations for Tearing in Dalton
Not all Dalton points are serrated, and serrated forms vary, which has led researchers to
propose a variety of functional hypotheses. Morse (1971) proposed that variation in body
shapes result from resharpening and changes in point use over time (see also Goodyear
1974). According to Morse, early in their use-lives Dalton points were socketed into
foreshafts, allowing them to be used both as dart points and as hafted knives. He suggests
that the jagged edges of serrated Dalton points, with fine serrations at the tip and longer
serrations at the shoulder, could help perform several functions related to the butchering
of deer.
O’Brien and Wood (1998) propose an alternative hypothesis, suggesting that the varia-
tion in Dalton shape was the result of primary engineering design. Thus, Dalton point
makers initially manufactured points with distinct blade characteristics (e.g., beveling,
11554_011b.indd 221 11/17/2017 6:14:27 PM
222 Ashley M. Smallwood and colleagues
serrations). They reference variation in modern archery equipment to argue all Dalton point
shape variants were designed to be projectiles (O’Brien and Wood 1998).
Both scenarios rely on a common fundamental function—the teethlike serrations on
Dalton points are ultimately used for tearing. In Morse’s (1971) explanation, the fine
serrations on a Dalton point tip could tear through deer skin for an initial cut, assist in
gutting, and, when used in a sawing-like motion, tear apart the front quarters from the
hindquarters. Similarly, in O’Brien and Wood’s (1998) hypothesis, serrated Dalton points
could have been particularly lethal because the teeth-like serrations tear open wounds,
causing internal damage, perhaps even encouraging a prominent blood trail for tracking.
Both scenarios also help explain adaptive responses to environmental changes at the end
of the Pleistocene. After the extinction of the megafauna and the competitive release of
deer populations, serrated point edges offered Dalton populations the selective functional
advantage of increased wound tearing to track blood trails of smaller, quicker prey and to
process that prey during butchering.
Serrations for Tearing in Scallorn?
Scallorn points are associated with a major shift in projectile technology—the adoption of
the bow and arrow. The adoption and social impacts of the spread of bow-and-arrow tech-
nology in North America varied from region to region with unique environmental, social,
and historical conditions (e.g., Blitz and Porth 2013; Shott 1996; VanPool and O’Brien
2013). In our study area, adoption of the technology has been associated with bow-based
warfare and enforced cooperation. Nassaney and Pyle (1999) conclude that bow-and-arrow
technology was abruptly adopted in central Arkansas during the Late Woodland period,
approximately 1,400 B.P., and that social factors such as warfare played a role. Populations
to the south and west, in Texas and Oklahoma, were refining their use of bow technology
and, with the advantages of this new technology, expanding their territories.
Nassaney and Pyle (1999) argue that warfare was an important incentive for the adop-
tion of the technology in Arkansas, and Blitz and Porth (2013) conclude that social change
took a different trajectory in the lower Southeast, an area including East Texas and Louisi-
ana. There, wild-plant foods were abundant, making food production less of a subsistence
focus. Populations aggregated in large civic-ceremonial centers, and the bow may have
helped enforce communal cooperation among large, nonkin-related populations. In the
western portion of the area discussed here, the adoption of the bow and arrow seems to
have coincided with the onset of widespread drought conditions, and bow-based warfare
was perhaps driven by population pressure and territorial disputes (Prewitt 1981). Thus,
throughout the ecologically diverse study area, Scallorn points emerged in the midst of
complex technological and social changes.
Although no studies have directly addressed the question of why Scallorn points are
serrated, a closer examination of archeological contexts shows a possible selective advan-
tage: a weapon for killing not just animals but humans as well. Across large parts of the
11554_011b.indd 222 11/17/2017 6:14:27 PM
The Convergent Evolution of Serrated Points 223
study area, including in Austin-interval burial contexts, Scallorn points are associated
with evidence of widespread violence, with numerous incidents of death caused by arrow
wounds (Boyd 1997, 2004; Collins 2004; Greer and Benfer 1975; Hall, Hester, and Black
1986; Hester 2004; Hester and Collins 1969; Hester, Wilson, and Headrick 1993; Huebner
and Comuzzie 1992; Prewitt 1974; Prikryl 1990; Ricklis 2004b). In this context, serrat-
ing Scallorn points potentially provided an important functional advantage in a social, as
opposed to an ecological, adaptation. Function remained the same, as the increased internal
tearing damage caused by serrations would have intensified impact shock and made it
more likely that points would remain embedded to cause additional damage, regardless
of the intended target.
Converging on Serrations
The tearing function of serrated point margins offered a selective advantage that explains
their adoption in the Late Paleoindian and Woodland–Mississippian/Late Prehistoric
periods. For Dalton populations at the end of the Ice Age, serrations would have improved
bloodletting during a deer hunt and could also have served a useful role as a hafted knife
for butchery. For Scallorn populations engaged in warfare, serrations would have improved
the shock, awe, and lethalness of arrows.
Conclusions
The study discussed here is an exercise in understanding how to detect cases of conver-
gence in the stone-tool record. Rather than using the entire scale of reduction technologies
or artifact types, which might lend themselves to more intuitively obvious evolutionary
and adaptive explanations, we have focused on convergence at the scale of a single tool
attribute. We used cladistics to support the interpretation that Dalton and Scallorn serrated
points are evolutionarily unrelated and support the hypothesis that both populations con-
verged on the use of serrations. We used geometric morphometric analysis to demonstrate
differences in overall point-blade designs and show that, despite these differences, both
populations converged on the same shape of serrations.
Serrations are often noted in point-type descriptions, but the functional role serrations
played through time is rarely considered in detail. We show that two unrelated populations,
Dalton and Scallorn, experiencing distinct environmental and social conditions, converged
on the application of serrations to point margins to address a shared adaptive functional
need for increased tearing.
References
Adams, D. C., Rohlf, F. J., & Slice, D. E. (2004). Geometric Morphometrics: Ten Years of Progress Following
the “Revolution.” Italian Journal of Zoology, 71, 5–16.
11554_011b.indd 223 11/17/2017 6:14:27 PM
224 Ashley M. Smallwood and colleagues
Adams, D. C., Rohlf, F. J., & Slice, D. E. (2013). A Field Comes of Age: Geometric Morphometrics in the 21st
Century. Hystrix, 24, 7–14.
Akerman, K., Fullagar, R., & van Gijn, A. (2002). Weapons and Wunan: Production, Function and Exchange of
Kimberley Points. Australian Aboriginal Studies Issue, 1, 13–42.
Anderson, D. G., & Bissett, T. G. (2015). The Initial Colonization of North America: Sea Level Change, Shoreline
Movement, and Great Migrations. In M. Frachetti & R. Spengler (Eds.), Mobility and Ancient Society in Asia
and the Americas: Proceedings of the Second International Conference on “Great Migrations” (pp. 59–88).
New York: Springer.
Anderson, D. G., & Mainfort, R. C., Jr. (2002). An Introduction to Woodland Archaeology in the Southeast. In
D. G. Anderson & R. C. Mainfort, Jr., (Eds.), The Woodland Southeast (pp. 1–19). Tuscaloosa: University of
Alabama Press.
Anderson, D. G., & Sassaman, K. E. (2012). Recent Developments in Southeastern Archaeology: From Coloniza-
tion to Complexity. Washington, D.C.: Society for American Archaeology Press.
Anderson, D. G., & Smith, S. D. (2003). Archaeology, History, and Predictive Modeling: Research on Fort Polk
1972–2002. Tuscaloosa: University of Alabama Press.
Anderson, D. G., Smallwood, A. M., & Miller, D. S. (2015). Early Human Settlement in the Southeastern United
States: Current Evidence and Future Directions. PaleoAmerica, 1, 1–45.
Ballenger, J. A. M. (2001). Dalton Settlement in the Arkoma Basin of Eastern Oklahoma. Robert E. Bell Mono-
graphs in Anthropology, no. 2, Sam Noble Museum of Natural History. Norman: University of Oklahoma.
Blitz, J. H., & Porth, E. S. (2013). Social Complexity and the Bow in the Eastern Woodlands. Evolutionary
Anthropology, 22, 89–95.
Bookstein, F. L. (1991). Morphometric Tools for Landmark Data: Geometry and Biology. New York: Cambridge
University Press.
Boyd, D. K. (1997). Caprock Canyonlands Archeology: A Synthesis of the Late Prehistory and History of Lake
Alan Henry and the Texas Panhandle Plains , vol. II. Reports of Investigations, no. 110. Austin, Texas: Prewitt
and Associates.
Boyd, D. K. (2004). The Palo Duro Complex. In T. K. Perttula (Ed.), The Prehistory of Texas (pp. 296–330).
College Station: Texas A&M University Press.
Bradley, B. A. (1997). Sloan Site Biface and Projectile Point Technology. In D. F. Morse (Ed.), Sloan: A Paleo-
indian Dalton Cemetery in Arkansas (pp. 53–57). Washington, D.C.: Smithsonian Institution Press.
Collard, M., Shennan, S. J., & Tehrani, J. J. (2006). Branching, Blending, and the Evolution of Cultural Similari-
ties and Differences among Human Populations. Evolution and Human Behavior, 27, 169–184.
Collins, M. B. (2004). Archeology in Central Texas. In T. K. Perttula (Ed.), The Prehistory of Texas (pp.
101–126). College Station: Texas A&M University Press.
Eren, M. I. (2012). On Younger Dryas Climate Change as a Causal Determinate of Prehistoric Hunter-Gatherer
Culture Change. In M. I. Eren (Ed.), Hunter-Gatherer Behavior: Human Response during the Younger Dryas
(pp. 11–23). Walnut Creek, Calif.: Left Coast Press.
Fenneman, N. M., & Johnson, D. W. (1946). Physiographic Divisions of the Conterminous U.S. Washington,
D.C.: U.S. Geological Survey.
Fox, G. (1939). Field Data and Reports on the Harrell Site (NT-5). Austin: Texas Archeological Research
Laboratory.
Gillam, J. C. (1996). A View of Paleoindian Settlement from Crowley’s Ridge. Plains Anthropologist, 157,
273–286.
11554_011b.indd 224 11/17/2017 6:14:27 PM
The Convergent Evolution of Serrated Points 225
Goodyear, A. C. (1974). The Brand Site: A Techno-functional Study of a Dalton Site in Northeast Arkansas.
Research Series no. 7. Fayetteville: Arkansas Archeological Survey.
Goodyear, A. C. (1982). The Chronological Position of the Dalton Horizon in Southeastern United States.
American Antiquity, 47, 382–395.
Grayson, D. K., & Meltzer, D. J. (2015). Revisiting Paleoindian Exploitation of Extinct North American
Mammals. Journal of Archaeological Science, 56, 177–193.
Greer, J., & Benfer, R. A. (1975). Austin Phase Burials at the Pat Parker Site, Travis County, Texas. Bulletin of
the Texas Archeological Society, 46, 189–216.
Hall, G. D., Hester, T. R., & Black, S. L. (1986). The Prehistoric Sites at Choke Canyon Reservoir, Southern
Texas: Results of Phase II Archaeological Investigations. Choke Canyon Series, no. 10. Center for Archaeological
Research, University of Texas at San Antonio.
Haynes, G. (2009). American Megafaunal Extinctions at the End of the Pleistocene. New York: Springer.
Hester, T. R. (2004). The Prehistory of South Texas. In T. K. Perttula (Ed.), The Prehistory of Texas (pp. 127–151).
College Station: Texas A&M University Press.
Hester, T. R., & Collins, M. B. (1969). Burials from the Frisch Auf! Site: 41FY42. Texas Journal of Science,
20, 261–272.
Hester, T. R., & Hill, T. C., Jr. (1975). Some Aspects of Late Prehistoric and Protohistoric Archaeology in
Southern Texas. Special Report, no. 1. Center for Archaeological Research, University of Texas at San Antonio.
Hester, T. R., Wilson, D., & Headrick, P. (1993). An Austin Phase Burial from Frio County, Southern Texas:
Archaeology and Physical Anthropology. Terra (Helsinki, Finland), 20, 5–8.
Hill, T. C., Jr., & Hester, T. R. (1973). A Preliminary Report on the Tortuga Flat Site: A Protohistoric Campsite
in Southern Texas. Newsletter of the Dallas Archeological Society, 17, 10–14.
Hixson, C. A. (2003). Horizontal Stratigraphy at the Graham/Applegate Site (41LL419). Newsletter of the
Council of Texas Archeologists, 27, 12–16.
Huebner, J. A., & Comuzzie, A. G. (1992). The Archeology and Bioarcheology of Blue Bayou: A Late Archaic
and Late Prehistoric Mortuary Locality in Victoria County, Texas. Studies in Archeology, no. 9. Austin: Texas
Archeological Research Laboratory, University of Texas.
Hughes, J. T. (1942). An Archeological Report on the Harrell Site of North-central Texas. M.A. thesis, University
of Texas. Austin.
Inman, B., Hill, T. C., Jr., & Hester, T. R. (1998). Archeological Investigations at the Tortuga Flat Site (41ZV155),
Zavala County, South Texas. Bulletin of the Texas Archeological Society, 69, 11–33.
Jacques, R., Kogon, S., & Shkrum, M. (2014). An Experimental Model of Tool Mark Striations by a Serrated
Blade in Human Soft Tissues. American Journal of Forensic Medicine and Pathology, 35, 59–61.
Kelley, C. J. (1947). The Lehmann Rock Shelter: A Stratified Site of the Toyah, Uvalde, and Round Rock Foci.
Bulletin of the Texas Archeological Society, 18, 115–128.
Kelly, R. L. (1995). The Foraging Spectrum. Washington, D.C.: Smithsonian Institution Press.
Koldehoff, B., & Walthall, J. A. (2009). Dalton and the Early Holocene Midcontinent: Setting the Stage. In T.
E. Emerson, D. L. McElrath, & A. C. Fortier (Eds.), Archaic Societies: Diversity and Complexity across the
Midcontinent (pp. 137–151). Albany: State University of New York.
Loendorf, C., Oliver, T. J., Tiedens, S., Plumlee, R. S., Woodson, M. K., & Simon, L. (2015). Flaked-Stone
Projectile Point Serration: A Controlled Experimental Study of Blade Margin Design. Journal of Archaeological
Science: Reports, 3, 437–443.
11554_011b.indd 225 11/17/2017 6:14:27 PM
226 Ashley M. Smallwood and colleagues
Lohse, J. C., Black, S. L., & Cholak, L. M. (2014). Toward an Improved Archaic Radiocarbon Chronology for
Central Texas. Bulletin of the Texas Archeological Society, 52, 65–89.
Lommel, A. (1997). The Unambal: A Tribe in Northwest Australia. Sydney: Takarakka Publications.
Lycett, S. J., von Cramon-Taubadel, N., & Foley, R. A. (2006). A Crossbeam Co-ordinate Caliper for the
Morphometric Analysis of Lithic Nuclei: A Description, Test and Empirical Examples of Application. Journal
of Archaeological Science, 33, 847–861.
Lyman, R. L., VanPool, T. L., & O’Brien, M. J. (2008). Variation in North American Dart Points and Arrow
Points When One or Both Are Present. Journal of Archaeological Science, 35, 2805–2812.
McGahey, S. O. (2000). Mississippi Projectile Point Guide. Jackson: Mississippi Department of Archives and
History.
Meltzer, D. J., & Holliday, V. T. (2010). Would North American Paleoindians Have Noticed Younger Dryas Age
Climate Changes? Journal of World Prehistory, 23, 1–41.
Miller, D. S., & Gingerich, J. A. M. (2013). Regional Variation in the Terminal Pleistocene and Early Holocene
Radiocarbon Record of Eastern North America. Quaternary Research, 79, 175–188.
Mitteroecker, P., Gunz, P., Windhager, S., & Schaefer, K. (2013). A Brief Review of Shape, Form, and Allometry
in Geometric Morphometrics, with Applications to Human Facial Morphology. Hystrix, 24, 59–66.
Morse, D. F. (1971). Recent Indications of Dalton Settlement Pattern in Northeast Arkansas. Southeastern
Archaeological Conference Bulletin, 13, 5–10.
Morse, D. F. (Ed.). (1997). Sloan: A Paleoindian Dalton Cemetery in Arkansas. Washington, D.C.: Smithsonian
Institution.
Nassaney, M. S., & Pyle, K. (1999). The Adoption of the Bow and Arrow in Eastern North America: A View
from Central Arkansas. American Antiquity, 64, 243–263.
O’Brien, M. J., & Lyman, R. L. (2003). Cladistics and Archaeology. Salt Lake City: University of Utah Press.
O’Brien, M. J., & Wood, W. R. (1998). The Prehistory of Missouri. Columbia: University of Missouri Press.
O’Brien, M. J., Boulanger, M. T., Buchanan, B., Collard, M., Lyman, R. L., & Darwent, J. (2014). Innovation
and Cultural Transmission in the American Paleolithic: Phylogenetic Analysis of Eastern Paleoindian Projectile-
Point Classes. Journal of Anthropological Archaeology, 34, 100–119.
Pounder, D. J., Bhatt, S., Cormack, L., & Hunt, B. (2011). Tool Mark Striations in Pig Skin Produced by Stabs
from a Serrated Blade. American Journal of Forensic Medicine and Pathology, 32, 93–95.
Pounder, D. J., Cormack, L., Broadbent, E., & Millar, J. (2011). Class Characteristics of Serrated Knife Stabs
to Cartilage. American Journal of Forensic Medicine and Pathology, 32, 157–160.
Prewitt, E. (1974). Archeological Investigations at the Loeve-Fox Site, Williamson County, Texas. Texas Archeo-
logical Survey Research Report, no. 49. Austin: University of Texas.
Prewitt, E. (1981). Cultural Chronology in Central Texas. Bulletin of the Texas Archeological Society, 85,
259–287.
Prikryl, D. J. (1990). Lower Elm Fork Prehistory: A Redefinition of Cultural Concepts and Chronologies along
the Trinity River, North-central Texas. Report, no. 37. Austin: Office of the State Archeologist, Texas Historical
Commission.
Ricklis, R. A. (2004a). Prehistoric Occupation of the Central and Lower Texas Coast. In T. K. Perttula (Ed.),
The Prehistory of Texas (pp. 155–180). College Station: Texas A&M University Press.
Ricklis, R. A. (2004b). The Archeology of the Native American Occupation of Southeast Texas. In T. K. Perttula
(Ed.), The Prehistory of Texas (pp. 181–202). College Station: Texas A&M University Press.
11554_011b.indd 226 11/17/2017 6:14:27 PM
The Convergent Evolution of Serrated Points 227
Rohlf, F. J. (1999). Shape Statistics: Procrustes Superimpositions and Tangent Spaces. Journal of Classification,
16, 197–223.
Rohlf, F. J. (2008a). TpsDig (version 2.12). Department of Ecology and Evolution, State University of New
York, Stony Brook.
Rohlf, F. J. (2008b). Relative Warps (version 1.45). Stony Brook: Department of Ecology and Evolution, State
University of New York.
Rohlf, F. J., & Slice, D. E. (1990). Extensions of the Procrustes Method for the Optimal Superimposition of
Landmarks. Systematic Zoology, 39, 40–59.
Sanderson, M. J., & Donoghue, M. J. (1989). Patterns of Variation in Levels of Homoplasy. Evolution; Inter-
national Journal of Organic Evolution, 43, 1781–1795.
Schiffer, M. B. (1975). An Alternative to Morse’s Dalton Settlement Pattern Hypothesis. Plains Anthropologist,
20, 253–266.
Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 Years of Image Analysis.
Nature Methods, 9, 671–675.
Sherwood, S. C., Driskell, B. N., Randall, A. R., & Meeks, S. C. (2004). Chronology and Stratigraphy at Dust
Cave, Alabama. American Antiquity, 69, 533–554.
Shott, M. J. (1996). Innovation and Selection in Prehistory: A Case Study from the American Bottom. In G.
H. Odell (Ed.), Stone Tools: Theoretical Insights into Human Prehistory (pp. 279–313). New York: Plenum.
Smith, H. L., & DeWitt, T. J. (2016). The Northern Fluted Point Complex: Technological and Morphological
Evidence of Adaptation and Risk in the Late Pleistocene Arctic. Archaeological and Anthropological Sciences.
doi: 10.1007/s12520-016-0335-y.
Smith, H. L., Smallwood, A. M., & DeWitt, T. (2014). A Geometric Morphometric Exploration of Clovis
Fluted-point Shape Variability. In A. M. Smallwood & T. A. Jennings (Eds.), Clovis: On the Edge of a New
Understanding (pp. 161–180). College Station: Texas A&M University Press.
Swofford, D. (1998). PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods) (version 4).
Sunderland, Mass.: Sinauer.
Thomas, D. H. (1978). Arrowheads and Atlatl Darts: How the Stone Got the Shaft. American Antiquity, 43,
461–472.
Turner, E. S., Hester, T. R., & McReynolds, R. L. (2011). Stone Artifacts of Texas Indians. Lanham, Md.: Taylor.
VanPool, T. L., & O’Brien, M. J. (2013). Sociopolitical Complexity and the Bow and Arrow in the American
Southwest. Evolutionary Anthropology, 22, 111–117.
Waters, M. R., & Stafford, T. W. (2007). Redefining the Age of Clovis: Implications for Peopling of the Americas.
Science, 23, 1122–1126.
Waters, M. R., Stafford, T., McDonald, G., Gustafson, C., Rasmussen, M., Cappellini, E., et al. (2011). Pre-Clovis
Mastodon Hunting 13,800 Years Ago at the Manis Site, Washington. Science, 334, 351–353.
Wilkins, J., Schoville, B. J., & Brown, K. S. (2014). An Experimental Investigation of the Functional Hypothesis
and Evolutionary Advantage of Stone-Tipped Spears. PLoS One, 9(8), e104514.
Zelditch, M. L., Swiderski, D., Sheets, H. D., & Fink, W. (2004). Geometric Morphometrics for Biologists: A
Primer. Amsterdam: Elsevier.
11554_011b.indd 227 11/17/2017 6:14:27 PM
11554_011b.indd 228 11/17/2017 6:14:27 PM
