Description and Classification of Hawaiian Adzes

Table of Contents

1 Project Tracking

1.1 Adze Weight

Quantiles of adze weight in grams
0%25%50%75%100%
4391388413466
  • The weights of complete adze blades are plotted on a log scale to differentiate among the lighter blades
  • The division between "light" and "heavy" blades is about 398 g

r/adze_wt_log.png

Histogram of adze weights on a logarithmic scale

1.2 Adze Length

Quantiles of adze length in mm
0%25%50%75%100%
2561100201368
  • The lengths of the measured adze blades are plotted on a log scale to differentiate among the smaller blades
  • The dividing line between "short" and "long" blades is about 120 mm

r/adze_len_log.png

Histogram of adze lengths on a logarithmic scale

1.3 Edge Width

Quantiles of whole adze edge widths in mm
0%25%50%75%100%
7263553119

r/adze_cutting_edge.png

Histogram of cutting edge widths in mm

1.4 Edge Angle

Quantiles of directly measured edge angles in degrees
0%25%50%75%100%
545506095

r/adze_edge_angle.png

Histogram of directly measured cutting edge angles

Quantiles of calculated edge angles in degrees
0%25%50%75%100%
1237434876

r/adze_trig_angle.png

Histogram of trigonometric measure of cutting edge angle

1.5 Adze Type

Secondary adzes have been fashioned using another tool as the raw material. They represent recycling of broken tools.

Distribution of primary and secondary adzes
Var1Freq
primary855

1.6 Bevel Shape

Bevel shape
Var1Freq
3
concave1
convex813
flat38

1.7 Edge Shape A

  • Viewed from the back of the blade
Edge shape
Var1Freq
angled27
convex245
straight583

1.8 Edge Shape B

  • Viewed with the front superior
Edge shape
Var1Freq
convex64
straight791

1.9 Face Reduction

  • Reduction of the face near the poll to facilitate lashing the blade to the handle
Reduction of face
Var1Freq
false141
true714

1.10 Axes

Under the heading "Axes and Adzes," Brigham discusses two axes, one (4603) a ground stone implement that he describes as "a great stone wedge" (p. 74). He also illustrates as "a more common form of axe" tool 3141 on Plate LVII "Hawaiian Stone Adzes."

identifierstorage_location
150-OA-B1-86-337Tray 2
250-OA-B1-75-518Tray 2
3O10-200Tray 1
450-0a-G6-32-Feature 113Tray 5
550-Oa-G5-152-150Tray 5
6H100 E10-2Tray 6
731271) 3121-3127 Hawaiian Adzes Unhafted
831413) 3140-3144
945614560-4572
101045410441-10463
111102111021-11102
12B.1397B.1392-B.1438
13B2518B.2508 - B.3714
14B.1700.KB.1700-1800
15B.1671C.1671-C.1710
16D.4030D.4076-4083

2 Exploratory Analyses

2.1 Pairwise Comparison

  • Jenny comments "Were you thinking of running some stats too at the end – just to see in fact what attributes are most highly correlated with one another? This may allow us to define grouped functional variables."
  • This plot compares adze blade weight, length, edge width, shoulder width at front, shoulder width at back and shoulder thickness

r/adze_pairs.png

Pairwise comparison of six metric variables

2.2 Percussion

  • An attempt to measure percussive potential of blades
  • Percussion index is defined as weight/width_edge
  • On the scatterplot values above the smooth line have high percussive potential

r/percussion.png

Scatterplot of measures used to calculate percussion index

  • Percussive potential increases to the right on the histogram
  • A percussion index of 8 divides the adzes into blades intended for low impact and those designed to transmit lots of force to the work

r/percussion-index.png

Histogram of percussion index

2.3 Blade Plan Index

  • The blade plan index relates the shoulder width to the edge width
  • The plan can flare to maximize edge width, or narrow to minimize edge width
  • Values less than 1 indicate a narrow edge
  • Values greater than 1 indicate a wide edge
Quantiles of blade plan index
0%25%50%75%100%
0.521.041.121.221.95

r/blade-plan.png

Histogram of blade plan index

Error in 1:ncol(xx) : argument of length 0

2.4 Indices Remarquables

  • These are from Jose Garanger's paper, Herminettes lithiques oceaniennes: Elements de typologie

2.4.1 Shoulder index

  • indice de la section transversale principale = \( \dfrac{R-7 \times 100}{R-9} \), where
    • R-7 = hauteur de la section transversale principale (thickness of the standard cross-section)
    • R-9 = largeur de la section transversale principale (width of the standard cross-section)
  • This is due to Buck (1944)
Quantiles of Buck's shoulder index
0%25%50%75%100%
20.0051.8566.6788.69211.11

r/buck-basics.png

Scatterplot of measures used to calculate Buck's shoulder index

r/buck-shoulder-hist.png

Histogram of Buck's shoulder index

r/buck-long-short.png

Buck's shoulder index for long and short adzes

r/buck-heavy-light.png

Buck's shoulder index for heavy and light adzes

2.4.2 Blade size

  • indice de la largeur de la lame = \( \dfrac{R-2 \times 100}{R-9} \), where
    • R-2 = longeur de la face frontale (blade length)
    • R-9 = largeur de la section transversale principale (width of the standard cross-section)
  • This is due to Green and Green (1960)
Quantiles of blade size index
0%25%50%75%100%
45.71145.08196.08266.08577.78

r/green-basics.png

Scatterplot of measures used to calculate blade size index

r/blade-size-hist.png

Histogram of blade size index

Blade size outliers
identifierstorage_locationlength_shoulderwidth_shoulder_frontwidth_shoulder_back
152461813) 4603-5304163534

r/blade-size-heavy.png

Blade size index for heavy and light adzes

r/blade-size-length.png

Blade size index for long and short adzes

2.4.3 Adze length

  • indice de la longeur de l'herminette = \( \dfrac{R-1 \times 100}{R-9} \)
    • R-1 = longeur maximum de l'herminette (length of the tool)
    • R-9 = largeur de la section transversale principale (width of the standard cross-section)
Quantiles of adze length index
0%25%50%75%100%
147.06259.09325.71429.891166.67

r/len-index-basics.png

Scatterplot of measures used to calculate the adze length index

r/adze-length-hist.png

Histogram of adze length index

Error in 1:ncol(xx) : argument of length 0

r/adze-length-long-short.png

Histogram of adze length index for long and short adzes

r/adze-length-heavy.png

Histogram of adze length index for heavy and light adzes

2.4.4 Edge width

  • indice de longeur du tranchant = \( \dfrac{R-4 \times 100}{R-1} \), where
    • R-1 = longeur maximum de l'herminette (length of the tool)
    • R-4 = longeur du fil du tranchant (width of the edge)
  • This is due to Kellum (1966)
Quantiles of edge width index
0%25%50%75%100%
7.4827.0235.4543.7575.61

r/kellum-basics.png

Scatterplot of measures used to calculate Kellum's edge width index

r/edge-width-hist.png

Histogram of edge width index

Error in 1:ncol(xx) : argument of length 0

r/edge-width-long-short.png

Edge width index of long and short adzes

r/edge-width-heavy.png

Edge width index of light and heavy adzes

2.4.5 Butt length

  • indice de longeur du tranchant = \( \dfrac{R-1 \times 100}{R-3} \), where
    • R-1 = longeur maximum de l'herminette (length of the tool)
    • R-3 = longeur du talon (length of the butt)
  • The index is calculated only for tanged adzes
Quantiles of butt length index
0%25%50%75%100%
138.71233.97266.01296.81Inf

r/butt-basics.png

Scatterplot of measures used to calculate the butt length index

r/butt-length-hist.png

Histogram of butt length index

Error in 1:ncol(xx) : argument of length 0

2.5 Exploratory Views

  • Adzes with chins thicker than shoulders prompted this series of graphs

2.5.1 Chin thickness index

  • Chin thickness / shoulder thickness
Chin thickness index Quantiles of chin thickness index
0%25%50%75%100%
0.290.620.730.851.58

r/chin-thick-scatter.png

Scatterplot of measures used to calculate chin thickness index

r/chin-thickness-hist.png

Histogram of the chin thickness index

Error in 1:ncol(xx) : argument of length 0

2.6 Cross-section Classification

  • This is due to Emory, "East Polynesian Adze Relationships"
  • It is based on the relationship of front width (FW) at the shoulder to back width (BW) at the shoulder
  • Emory didn't partition exhaustively
  • The classification used here is:
    • Triangular: \( FW \leq 0.15 \times BW \)
    • Subtriangular: \( 0.15 \times BW < FW \leq 0.3 \times BW \)
    • Trapezoidal: \( 0.3 \times BW < FW \leq 0.8 \times BW \)
    • Quadrangular: \( 0.8 \times BW < FW \leq 1.2 \times BW \)
    • Reversed Trapezoidal: \( 1.2 \times BW < FW \leq 1.7 \times BW \)
    • Reversed Subtriangular: \( 1.7 \times BW < FW \leq 1.85 \times BW \)
    • Reversed Triangular: \( FW > 1.85 \times BW \)
Distribution of cross sections
x_sectionFreq
1Triangular1
2Subtriangular1
3Trapezoidal12
4Quadrangular724
5Reversed Trapezoidal91
6Reversed Subtriangular4
7Reversed Triangular22

r/emory.png

Dot chart of cross-section classification

r/emory-diag.png

Diagnostic scatterplot of cross-section classification

2.6.1 Comments

  • Emory's system is designed to group the Hawaiian adzes in a way that sets them apart from adzes fashioned elsewhere in East Polynesia
  • Its object is the opposite of what we are trying to do
  • There is little variability in the relationship between shoulder width at front and back
  • Reversed triangular adzes tend to be small

2.7 Cross-section Area

This is due to McAlister (2011), Methodological Issues in the Geochemical Characterisation and Morphological Analysis of Stone Tools: A Case Study from Nuku Hiva, Marquesas Islands, East Polynesia.

r/cross-section-area-histogram.png

Histogram of cross-section area

2.8 Shoulder height by adze size

  • In the lab it seems that tall adzes are typically large
  • Here, "large" is captured by the length of the tool

r/shoulder-height-size.png

Scatterplot showing the relationship between adze blade length and shoulder height index

  • Here, "large" is captured by weight

r/shoulder-height-weight.png

Scatterplot showing the relationship between adze blade weight and shoulder height index

2.9 Multivariate Views

  • A simple correspondence analysis on all metric variables

r/ca-1.png

3 EDXRF

  • Peter Mills and Steven Lundblad are characterizing the chemical composition of the BPBM adzes using a non-destructive technique
  • Their laboratory produces its most precise results for the mid-Z trace elements Rb, Sr, Y, Zr, and Nb

3.1 Elemental Review

3.1.1 Sodium

r/edxrf-na2o.png

Histogram of Na2O %

3.1.2 Magnesium

r/edxrf-mgo.png

Histogram of MgO %

3.1.3 Aluminum

r/edxrf-al2o3.png

Histogram of Al2O3 %

3.1.4 Silicon

r/edxrf-sio2.png

Histogram of Silicon dioxide %

3.1.5 Potassium

r/edxrf-k2o.png

Histogram of K2O %

3.1.6 Calcium

r/edxrf-cao.png

Histogram of Calcium oxide %

3.1.7 Titanium

r/edxrf-tio2.png

Histogram of Titanium dioxide %

3.1.8 Vanadium

r/edxrf-v.png

Histogram of Vanadium in parts per million

3.1.9 Manganese

r/edxrf-mno.png

Histogram of Manganese oxide in parts per million

3.1.10 Iron

r/edxrf-fe.png

Histogram of Iron %

3.1.11 Nickel

r/edxrf-ni.png

Histogram of Nickel in parts per million

3.1.12 Copper

r/edxrf-cu.png

Histogram of Copper in parts per million

3.1.13 Zinc

r/edxrf-zn.png

Histogram of Zinc in parts per million

3.1.14 Rubidium

r/edxrf-rb.png

Histogram of Rubidium in parts per million

3.1.15 Strontium

r/edxrf-sr.png

Histogram of Strontium in parts per million

3.1.16 Yttrium

r/edxrf-y.png

Histogram of Yttrium in parts per million

3.1.17 Zirconium

r/edxrf-zr.png

Histogram of Zirconium in parts per million

3.1.18 Niobium

r/edxrf-nb.png

Histogram of Niobium in parts per million

3.1.19 Barium

r/edxrf-ba.png

Histogram of Barium in parts per million

3.2 Major Elements

  • SiO2, Al2O3, Fe2O3, MgO, CaO, Na2O, K2O

3.2.1 Major Element Pairs

r/edxrf-major-pairs.png

Pairs plot of major elements

3.2.2 Total Alkali Silica Diagram

  • Need to add classification

r/edxrf-TAS.png

3.3 Minor Elements

3.3.1 Minor Element Pairs

r/edxrf-minor-pairs.png

Pairs plot of minor elements

3.3.2 Mid-Z trace element pairs

r/edxrf-mid-z-pairs.png

Pairs plot of mid-Z trace elements

3.4 Multiple ANOVA

  • Comparisons of UH and UH Hilo lab results for mid-Z trace elements and major elements both indicate inter-laboratory variability too great for direct comparability

3.4.1 Mid-Z trace elements

  • Multiple analysis of variance for mid-Z trace elements suggests that there are large differences between the UH and UH Hilo laboratories
  • The differences are especially pronounced for Niobium and Yttrium
  • Results suggest that analyses of mid-Z trace elements from these two labs are not directly comparable

null device 1 null device 1

Comparison of median concentrations of Mid-Z trace elements
labrb_ppmsr_ppmy_ppmzr_ppmnb_ppm
1UH30.00538.0038.50313.5043.50
2UH-Hilo30.10576.7943.33323.5736.44
null device 1 null device 1
Comparison of median concentrations of Mid-Z trace elements
labrb_ppmsr_ppmy_ppmzr_ppmnb_ppm
1UH30.00538.0038.50313.5043.50
2UH-Hilo30.10576.7943.33323.5736.44
labrb_ppmsr_ppmy_ppmzr_ppmnb_ppm
1UH30.00538.0038.50313.5043.50
2UH-Hilo30.10576.7943.33323.5736.44
Summary of analysis of variance
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.090.090.010.94
Residuals820.0013814.6016.85
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.090.090.010.94
Residuals820.0013814.6016.85
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.090.090.010.94
Residuals820.0013814.6016.85
Comparison of median concentrations of major elements
labna2o_percentmgo_percental2o3_percentsio2_percentk2o_percentcao_percenttio2_percent
1UH3.095.0213.4948.211.159.384.00
2UH-Hilo2.573.2012.9045.130.948.513.63
labna2o_percentmgo_percental2o3_percentsio2_percentk2o_percentcao_percenttio2_percent
1UH3.095.0213.4948.211.159.384.00
2UH-Hilo2.573.2012.9045.130.948.513.63
labna2o_percentmgo_percental2o3_percentsio2_percentk2o_percentcao_percenttio2_percent
1UH3.095.0213.4948.211.159.384.00
2UH-Hilo2.573.2012.9045.130.948.513.63

3.4.2 Major elements

  • There appear to be significant inter-laboratory differences for most major elements
Summary of analysis of variance in concentrations of major elements
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.630.634.080.04
Residuals820.00126.540.15
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.630.634.080.04
Residuals820.00126.540.15
DfSum SqMean SqF valuePr(>F)
as.factor(lab)1.000.630.634.080.04
Residuals820.00126.540.15

3.5 Diagnostic Plots

  • The plots in this section currently show the adzes measured using EDXRF at Hilo against material from quarries contained in the spreadsheet QuarryData.xls
  • Bishop Museum adzes are indicated in the plots as 'unknown'
  • Quarry data reported by Sinton and Sinoto are not included because EDXRF data from Hilo are not directly comparable with them
  • Plots using the most reliably measured elements are followed by plots using elements measured with less precision
  • Note that issues of measurement reliability are unlikely to yield false positive results

3.5.1 Strontium Zirconium plot

  • This plot is a favorite of Mills and Lundblad, based on PCA of Mid-Z trace elements Rb, Sr, Y, Zr, and Nb

r/sr-zr-scatter.png

Scatterplot of Strontium and Zirconium

3.5.2 Strontium Niobium plot

r/sr-nb-scatter.png

Scatterplot of Strontium and Niobium

3.5.3 Zirconium Niobium plot

r/nb-zr-scatter.png

Scatterplot of Zirconium and Niobium

3.5.4 Weisler and Kirch

  • This plot was used by Weisler and Kirch to source Cook Island polished adze flakes analyzed by EDXRF (M.I. Weisler and P.V. Kirch, Interisland and interarchipelago transfer of stone tools in prehistoric Polynesia, PNAS 93:1381-5)

r/nb-zr-sr-scatter.png

Scatterplot of Niobium to Strontium and Zirconium to Strontium ratios

3.5.5 Fractionation Trends

  • This plot is after Melinda Allen and Kevin Johnston's work in the Cook Islands

r/fractionation.png

Fractionation trends at quarries and among adzes

3.5.6 Titanium Iron plot

r/ti-fe-scatter.png

Scatterplot of Titanium and Iron

3.5.7 Titanium Zirconium plot

r/ti-zr-scatter.png

Scatterplot of Titanium and Zirconium

3.5.8 Zirconium Nickel plot

  • Note that Nickel usefully discriminates most of the O`ahu quarry rock from Hawai`i Island

r/ni-zr-scatter.png

Scatterplot of Zirconium and Nickel

3.6 Discriminant Analysis

  • Linear discriminant analysis appears to work with the EDXRF quarry data, using the mid-z trace elements Rubidium, Strontium, Yttrium, Zirconium, and Niobium
Accuracy of the linear discriminant analysis
0.9985
0.9985
0.9985
  • Linear discriminant analysis using jacknifed predictions indicates that the mid-z elements discriminate among the six quarries with a high degree of accuracy
 Haleakala   Kilauea Mauna Kea      Nu`u    Pololu  Waiahole 
1.0000000 0.9846154 1.0000000 1.0000000 1.0000000 0.9935484
  • Here is how the linear discriminate analysis assigns adzes to quarries
Assignment of unknowns to sources
BPBMHaleakalaKilaueaMauna KeaNu`uPololuWaiaholeunknown
1985.134.01 adze00.10000.640.27
50-HA-1349-G-5 adze00.8500000.14
50-HA-900-P24-1 microadze0010000
50-M9-B8 38 #3 microadze00.040.13000.050.78
50-Ma-B8-41-TP4-8 microad00.8600000.14
50-MA-B9-26-5 adze1000000
H100 E10-2 microadze00.610000.250.14
H14 #3 adze00.730000.220.05
H14 23700.750000.040.2
H14 390010000
H14 40 adze00.900000.1
H14-266 adze00.710000.010.28
H14-267 adze0000010
H1E15 40010000
H2-5 (H52-5 on adze)00.340000.030.63
H2ETI-380001000
H2WT4-21 adze00.780000.130.09
H-47 A 1-1 microadze00.7300000.27
H50-5(Q3) adze00.70000.010.29
H8 F12-100.8600000.14
H8 Sur. 4 adze0010000
H9826 16-6 small adze00.70000.050.25
HA B22 65 D-102 microadze00.8600000.14
HA B22-22 TP3-2 microadze00.8800000.11
HA B22-65 B-65 1 microadz00.770000.030.2
HA B22-B5-B-66 narrow mic00.760000.030.2
HA-B22-65-G31 microadze000000.990.01
HA-D4-27-1 microadze00.830000.020.15
HA-D4-48-6 microadze00.580000.090.33
Haw 131 adze0010000
Haw 66 adze00.9300000.07
HAW130 adze0100000
HAW-167 adze00.750000.020.23
HAW-7 adze00.8200000.18
HAW-88 adze0000010
HAW-89 adze00.8600000.13
Haw-90 adze00.70000.010.29
HBC 524 adze00.8600000.14
HI E151 Frag 2 out of 200.9500000.05
HI G2-3 Frag 1 out of 200.5700000.42
HV-76 E9-7 microadze00.7800000.21
K 16 1 2-1 adze00.6800000.32
No Number 1 out of 2 adze00.790000.090.12
No Number 2 out of 2 adze0000010
S 27-3 1-664 adze0000100
BPBMHaleakalaKilaueaMauna KeaNu`uPololuWaiaholeunknown
1985.134.01 adze00.10000.640.27
50-HA-1349-G-5 adze00.8500000.14
50-HA-900-P24-1 microadze0010000
50-M9-B8 38 #3 microadze00.040.13000.050.78
50-Ma-B8-41-TP4-8 microad00.8600000.14
50-MA-B9-26-5 adze1000000
H100 E10-2 microadze00.610000.250.14
H14 #3 adze00.730000.220.05
H14 23700.750000.040.2
H14 390010000
H14 40 adze00.900000.1
H14-266 adze00.710000.010.28
H14-267 adze0000010
H1E15 40010000
H2-5 (H52-5 on adze)00.340000.030.63
H2ETI-380001000
H2WT4-21 adze00.780000.130.09
H-47 A 1-1 microadze00.7300000.27
H50-5(Q3) adze00.70000.010.29
H8 F12-100.8600000.14
H8 Sur. 4 adze0010000
H9826 16-6 small adze00.70000.050.25
HA B22 65 D-102 microadze00.8600000.14
HA B22-22 TP3-2 microadze00.8800000.11
HA B22-65 B-65 1 microadz00.770000.030.2
HA B22-B5-B-66 narrow mic00.760000.030.2
HA-B22-65-G31 microadze000000.990.01
HA-D4-27-1 microadze00.830000.020.15
HA-D4-48-6 microadze00.580000.090.33
Haw 131 adze0010000
Haw 66 adze00.9300000.07
HAW130 adze0100000
HAW-167 adze00.750000.020.23
HAW-7 adze00.8200000.18
HAW-88 adze0000010
HAW-89 adze00.8600000.13
Haw-90 adze00.70000.010.29
HBC 524 adze00.8600000.14
HI E151 Frag 2 out of 200.9500000.05
HI G2-3 Frag 1 out of 200.5700000.42
HV-76 E9-7 microadze00.7800000.21
K 16 1 2-1 adze00.6800000.32
No Number 1 out of 2 adze00.790000.090.12
No Number 2 out of 2 adze0000010
S 27-3 1-664 adze0000100

3.7 Principal Components Analysis

Mills et al. published a revealing PCA plot of Kaua`i materials. This is an attempt to replicate their analysis with the Bishop Museum adze data. The best conclusion to draw from this analysis might be that the discriminate analysis depends fairly strongly on the characterizations of the known quarries. As additional source data are collected, the results of the principal components analysis are likely to change markedly.

r/pca-results.png

4 Database Structure

'data.frame':   855 obs. of  31 variables:
 $ id                  : int  1 3 7 13 14 18 20 30 31 23 ...
 $ identifier          : chr  "OA B1-30-29" "OA-B1-30-27" "50-OA-B1-30-R6-1" "OA-C4-171#1 MVP.60" ...
 $ storage_location    : chr  "Tray 1" "Tray 1" "Tray 1" "Tray 1" ...
 $ site                : chr  "50-Oa-B1-30" "50-Oa-B1-30" "50-Oa-B1-30" "50-OA-C4-171" ...
 $ weight              : int  111 35 24 31 29 245 407 1355 602 263 ...
 $ adze_type           : chr  "primary" "primary" "primary" "primary" ...
 $ bevel               : chr  "single" "single" "single" "single" ...
 $ edge_present        : chr  "true" "true" "true" "true" ...
 $ chin_present        : chr  "true" "true" "true" "true" ...
 $ shoulder_present    : chr  "true" "true" "true" "true" ...
 $ poll_present        : chr  "true" "true" "true" "true" ...
 $ length_poll         : int  92 58 51 65 58 128 135 253 186 118 ...
 $ length_shoulder     : int  48 29 21 31 35 72 91 155 110 64 ...
 $ length_chin         : int  11 9 5 5 10 10 14 40 30 26 ...
 $ width_edge          : int  33 21 25 21 23 46 52 49 51 40 ...
 $ width_shoulder_front: int  29 21 22 17 20 45 44 46 41 36 ...
 $ width_shoulder_back : int  30 20 22 18 20 46 40 39 39 34 ...
 $ thickness_shoulder  : int  16 10 8 7 10 17 29 58 38 26 ...
 $ thickness_chin      : int  11 8 5 4 9 11 16 38 19 24 ...
 $ edge_angle          : int  36 36 34 34 34 38 37 46 39 40 ...
 $ bevel_shape         : chr  "convex" "convex" "convex" "convex" ...
 $ edge_shape_a        : chr  "straight" "straight" "straight" "straight" ...
 $ edge_shape_b        : chr  "straight" "straight" "straight" "straight" ...
 $ face_reduced        : chr  "true" "true" "true" "true" ...
 $ butt_angle          : int  10 10 0 14 0 8 6 5 10 0 ...
 $ color_value         : chr  "4" "4" "3" "2.5" ...
 $ complete            : chr  "complete" "complete" "complete" "complete" ...
 $ broken              : chr  "" "" "" "" ...
 $ reworked            : chr  "" "" "" "" ...
 $ polish              : chr  "present" "present" "present" "present" ...
 $ notes               : chr  "Small dip present at the shoulder.\r\nBlade is partially chipped in back." "Small chip on the blade/top shoulder." "Edge is jagged and has been chipped/broken with use, although there is no apparent reworking." "" ...

5 Problems

5.1 Duplicate entries

If there are duplicate entries, then they will show in a table below. Otherwise, nothing will appear.

5.2 Missing measures on whole adzes

The queries in this section check for missing values. If missing values are found, a table is produced. If no missing values are found, then no table is produced.

Error in 1:ncol(xx) : argument of length 0

Error in 1:ncol(xx) : argument of length 0

5.3 Logically incorrect entries

These two queries catch instances of logically impossible values for length values. The first finds shoulder lengths longer than poll lengths. The second finds chin lengths longer than shoulder lengths.

Error in 1:ncol(xx) : argument of length 0

Incorrect shoulder or chin length
identifierstorage_locationlength_shoulderlength_chin
315B.1523-O./L.996B.1441-B.15544041
593D.73C.10349-D.6102025

5.4 Untanged adzes

This code block is designed to catch instances of shoulder measures that stray from the convention that the shoulder of an untanged adze is its mid-point. It selects for untanged adzes where the shoulder length is more than a mm away from the midpoint.

Cross section not at midpoint of untanged adze
identifierstorage_locationlength_polllength_shoulder
26O2-R12-Ma-102Tray 12517
28O10-24Tray 15743
6450-0a-G6-32-F123-surf-23Tray 53325

6 Software Version Information

6.1 Org-mode

Org-mode version 7.8.09 (release_7.8.09.201.g55cdbc)

6.2 R

platform
x86_64-apple-darwin9.8.0
arch
x86_64
os
darwin9.8.0
system
x86_64, darwin9.8.0
status
major
2
minor
14.0
year
2011
month
10
day
31
svn rev
57496
language
R
version.string
R version 2.14.0 (2011-10-31)

7 Reading

7.1 Turner thesis

7.1.1 Function types

7.1.1.1 Type 1

Tools designed for "[h]eavy roughing out work to remove excess timber and for tree felling" 122;

Requires a "large wide-bladed, thick, heavy tool with a robust edte angle for chopping and splitting out timber between scarfs" 122;

Functional requirements:

  • Long and heavy for maximum impact
  • Thick cross section to prevent breakage and to allow maximum force to be used
  • Robust edge angle, around 50 degrees
  • Wide blade to maximize size of wood chunks removed
  • Symmetry of form
7.1.1.2 Type 2

Tools designed for "[f]ine timber shaping and dressing" 148; require shallow sectioned, wide bladed and low edge angle tools.

Functional requirements:

  • Shallow section to allow a low edge angle
  • Sufficient length for a follow through stroke
  • Wide blade to maximize contact area
  • Well balanced for an even stroke requires symmetry
  • Relatively straight blade
7.1.1.3 Type 3

Tools designed for "[s]haping curved (concave) surfaces, edge trimming and other precision shaping tasks" 168; require rounded shallow sectioned tools with curved blades.

Functional requirements:

  • Some length for the follow-through stroke and for reaching difficult areas
  • Frontal convexity, marked blade curvature and low edge angles
  • Blade width relative to curvature of area being adzed
  • Size relative to area being adzed
7.1.1.4 Type 4

Tools designed for "[m]aking deep V-shaped cuts in wood to allow wider bladed tools access to split and chop out large chunks of timber" 180; require thick, heavy, large narrow-bladed tool with robust edge angle.

Functional requirements:

  • Large size for maximum impact
  • Thick cross section to withstand for of impact
  • Robust edge angle to absorb impact and overcome wood resistance
  • Narrow blade for maximum penetration
7.1.1.5 Type 5

Laterally hafted tools, axes

7.1.1.6 Type 6

Tools for gouging and chiseling

7.1.1.7 Type 7

Small to medium tools for everyday chopping and trimming tasks

Date: 2012-05-13T17:35-1000

Author: Thomas S. Dye

Org version 7.8.09 with Emacs version 23

Validate XHTML 1.0