Professional Documents
Culture Documents
PDF Archaeologist S Laboratory The Analysis of Archaeological Evidence 2Nd Edition Eb Banning Ebook Full Chapter
PDF Archaeologist S Laboratory The Analysis of Archaeological Evidence 2Nd Edition Eb Banning Ebook Full Chapter
https://textbookfull.com/product/food-analysis-laboratory-
manual-3rd-edition-s-suzanne-nielsen-auth/
https://textbookfull.com/product/data-analysis-for-business-
decision-making-a-laboratory-manual-2nd-edition-andres-fortino/
https://textbookfull.com/product/the-oxford-handbook-of-
archaeological-ceramic-analysis-oxford-handbooks-1st-edition-
alice-m-w-hunt-editor/
https://textbookfull.com/product/fingerprint-analysis-laboratory-
workbook-second-edition-daluz/
Qualitative Research Methods: Collecting Evidence,
Crafting Analysis, Communicating Impact 2nd Edition
Sarah J. Tracy
https://textbookfull.com/product/qualitative-research-methods-
collecting-evidence-crafting-analysis-communicating-impact-2nd-
edition-sarah-j-tracy/
https://textbookfull.com/product/multivariate-analysis-for-the-
behavioral-sciences-2nd-edition-brian-s-everitt/
https://textbookfull.com/product/my-people-as-your-people-a-
textual-and-archaeological-analysis-of-the-reign-of-jehoshaphat-
american-university-studies-chris-mckinny/
https://textbookfull.com/product/arts-crafts-and-trades-in-
ancient-and-byzantine-thessaloniki-archaeological-literary-and-
epigraphic-evidence-anastassios-ch-antonaras/
https://textbookfull.com/product/alternative-iron-ages-social-
theory-from-archaeological-analysis-1st-edition-brais-x-curras-
editor/
Interdisciplinary Contributions to Archaeology
Edward B. Banning
The
Archaeologist’s
Laboratory
The Analysis of Archaeological Evidence
Second Edition
Interdisciplinary Contributions to Archaeology
Series Editor
Jelmer Eerkens
University of California, Davis
Davis, CA, USA
More information about this series at http://www.springer.com/series/6090
Edward B. Banning
The Archaeologist’s
Laboratory
The Analysis of Archaeological Evidence
Second Edition
Edward B. Banning
Department of Anthropology
University of Toronto
Toronto, ON, Canada
ISSN 1568-2722
Interdisciplinary Contributions to Archaeology
ISBN 978-3-030-47990-9 ISBN 978-3-030-47992-3 (eBook)
https://doi.org/10.1007/978-3-030-47992-3
This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
For Lyn and Jordan
Preface
A new edition of The Archaeologist’s Laboratory has been long overdue. Not only have things
changed in our discipline over the last two decades, I have long wanted to give the book better
focus and to illustrate its concepts with recent case studies. Although it is impossible to separate
lab work completely from fieldwork, the emphasis here is on the former, the activities that
archaeologists and their colleagues carry out after at least some fieldwork is complete. How-
ever, we should always keep in mind that fieldwork and laboratory work go hand-in-hand, and
the best research designs are not compartmentalized in one or the other.
A thread woven throughout the book is the quality and validity of archaeological arguments
and the data we use to support them. No matter what our theoretical orientation, we want to
make academic arguments and heritage recommendations that are convincing. At a bare
minimum, we want to convince ourselves and others that the results we seem to have are not
just due to measurement errors or poor research design. We want to draw inferences that are at
least reasonable given the evidence available to us, and to present conclusions or hypotheses
that arguably fit the evidence better than some competing hypotheses, and especially the
competing hypothesis that apparent patterns are only due to sampling problems, investigator
bias, or measurement error.
Consequently, I have aimed in this book to present basic information about many of the most
common laboratory activities that archaeologists carry out, and a few less common ones, but in
the context of ensuring, to the extent that it is possible, that these activities result in
observations, measurements, and conclusions that are accurate, valid, and relevant to the
research problems at hand. While it has been necessary to subdivide laboratory research into
a number of arguably arbitrary categories, it is important to realize that archaeology is a holistic
discipline that works best when we consider different kinds of evidence, such as pottery and
plant and animal remains, in tandem. It would be impossible to understand past food and
culinary practices properly, for example, by examining only one of these. That is just one of the
reasons that archaeologists typically work in research teams that combine a variety of expertise.
I also firmly believe that the way we present our research results to others is just as important
as the way we conduct research, so I devote space to the presentation of data, mainly in graphs
and artifact illustrations, but with at least brief reference to 3D imaging and the media of
archaeological dissemination, which today go well beyond traditional monographs and aca-
demic journals.
I have tried to give the book an international and inclusive flavor by including examples and
case studies from around the world and from the Paleolithic to historical archaeology. Archae-
ology in different parts of the world has different challenges and different research problems
that sometimes call for very different research strategies, even though many aspects of research
are more universal.
This book has benefited from advice and suggestions I received from and conversations I
have had with a great many students and colleagues. I owe some of my thinking on topics
covered in this book to the early influences of Ron Farquhar, Ted Litherland, Ron Hancock, and
the late Jack Holladay. In addition to those whom I thanked in my first edition, I would like to
acknowledge valuable suggestions for improvements to one or more chapters by Caitlin Buck,
vii
viii Preface
Ben Collins, Costis Dallas, Tom Dye, Kevin Gibbs, the late Stuart Laidlaw, Lee Lyman,
Danielle MacDonald, Lisa Maher, Joy McCorriston, Katherine Patton, Trevor Ringrose,
Branden Rizzuto, Aaron Shugar, Jim Skibo, James Stemp, Susan Stock, Todd Surovell, John
Triggs, and John Whittaker. I also would like to thank Teresa Krauss, Christi Lue, and Sowmya
Thodur at Springer; Sophia Arts for her assistance with copy-editing and the index; and
Emerson Grossmith, Seiji Kadowaki, Steve Rhodes, Joy McCorriston, Emma Yasui, and
Emily Hubbard for photos of potting, core refitting, bone awls, charred rachis segments, starch
and phytoliths. Arno Glasser, Steve Rhodes, and Stuart Laidlaw helped with some of the other
photos, and William Wadsworth and Yara Salama drew many of the illustrations, while Julia
Pfaff and Catharine Solomon provided several lithic illustrations. The remaining photographs
and illustrations, unless noted otherwise, are by the author.
ix
x About the Book
References
Barcelo, J. A., & Bogdanovic, I., Eds. (2015). Mathematics and archaeology. Boca Raton:
CRC Press.
Baxter, M. J. (2003). Statistics in archaeology. New York: Wiley.
Buck, C. E., Cavanagh, W. G., Litton, C. D. (1996). Bayesian approach to interpreting
archaeological data. New York: Wiley.
Drennan, R. D. (2010). Statistics for archaeologists. A common sense approach. New York:
Springer.
About the Book xi
Fletcher, M., & Lock, G. R. (1991). Digging numbers: Elementary statistics for archaeologists.
Oxford: Oxford University Committee for Archaeology.
Rowntree, D. (2018). Statistics without tears: A primer for non-mathematicians. London:
Penguin Books.
Shennan, S. (1997). Quantifying archaeology (2nd ed.). Edinburgh: Edinburgh University
Press.
Contents
xiii
xiv Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
About the Author
xxi
List of Figures
Fig. 1.1 Manual caliper and close-up of the measuring scale. Note how the “zero”
tick on the mm scale falls just a little to the left of 18 mm. That suggests that
the measurement should be close to 17.7 or 17.8 mm. Now, looking along
the bottom scale (the 0.05 mm scale), the tick that best lines up with the tick
above it is the one between “7” and “8,” indicating that the measurement is
17.75 mm (0.05 is its resolution or precision). Given that this is a rather
cheap and possibly not highly accurate caliper, one might be more com-
fortable citing 0.1 mm resolution (or 17.8 mm) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Fig. 1.2 The three triangles in (a) indicate three darts aimed very precisely but not
very accurately, while the three in (b) are fairly accurate but less precise
because they are not as tightly clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Fig. 1.3 Scatterplot of the relationship of total floor area to population size in an
ethnographic sample of largest settlements (Modified from Naroll 1962).
The graph omits the largest site (Cuzco at 16.7 ha and 200,000 people) and
has linear, rather than logarithmic scales. The dashed diagonal indicates an
alleged relationship of 10 m2 of floor area per person but note how poorly
the data fit the regression line, which is heavily influenced by the three
largest sites, especially Cuzco. Small error bars have been added just to
illustrate how to show estimated measurement errors on area and population
statistics . . . . .. . . . .. . . . . .. . . . .. . . . .. . . . . .. . . . .. . . . .. . . . .. . . . . .. . . . .. . . . .. . . . .. . . . . .. . . 12
Fig. 2.1 The effect of symmetry and skewness on some common measures of central
tendency. In a normal distribution, the sample mean, median and mode are
equal and in the center of the curve. In skewed distributions, however, the
sample mean deviates considerably from the median and mode because
extreme values in the long tail pull it to left or right. In such cases, the mean
does not provide a good sense of central tendency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Fig. 2.2 A histogram showing how to calculate a mean from grouped data, as you
might find in a publication. Here we would multiply the number of
observations in each interval by the mid-point of that interval (indicated by
ticks on top of the bars), sum those values, and then divide the sum by the
number of observtions. In this example, the sum would be 5 + (2*9) +
(4*13) + (3*17) + (2*21) + (2*25) + 29, so the mean would be 247/15 ¼
16.5 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Fig. 3.1 A taxonomic classification of arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Fig. 3.2 A taxonomic classification of Owasco pottery from New York State (After
Whallon 1972: 17). Note that the distinctions made in the left side of the
“tree” are quite different from those on the right . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
xxiii
xxiv List of Figures
Fig. 4.5 A few examples of lexical items and interactions for describing cylinder
seals. Top: interactions among people and animals. Middle: holding an
object. Bottom: people, objects and animals and their interactions on
vehicles. Lexical items include animal subject (Sa), animal object (Oa),
subject man (Sm), object man (Om), hybrid creature being led (S2h), neutral
man (Nm), neutral container (Nc), and actions include holding two-handed
to left (2 h L), one-handed on one shoulder (1 h 1sh), two-handed on both
shoulders (2 h 2sh), and two-handed on head. (2 h hd; modified from Digard
et al. 1975: 43, 45, 249, 325) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Fig. 4.6 A subset of segmentation rules for Samarran house plans in Mesopotamia,
using the orientation rule that the widest part of the house is shown at the
bottom. (After Banning 1997: 24) . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . . . . .. . . . . 48
Fig. 4.7 Example of a portion of an Entity-Relationship Model (ERM) for a project
with multiple sites and excavations. In this style of ERM, rectangles are
entities or entity types, ovals are some attributes of those entities, and
diamonds are relationships. On the connecting lines, the vertical bar at one
end and “crow’s foot” at the other means that multiple members of the entity
attached by the “crow’s foot” may participate in the relationship with only
one member the other entity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Fig. 4.8 A structure chart for a file to record “sites” information, with a key attribute
(site number) and unique “site name” field above the dashed line, and a
number of fields to describe other attributes of each site in a regional survey
below. Bolded letters indicate whether the fields are alphanumeric (A),
numeric (N), date (D), boolean (B), or text (T) fields. Note that “Site
number” is not a numeric field because the numbers are arbitrary labels . . . . . 50
Fig. 4.9 A structure chart depicting a relation between a “lithics” file and a “context”
file for an archaeological excavation. Only some of the fields are shown.
Bolded letters indicate whether the fields are alphanumeric (A), numeric
(N), date (D), boolean (B), or text (T) fields. The “crow’s foot notation” on
the connecting lines indicate that each lithic has a mandatory relationship to
a single member of the “context” file (double bar) while each context has an
optional relationship to potentially many members of the “lithics” file
(“crow’s foot”). Note that “site number” in the context file could be an
attribute pointer to a key attribute in a “sites” file as in Fig. 4.8 . . .. . . . .. . . . .. . 50
Fig. 4.10 An example of a portion of a Data Flow Diagram (DFD) to sketch out what
happens when someone carries out a process that sorts artifacts into groups
on the basis of their stratigraphic context. For each artifact in the “artifacts”
file, the process has to look up its stratigraphic placement by reference to the
“context” file and then assigns it to one of the stratigraphic groupings
at right . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Fig. 4.11 Example of two pages from a data dictionary, one defining a data element,
the other a data flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Fig. 5.1 Stem-and-leaf plot (left) and a corresponding tally (right) to show two
different ways to record the frequency distribution of blade lengths in
millimeters. The former includes all the raw data, including the last digit,
while the latter, like a histogram, only tells you how many observations fell
within each interval .. . . . . . . . . . .. . . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . 61
Fig. 5.2 Box-plot of edge straightness of bifaces from Olduvai Gorge Beds II and IV.
(After Shipton 2018: 117) . . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . . .. . 61
Fig. 5.3 Azimuth plot of orientations in 132 Irish “court cairn” tombs. (Y. Salama,
After De Valera 1959: plate 35) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
xxvi List of Figures
Fig. 5.4 Bar graphs to show the distribution of surface treatments among miniature
vessels and the general assemblage at the Homol’ovi I site in northeastern
Arizona. (After Fladd and Barker 2019: 114) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Fig. 5.5 Two back-to-back bar graphs to facilitate comparison of distributions of
body parts by NISP among deer remains in pits and other cultural layers at
the site of Gomolava, Serbia. (Data from Orton 2012: 334) . . . . . . . . . . . . . . . . . . . 63
Fig. 5.6 “Battleship curve” (a) of tombstones from New England. (After Deetz
1967) and (b) the same data shown as histograms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Fig. 5.7 Stacked bar graph of carbonized plant taxa by time period at Santa Cruz
Island (after Arnold & Martin 2014: 241). Note the distracting moiré effects
(patterns not used in the original), and that these data could have been shown
effectively in a time series instead . .. .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . 65
Fig. 5.8 Time-series bar graph indicating the times (as once understood) when each
of several Near Eastern sites was occupied. Gaps in the bars indicate site
abandonments (modified from Singh 1974). Note the time arrangement with
oldest at the bottom, as in stratigraphic order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Fig. 5.9 Pie chart of the distribution of four different sources of fat following analysis
of chemical residues in pottery. (Y. Salama) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Fig. 5.10 Map with pie charts to show the proportional variation in food groups across
space in the Upward Sun River site in central Alaska. (Y. Salama, after Choy
et al. 2016: 9760) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Fig. 5.11 Windflowers to show the distribution of horizontal orientations (compass
bearings) and dip angles (inclination) of artifact refits in layer 2 at La
Ferrassie, France. (Data from McPherron 2018: 12) .. . . . . . . . . . .. . . . . . . . . . .. . . . . 67
Fig. 5.12 Rose plots to compare the angular distributions of body orientations in Early
Bronze Age Grave Burials (A) and Storage Pit Burials (B) in Moravia. (Y.
Salama, after Pankowská & Monik 2017: 922) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Fig. 5.13 Two histograms with continuous data, shown back-to-back to facilitate
comparison of distributions of lengths of projectile points at two early
Neolithic sites in Lebanon and Israel. (After Rhodes et al. 2020) . . . . . . . . . . . . . 68
Fig. 5.14 Histogram with discrete data, in this case, a Poisson distribution with a mean
of 4 (see Chap. 8) . . . . . . . . . . . .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . .. . . . . . . . 68
Fig. 5.15 The histogram in (a) distorts the information by treating unequal intervals as
though they are equal (and also has time in reverse order), that in (b) corrects
for the period durations but, unless all the sites were continuously occupied
throughout their period, distorts histogram shape by using heights, and (c)
corrects this effect by making the areas in each interval proportional to
frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Fig. 5.16 Effect of “bin” size and location on histogram shape. All four histograms
show the same data, but with different “bin” boundaries . . . . . . . . . . . . . . . . . . . . . . 70
Fig. 5.17 Broken-line graph showing interpolated variation in artifact density along a
transect across an artifact scatter in Lefkas, Greece. (Data from Gallant
1986: 407) . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . 71
Fig. 5.18 Time-series graph showing how estimated momentary population sizes
changed over time (after Ortman 2016). Steep slopes on the broken lines
indicate rapid population growth or decline and bars on total population
represent standard errors (here conjectural) . . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . .. . 71
Fig. 5.19 Misleading time-series with unequal periods shown as though they were of
equal duration (a) and a corrected version (b) with the approximate duration
of each period indicated in years and points at the midpoints of periods (data
from Morris 1987: 141). The latter more accurately represents rates
of change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
List of Figures xxvii
Fig. 5.21 Comparing wealth inequality in graves by means of ogives. The Lorenz
curve compares inequality by plotting the cumulative proportion of metal
artifacts against the cumulative proportion of graves, ranked from “poorest”
to “richest” graves. The diagonal line (y ¼ x) would indicate an “equitable”
distribution. The curves closest to the lower right corner of the graph show
the greatest inequality, with most graves showing no metal artifacts at all
(modified from Morris 1987: 142). The maximum vertical difference (D)
between curves for early-mid and late Geometric periods provides the
Kolmogorov-Smirnov statistic for the difference between these periods’
wealth distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Fig. 5.20 Specialized time-series with bottom-to-top orientation to plot oxygen-
isotope ratios (δ18O ‰) in the GISP-2 ice core from Greenland as a climate-
change proxy. (Curve from Buizert et al. 2014: 1178) . . . . . . . . . . . . . . . . . . . . . . . . . 73
Fig. 5.22 Scatterplot showing a Factor Analysis (see Chap. 3) of chemical
compositions of proto-porcelain sherds from a cemetery and four kiln sites
in China. Adding dashed elipses to the graph helps to highlight important
clusters in the data. Note the outlier at upper right.
(After Yu et al. 2018: 28) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Fig. 5.23 Four-dimensional scatterplot (or “rugplot”) with sides of the “boxes” shown
side-by-side. The dashed line segments highlight the location of a selected
artifact in multiple dimensions, and the short line segments connect the
graphs by their shared points (cf Tufte 1983: 135) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Fig. 5.24 Scatterplot of the lengths and widths of Clovis points from the Hogeye &
Graffenried caches in Texas. (Redrawn with data from Waters et al. 2011:
116) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Fig. 5.25 Ternary graph of the percentages of domesticated cattle, caprines and pigs at
Neolithic sites in the western Balkans (after Orton et al. 2016: 12). Circles
represent open-air sites and squares cave sites, and the colors represent
different regions. Note that cave sites are high in caprines, and open-air sites
high in cattle, while pigs occur more rarely at both kinds of sites . . . . .. . . . .. . . 76
Fig. 5.26 Maps of basalt microrefuse densities on a house floor in a Neolithic site in
Jordan (Ullah et al. 2015), with (a) a chloropleth map showing number of
basalt fragments per cm2, (b) interpolation to smooth the data to make an
isopleth map that is not constrained by the grid, and (c) with the densities
converted to Z scores and areas from (b) that are at least 1 Z above the mean
in red and ones more than 1 Z below the mean in yellow (see pp. xx). The
large object at center is a mortar, distortion to the grid was caused by rocks
and other obstructions, and dark grey marks a house wall . . . . . . . . . . . . . . . . . . . . . 76
Fig. 5.27 Spatial histogram (or stepped statistical surface) of potsherds at the entrance
to Grave 9 at Fjälkinge, southern Sweden (Y. Salama, after Shanks and
Tilley 1987: 166). Note that this image exaggerates the density of sherds at
lower right. (Because that unit is much larger than the others, it should have
been scaled by sherd density, not count) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
xxviii List of Figures
Fig. 5.28 Example of a simple control chart (a), showing Upper (UCL) and Lower
Control Limits (LCL) and target value of 0 mean difference between rim
diameters by expert and lab personnel. The large arrow highlights 2 weeks
where personnel suddenly began to overestimate rim diameter, suggesting
perhaps some change in personnel or the expert. Example of a difference-
against-mean chart (b) for measurements of projectile point length by the
same analyst at two different times (after Lyman and VanPool 2009: 497).
As the scatter is horizontal, there is no magnitude-dependent trend, the mean
difference is close to zero, and most of the points are within two standard
deviations of this mean difference. However, the points are somewhat more
widely scattered about the mean for smaller measurements than
for large ones . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . . .. . . 78
Fig. 5.29 Distortion in a time series (a) caused by cutting out the bottom of the y-axis
makes the slope of population change appear remarkably steep around 1200
and 1400 CE when an honest graph (b) shows that the slope is less
pronounced and most of the differences may be within measurement
errors . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . . . . . 80
Fig. 5.30 Use of meaningless 3D effects (a) or icons sized to their heights (b) distorts
the data by jeopardizing the proportionality of area and also distracts
viewers from the data. (Projectile point images from
Boule 1923: 378, 380) . .. . .. . .. . .. .. . .. . .. . .. . .. . .. . .. .. . .. . .. . .. . .. . .. . .. .. . .. . .. . 81
Fig. 5.31 Bar graph of axe types found in single finds and hoards from Middle
Neolithic IIB votive deposits at Skane, with a linear scale (a) and with a
square-root transformation (b). The latter violates the principle of
proportionality but makes it easier to show both large and small values on
the same graph (data from Tilley 1996: 289). Note that a log10 transfor-
mation is not possible in this case because there is a 0 value . . . . . . . . . . . . . . . . . . 82
Fig. 6.1 The “Scientific Cycle” . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . . .. . . 87
Fig. 6.2 The Hermeneutic Circle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Fig. 6.3 Spurious correlation between changes in the bore diameter of smoking pipes
in the United States (Heighten and Deagan 1971) and the price of wheat in
Münich (Kellenbenz et al. 1977: 218) over the period from 1688 to 1775
(correlation coefficient, Pearson’s r ¼ 0.85). The fact that wheat prices
happened to rise at the same time that pipestem diameters decreased is
purely coincidental . . .. . .. . .. . .. . .. . . .. . .. . .. . .. . .. . .. . . .. . .. . .. . .. . .. . . .. . .. . .. . .. . 90
Fig. 6.4 Examples of some basic sampling frames and resulting samples: (a) a simple
random sample from a collection of projectile points, (b) a proportionally
stratified cluster sample of artifacts with contexts selected randomly from
three kinds of context, (c) a systematic spatial sample of squares in a grid,
(d) a systematic, stratified, unaligned sample of points for coring, (e) a
systematic stratified spatial sample of transects in an archaeological survey
(randomly chosen start transect in each stratum), and (f) a disproportionate
stratified random sample of spaces in a large architectural complex with
three types of spaces (33% CY courtyards, 40% D domestic, and 30% S
storage). Selected elements are grey, and sampling was without
replacement . . . .. . . .. . . .. . .. . . .. . . .. . .. . . .. . . .. . . .. . .. . . .. . . .. . .. . . .. . . .. . .. . . .. . . .. . 96
List of Figures xxix
Fig. 6.5 PPS sampling: (a) point counting, only the particles or sites intersected by
systematic points, (b) line counting, only particles or sites intersected by
systematic lines, and (c) only particles or sites whose lower (or southern)
ends are within a ribbon or strip are counted. For (b) and narrow particles
whose length is less than the line spacing, the probability of intersection is
twice their length divided by π times the spacing between lines. An alter-
native to (c) is to count those whose centers are within the ribbon, but
centers are not as easy to determine accurately as ends. Note that, for all
these methods, large targets are intersected more often than small ones . . . . 97
Fig. 6.6 Example of sequential sampling for flint, pottery and basalt microrefuse at
the Neolithic site of Tabaqat al-Bûma, Jordan (Y. Salama, after Ullah et al.
2015). After the sample size for these cluster samples exceeds n ¼ 14, there
is no significant improvement in the relative standard error . . . . . . . . . . . . . . . . 101
Fig. 7.1 Effect of fragmentation on the value of NISP, here for 10 complete
specimens. At first fragmentation leads to higher values of NISP, but then
many fragments become unidentifiable, so that NISP approaches 0 at high
levels of fragmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . 108
Fig. 7.2 Conjectural pollen diagram with various taxa of arboreal and herbaceous
plants. Note how major peaks in one taxon tend to correspond with troughs
in the spectra of other taxa. (See also Fig. 16.14) . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Fig. 7.3 Illustration of Orton’s (2000: 65) ratio-of-ratios approach to estimating the
difference in taxonomic ratios among sites with similar preservation
conditions. In site A, the original ratios among three taxa are 1:1 and 1:1,
while in site B the ratios are 5:3 and 3:2. After those assemblages undergo
deterioration at rates that vary by taxon but are proportional at the two sites
(site B is twice as destructive as site A), the ratios among taxa in the sample
assemblages of the two sites preserve the original proportionality (e.g., the
ratio of 1:1 to 5:3 is the same as 2:1 to 10:3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Fig. 7.4 Graph of ranked site sizes follows Zipf’s Law, a power function that is a
fractal property, here illustrated with data from the Laconia Survey, Greece,
and a subset from the Chrysapha area within it (after Cavanagh 2009: 414).
The fact that both the full data set and the subset have the same slope means
that they have the same fractal dimension, demonstrating self-similarity at
different scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Fig. 7.5 Calculation of the fractal measure, D, for the sizes of flintknapping debris
(Y. Salama, after Brown et al. 2005). The axes are natural logarithms, r is
length of flakes in millimeters and N(>r) is the number of artifacts with
lengths greater than r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . 122
Fig. 7.6 Sequential sampling to determine how large a sample (n) is necessary to
obtain a stable estimate of ubiquity. In both examples shown here, the
probability that any particular sample element contained at least one item
was 0.25 (dashed line), but one sequential sample (squares) approached but
never reached this, while the other (circles) yielded an accurate estimate at n
¼ 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Fig. 7.7 When plotting richness (number of taxa) against the cumulative proportions
of taxa, population A is more diverse than population B despite having fewer
taxa because the cumulative proportions of taxa are higher (cf. Ringrose
1993b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Fig. 7.8 Plot of m against the expected species index, s(m). Sample A shows more
diversity than sample B because its curve is higher (cf. Ringrose 1993b) . 126
xxx List of Figures
Fig. 8.1 Binomial model for a probability of 0.4 of finding at least one maize kernel
in ten 1 m2 excavation units, B(10,0.4). Note that the most likely outcome is
that four units will have one or more maize kernels, and there is very little
chance that none, or as many as eight, will have a maize kernel. The bars are
separate line segments to signal that this is a discrete distribution . . . . . . . . . 132
Fig. 8.2 Poisson distribution for a density of 5 artifacts per square
meter (λ ¼ 5) .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . 133
Fig. 8.3 Normal or Gaussian distribution for the lengths of a population of projectile
points with a mean length of 43 mm and standard deviation (σ) of 7 mm. The
hatched area represents slightly more than 68% of the area under the curve,
and each tail represents almost 16%. The area within 2σ is about 95%, and
within 3σ is about 99.7% of the curve’s total area . .. . .. . .. . .. . .. . .. . .. . .. . .. 133
Fig. 8.4 Examples of a two-tail Z-test with two rejection regions that jointly add up
to 0.05 of the curve’s area (a) and a one-tail test, also with a significance of
0.05, but with all the area in one tail (b). In classical statistics, a result within
one of the tails (hatched) would result in rejection of the null hypothesis that
the mean ¼ μ at 95% confidence .. . . . .. . . .. . . .. . . . .. . . .. . . .. . . . .. . . .. . . .. . . . .. 134
Fig. 8.5 Test of the hypothesis that the flake density could at a site could be λ ¼ 3 if a
sample drawn from the population has a mean > 5. This uses a one-tail test
and the Poisson model and shows that this outcome has a probability of only
0.084 (grey bars), so that we would reject the hypothesis at 90% confidence
but could not reject it at 95% confidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Fig. 9.1 Schematic of a storage box containing silica gel to control humidity for
sensitive artifacts, such as metal ones (W. Wadsworth, after Cronyn 1990:
75). Note that it would be necessary to change the silica gel periodically, on
the basis of the humidity indicator strip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Fig. 9.2 Example of a tag to indicate the removal of an artifact from its place in a
drawer (left) and part of a page from a “Specimens Removed Log” (right)
that documents the date and purpose of removal and the person
responsible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Fig. 10.1 Laboratory step-stool for accessing high shelving . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Fig. 10.2 Using microscopes at ergonomic heights with plenty of illumination from a
large window . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. 153
Fig. 11.1 Plot of the first two Principal Components of archaeological samples and
quarry pieces analyzed by ICP-MS, a type of mass spectrometry, with
ellipses for the 95% confidence intervals for group membership. (After
Carballo et al. 2007: 36) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Fig. 11.2 Conchoidal fracture, resulting in a Hertzian cone that expands from the point
of impact by a hard object on a block of glass or flint . . . . . . . . . . . . . . . . . . . . . . 162
Fig. 11.3 Removal of a flake from the side of a core by conchoidal fracture . . . . . . . . 162
Fig. 11.4 Making a pyramidal core from an egg-shaped cobble (a, b) and
rejuvenating the core (c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Fig. 11.5 A student attempting to learn hard-hammer reduction . . . . . . . . . . . . . . . . . . . . . . 163
Fig. 11.6 Anatomy of a flake or blade, with some of its main measurements. In the
ventral view at right, the heavy lines represent the “box length” and width at
half box length, while box width is shown above. Box length and width
define a rectangle that encloses the flake when aligned to the platform. Axes
with arrows represent the axial length and axial width at half-length, axes
with long dashes represent long axis and width at half long axis, and lightly
dashed lines represent maximum length and width at half maximum (which
in this case is also equal to maximum width) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
List of Figures xxxi
Fig. 11.7 A classification of flake terminations as viewed in radial section: from left to
right, feather, step, hinge, and plunging terminations . . . . . . . . . . . . . . . . . . . . . . . 166
Fig. 11.8 Classification of retouch type and ordinal scale of invasiveness: (a) scaled,
(b) stepped, (c) sub-parallel, and (d) parallel retouch; (e) short, (f) long, (g)
invasive, and (h) covering retouch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Fig. 11.9 A classification for description of retouch position: (a) direct, (b) bifacial,
(c) alternate, and (d) inverse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Fig. 11.10 Measuring the “length” of retouch on an edge by the angle of arc in degrees
on a polar graph, here 127 . A similar scale can be used to record the
locations of polish or use wear. The artifact is positioned here by its axis of
flaking (axial length) and midline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Fig. 11.11 Taxonomy of debitage. (Modified from Sullivan and Rozen 1985) . . . . . . . . 169
Fig. 11.12 Flakes refitted to a core from the Late Neolithic site of al-Basatîn, Jordan
(From Kadowaki and Banning 2018: 68). “PC” is partially cortical . . . . . . . 170
Fig. 11.13 Map of connections between refitted lithics in Layer 15 at Oliveira Cave,
Torres Novas, Portugal. (Modified from Deschamps and Zilhão 2018) . . . 170
Fig. 11.14 Segmentation terminology for a lower milling stone or quern. (After Wright
1992: 497–498) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Fig. 11.15 Segmentation terminology for an upper milling stone or handstone. (After
Wright 1992: 497–498) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Fig. 11.16 Segmentation terminology for mortars and pestles. (After Wright 1992:
497–498) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Fig. 11.17 Tenth- or Eleventh-century whetstones from Starigard-Oldenburg,
Germany. (Y. Salama, after Gabriel and Kempke 2011) . . . . . . . . . . . . . . . . . . . . 174
Fig. 11.18 Scatterplot and difference-against-mean plots showing inter-analyst
variability for two analysts measuring the thickness of projectile points.
(After Lyman and VanPool 2009: 498) . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. 177
Fig. 11.19 Medians of lithic quality measures for different modes of learning how to
strike flakes from cores. Error bars indicate the 95% credible intervals
(Bayesian analogue to the Confidence Interval, after Morgan et al. 2014: 4) 178
Fig. 11.20 Changes in the probability of striking a viable flake along a transmission
chain. For verbal and gestural teaching, this probability declined markedly
with distance from the original teacher, so that there was little to no
difference among the different kinds of transmission by time 5. (After
Morgan et al. 2014: 4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Fig. 12.1 Archaeological descriptions of pottery emphasize the radial section (a cross-
section parallel to the pot’s axis of symmetry to provide a profile from rim to
base) and tangential view (i.e., view of the vessel’s exterior surface). A
transverse view (looking down at the rim or up at the base) is more important
for vessels that are not circular in plan or that have unusual features on base,
interior, or rim. Illustrations of pottery typically employ “cutaway” views
that show a radial section as well as a reconstruction of the exterior and
interior surfaces of the vessel, as though the illustrator had cut away
one-quarter of the vessel (left) . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . .. . 186
Fig. 12.2 An example of segmentation rules for painted decoration on the exterior of
Pueblo vessels (after Bunzel 1972). Vessels like the one at upper left are
divided into zones of actual or potential decoration both vertically and
radially. The rectangles show the segmentation of decorative panels as
viewed tangentially (from the side). The circles at right show the radial
segments as viewed transversely (from the top) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Another random document with
no related content on Scribd:
evident that the strategic value of its position is very high. Such again
is, and to a greater degree was, the position of England. The trade of
Holland, Sweden, Russia, Denmark, and that which went up the
great rivers to the interior of Germany, had to pass through the
Channel close by her doors; for sailing-ships hugged the English
coast. This northern trade had, moreover, a peculiar bearing upon
sea power; for naval stores, as they are commonly called, were
mainly drawn from the Baltic countries.
But for the loss of Gibraltar, the position of Spain would have been
closely analogous to that of England. Looking at once upon the
Atlantic and the Mediterranean, with Cadiz on the one side and
Cartagena on the other, the trade to the Levant must have passed
under her hands, and that round the Cape of Good Hope not far from
her doors. But Gibraltar not only deprived her of the control of the
Straits, it also imposed an obstacle to the easy junction of the two
divisions of her fleet.
At the present day, looking only at the geographical position of
Italy, and not at the other conditions affecting her sea power, it
would seem that with her extensive sea-coast and good ports she is
very well placed for exerting a decisive influence on the trade route to
the Levant and by the Isthmus of Suez. This is true in a degree, and
would be much more so did Italy now hold all the islands naturally
Italian; but with Malta in the hands of England, and Corsica in those
of France, the advantages of her geographical position are largely
neutralized. From race affinities and situation those two islands are
as legitimately objects of desire to Italy as Gibraltar is to Spain. If the
Adriatic were a great highway of commerce, Italy’s position would be
still more influential. These defects in her geographical
completeness, combined with other causes injurious to a full and
secure development of sea power, make it more than doubtful
whether Italy can for some time be in the front rank among the sea
nations.
As the aim here is not an exhaustive discussion, but merely an
attempt to show, by illustration, how vitally the situation of a country
may affect its career upon the sea, this division of the subject may be
dismissed for the present; the more so as instances which will further
bring out its importance will continually recur in the historical
treatment. Two remarks, however, are here appropriate.
Circumstances have caused the Mediterranean Sea to play a
greater part in the history of the world, both in a commercial and a
military point of view, than any other sheet of water of the same size.
Nation after nation has striven to control it, and the strife still goes
on. Therefore a study of the conditions upon which preponderance in
its waters has rested, and now rests, and of the relative military
values of different points upon its coasts, will be more instructive
than the same amount of effort expended in another field.
Furthermore, it has at the present time a very marked analogy in
many respects to the Caribbean Sea,—an analogy which will be still
closer if a Panama canal route ever be completed. A study of the
strategic conditions of the Mediterranean, which have received
ample illustration, will be an excellent prelude to a similar study of
the Caribbean, which has comparatively little history.
The second remark bears upon the geographical position of the
United States relatively to a Central-American canal. If one be made,
and fulfil the hopes of its builders, the Caribbean will be changed
from a terminus, and place of local traffic, or at best a broken and
imperfect line of travel, as it now is, into one of the great highways of
the world. Along this path a great commerce will travel, bringing the
interests of the other great nations, the European nations, close
along our shores, as they have never been before. With this it will not
be so easy as heretofore to stand aloof from international
complications. The position of the United States with reference to
this route will resemble that of England to the Channel, and of the
Mediterranean countries to the Suez route. As regards influence and
control over it, depending upon geographical position, it is of course
plain that the center of the national power, the permanent base,[19] is
much nearer than that of other great nations. The positions now or
hereafter occupied by them on island or mainland, however strong,
will be but outposts of their power; while in all the raw materials of
military strength no nation is superior to the United States. She is,
however, weak in a confessed unpreparedness for war; and her
geographical nearness to the point of contention loses some of its
value by the character of the Gulf coast, which is deficient in ports
combining security from an enemy with facility for repairing
warships of the first class, without which ships no country can
pretend to control any part of the sea. In case of a contest for
supremacy in the Caribbean, it seems evident from the depth of the
South Pass of the Mississippi, the nearness of New Orleans, and the
advantages of the Mississippi Valley for water transit, that the main
effort of the country must pour down that valley, and its permanent
base of operations be found there. The defense of the entrance to the
Mississippi, however, presents peculiar difficulties; while the only
two rival ports, Key West and Pensacola, have too little depth of
water, and are much less advantageously placed with reference to the
resources of the country. To get the full benefit of superior
geographical position, these defects must be overcome. Furthermore,
as her distance from the Isthmus, though relatively less, is still
considerable, the United States will have to obtain in the Caribbean
stations fit for contingent, or secondary, bases of operations; which
by their natural advantages, susceptibility of defense, and nearness
to the central strategic issue, will enable her fleets to remain as near
the scene as any opponent. With ingress and egress from the
Mississippi sufficiently protected, with such outposts in her hands,
and with the communications between them and the home base
secured, in short, with proper military preparation, for which she has
all necessary means, the preponderance of the United States on this
field follows, from her geographical position and her power, with
mathematical certainty.
“Strategy,” says Jomini, speaking of the art of war on land, “is the art
of making war upon the map, and comprehends the whole theater of
warlike operations. Grand tactics is the art of posting troops upon
the battlefield, according to the accidents of the ground; of bringing
them into action; and the art of fighting upon the ground in
contradistinction to planning upon a map. Its operations may extend
over a field of ten or twelve miles in extent. Strategy decides where to
act. Grand tactics decides the manner of execution and the
employment of troops,” when, by the combinations of strategy, they
have been assembled at the point of action.
... Between Strategy and Grand Tactics comes logically Logistics.
Strategy decides where to act; Logistics is the act of moving armies; it
brings the troops to the point of action and controls questions of
supply; Grand Tactics decides the methods of giving battle.
5. Fundamental Principles[21]
The situation here used in illustration is taken from the Thirty Years’
War, 1618–1648, in which the French House of Bourbon opposed the
House of Austria, the latter controlling Spain, Austria, and parts of
Germany. France lay between Spain and Austria; but if Spain
commanded the sea, her forces could reach the field of conflict in
central Europe either by way of Belgium or by way of the Duchy of
Milan in northern Italy, both of which were under her rule.
[The upper course of the Danube between Ulm and Ratisbon is
also employed to illustrate central position, dominating the great
European theater of war north of the Alps and east of the Rhine.—
Editor.]
The situation of France relatively to her two opponents of this
period—Spain and Austria—illustrates three elements of strategy, of
frequent mention, which it is well here to name and to define, as well
as to illustrate by the instance before you.
1. There is central position, illustrated by France; her national
power and control interposing by land between her enemies. Yet not
by land only, provided the coast supports an adequate navy; for, if
that be the case, the French fleet also interposes between Spanish
and Italian ports. The Danube is similarly an instance of central
position.
2. Interior lines. The characteristic of interior lines is that of the
central position prolonged in one or more directions, thus favoring
sustained interposition between separate bodies of an enemy; with
the consequent power to concentrate against either, while holding
the other in check with a force possibly distinctly inferior. An interior
line may be conceived as the extension of a central position, or as a
series of central positions connected with one another, as a
geometrical line is a continuous series of geometrical points. The
expression “Interior Lines” conveys the meaning that from a central
position one can assemble more rapidly on either of two opposite
fronts than the enemy can, and therefore can utilize force more
effectively. Particular examples of maritime interior lines are found
in the route by Suez as compared with that by the Cape of Good
Hope, and in Panama contrasted with Magellan. The Kiel Canal
similarly affords an interior line between the Baltic and North Sea, as
against the natural channels passing round Denmark, or between the
Danish Islands,—the Sound and the two Belts.[22] These instances of
“Interior” will recall one of your boyhood’s geometrical theorems,
demonstrating that, from a point interior to a triangle, lines drawn to
two angles are shorter than the corresponding sides of the triangle
itself. Briefly, interior lines are lines shorter in time than those the
enemy can use. France, for instance, in the case before us, could
march twenty thousand men to the Rhine, or to the Pyrenees, or
could send necessary supplies to either, sooner than Spain could
send the same number to the Rhine, or Austria to the Pyrenees,
granting even that the sea were open to their ships.
3. The position of France relatively to Germany and Spain
illustrates also the question of communications. “Communications”
is a general term, designating the lines of movement by which a
military body, army or fleet, is kept in living connection with the
national power. This being the leading characteristic of
communications, they may be considered essentially lines of
defensive action; while interior lines are rather offensive in
character, enabling the belligerent favored by them to attack in force
one part of the hostile line sooner than the enemy can reinforce it,
because the assailant is nearer than the friend. As a concrete
instance, the disastrous attempt already mentioned, of Spain in 1639
to send reinforcements by the Channel, followed the route from
Corunna to the Straits of Dover. It did so because at that particular
moment the successes of France had given her control of part of the
valley of the Rhine, closing it to the Spaniards from Milan; while the
more eastern route through Germany was barred by the Swedes, who
in the Thirty Years’ War were allies of France. The Channel therefore
at that moment remained the only road open from Spain to the
Netherlands, between which it became the line of communications.
Granting the attempt had been successful, the line followed is
exterior; for, assuming equal rapidity of movement, ten thousand
men starting from central France should reach the field sooner.
The central position of France, therefore, gave both defensive and
offensive advantage. In consequence of the position she had interior
lines, shorter lines, by which to attack, and also her communications
to either front lay behind the front, were covered by the army at the
front; in other words, had good defense, besides being shorter than
those by which the enemy on one front could send help to the other
front. Further, by virtue of her position, the French ports on the
Atlantic and Channel flanked the Spanish sea communications.
At the present moment, Germany and Austria-Hungary, as
members of the Triple Alliance, have the same advantage of central
and concentrated position against the Triple Entente, Russia, France,
and Great Britain.
Transfer now your attention back to the Danube when the scene of
war is in that region; as it was in 1796, and also frequently was
during the period of which we are now speaking.... You have seen
before, that, if there be war between Austria and France, as there so
often was, the one who held the Danube had a central position in the
region. Holding means possession by military power, which power
can be used to the full against the North or against the South—
offensive power—far more easily than the South and North can
combine against him; because he is nearer to each than either is to
the other. (See map.) Should North wish to send a big reinforcement
to South, it cannot march across the part of the Danube held, but
must march around it above or below; exactly as, in 1640,
reinforcements from Spain to the Rhine had, so to say, to march
around France. In such a march, on land, the reinforcement making
it is necessarily in a long column, because roads do not allow a great
many men to walk abreast. The road followed designates in fact the
alignment of the reinforcement from day to day; and because its
advance continually turns the side to the enemy, around whom it is
moving, the enemy’s position is said to flank the movement,
constituting a recognized danger. It makes no difference whether the
line of march is straight or curved; it is extension upon it that
constitutes the danger, because the line itself, being thin, is
everywhere weak, liable to an attack in force upon a relatively small
part of its whole. Communications are exposed, and the enemy has
the interior line....
This is an illustration of the force of Napoleon’s saying, that “War
is a business of positions.” All this discussion turns on position; the
ordinary, semi-permanent, positions of Center, North, and South; or
the succession of positions occupied by the detachment on that line
of communications along which it moves. This illustrates the
importance of positions in a single instance, but is by no means
exhaustive of that importance. Fully to comprehend, it is necessary
to study military and naval history; bearing steadily in mind
Napoleon’s saying, and the definitions of central position, interior
lines, and communications.
Take, for example, an instance so recent as to have been
contemporary with men not yet old,—the Turkish position at Plevna
in 1877. This stopped the Russian advance on Constantinople for
almost five months. Why? Because, if they had gone on, Plevna
would have been close to their line of communications, and in a
central position relatively to their forces at the front and those in the
rear, or behind the Danube. It was also so near, that, if the enemy
advanced far, the garrison of Plevna could reach the only bridge
across the Danube, at Sistova, and might destroy it, before help could
come; that is, Plevna possessed an interior line towards a point of the
utmost importance. Under these circumstances, Plevna alone