The refinement of handaxes, defined as increasing planview symmetry and profile thinness, has been used to distinguish Early and Late Acheulian assemblages. However, recent studies have found that this is not a ubiquitous trend throughout the Acheulian industry. Yet, research suggests that Late Acheulian handaxes differ from earlier forms in the complexity and extent of thinning procedures. To test the discriminatory power of thinning in distinguishing Early and Late Acheulian handaxes, cross-sectional shapes are compared through geometric morphometric techniques. Results suggest that analysing cross-sectional outlines can identify variance relating to reduction processes, which differentiate handaxes from Early and Late Acheulian assemblages.
Pilot study comparing the effects of thinning processes on the cross-sectional morphologies of Early and Late Acheulian handaxes
Review badges
Web of Science (Free Access)
Abstract
- Contribute
- pre-publication peer review (FINAL ROUND)
Decision Letter
2020/11/0101-Nov-2020
Dear Dr. Caruana:
I am writing to inform you that we are happy to accept your manuscript entitled "A Pilot Study Comparing the Effects of Thinning Processes on the Cross-Sectional Morphologies of Early from Late Acheulian Handaxes" in its current form for publication in Archaeometry.
Thank you for your contribution. On behalf of the Editors of Archaeometry, we look forward to your continued contributions to the Journal.
Yours sincerely,
Prof. Mark Pollard
Managing Editor, Archaeometry
mark.pollard@rlaha.ox.ac.ukDecision letter by
Cite this decision letter
Author Response
2020/10/15Reviewer Comments and Responses (Round 2)
I would like to thank Prof. Pollard for facilitating a second review of my manuscript, and the referee’s comment have further added small improvements to the argument presented in this study. I have systematically addressed the comments below and I believe that this manuscript has now sufficiently passed the review process. I note the referee’s positive comments regarding my responses to their initial round of comments and general support for the publication of this research.
I. Referee Comments
OPENING STATEMENT
“I would like to thank the author to take int consideration the most part of the corrections and suggestions. He has done a great effort to introduced them into the main text or in an appropriated Supplementary Information section.”
REPLY
I am very thankful to the referee for their encouraging words and helpful comments, which have greatly improved this manuscript. I hope that these final corrections will satisfy the requirements for publication.
MINOR COMMENTS
COMMENT 1. “Page 6, first paragraph, please refer to Figure 1 when explain the cross-sections. There is no a clear definition of this term. Including this call to Figure 1, the reader will find the correct explanation in the Figure caption.”
RESPONSE 1. I have added a reference to figure 1 in the first paragraph of page 6.
COMMENT 2. “Materials section: Due the small size of the sample and the quantity of previous studies, it is mandatory to make at least on paragraph explaining the main technological features. These must be compiled again in the discussion section. For example, the higher variation on tips are related to longer shaping sequences on Late Acheulean handaxes? Or is there more thinning invasive work or by contrast, the final retouch is marginal and concentrated on edges? Or perhaps the final retouch is only located tips? In other words, the morphometry requires a clear technological background behind. It is not necessary to be so extent because, as the author stated, everything is published. But it is necessary to include the main technological aspects.”
RESPONSE 2. I thank the referee for this suggestion, and I do agree that integration of the cross-sectional analysis and technological features of these handaxe samples is required. I have added text to two sections to discuss flaking patterns that characterize handaxes from the site analyzed in the study. In the Materials section on page 15, I have added descriptions of the handaxes from the assemblages, which are tied into the morphometric results in the Discussion section on pages 25-26.
COMMENT 3. “Together with the technological explanation, it is also mandatory to include one or two Figures with tools. The COVID has altered our routines and limited our capabilities, I understand the author’s explanation. Nevertheless, as the technology is published, I suggest to take some of the published pictures and create one/two new figure(s). Then, in the Figure caption, include: Modified from + reference.”
RESPONSE 3. I appreciate the referee’s concern here and typically it is very useful to provide an image of the artefacts. However, I have searched for images of handaxes from Cave of Hearths, Rietputs 15 and Sterkfontein and have found none of decent quality. Images of these assemblages can be found in the publication I have cited in the text, and I feel this is a better option for readers rather then me providing a figure of pixelated, unclear artefacts.
COMMENT 4. “Pages 19-20, Additional Analysis (SOM). This information is really useful but I think it must be integrated, form my point of view, in the Result section, as well as the fisrt paragraph of page 21.
‘Before the cross-sectional results are discussed, it should be noted that the analyses of planview and profile shapes in the SOM confirm that they have a relatively low discriminatory power for distinguishing Early from Late Acheulian handaxes. This corroborates previous studies suggesting that planview and profile perspectives do not predict the affiliation of Acheulian handaxes from South Africa to either the Early or Late periods of this industry (Li et al. 2018; Caruana and Herries, 2020).’
If the author does not readapt this part to integrate it in the main test, it must be moved to Supplem. Information.”
RESPONSE 4. I appreciate the referee’s concern here, although I think it is fair to report on all the GM results carried out in this study within the main text. Past research has confirmed that planview and profile shapes do not distinguish the Rietputs 15 and Cave of Hearths handaxes (i.e. Li et al., 2018). Although another referee wanted to see if the GM methods used in this study confirm this result, hence the additional analysis in the SOM. Using GM methods also show that planview and profile shapes are not great discriminators of Early vs. Late Acheulian handaxes and while not the focus of this study, it is a significant detail. It supports the argument that cross-sectional shapes better distinguish handaxes across time and I feel that this information, while supplementary, strengthens the findings of this study. I would prefer to leave this in as it also attracts attention to the SOM section.
COMMENT 5. “The Figure 2 is so clear. The main differences between all sites comes from tips. Nevertheless, the author could complement the PCA analysis from mid and base cross-sections (Figures 3 and 4) with some statistical analysis, to detect the intra-site variability: 1) Distance and significant test, to check the differences between the different groups; 2) MANOVA analysis with the first 10 PC to check how similar the scatters are.”
RESPONSE 5. I thank the reviewer for this suggestion, and I agree it would be useful to highlight the lack of clear discrimination in PCA graphs for midsection and base proportions. I have added results of MANOVA tests in a paragraph on page 23, which provide insight into the overlap in PCA results.
COMMENT 6. Page 24, Reduction intensity comparisons: in my previous review, I pointed out the low values of r2. In this case, I have to repeat the same idea. In spite of the p values, there is not a significant correlation between PC1 scores and SDI values. In fact, the aouthor say it in the sentence before.
RESPONSE 6. I respect the referee’s concern with my assessment of the SDI and I have noted the insignificance of the relationship between SDI and PC1 scores. I have provided an alternative explanation for this weak correlation on page 25.
Author response by
Cite this author response
- pre-publication peer review (ROUND 2)
Decision Letter
2020/10/1414-Oct-2020
Dear Dr. Caruana:
Manuscript ID ARCH-06-0288.R1 entitled "A Pilot Study Comparing the Effects of Thinning Processes on the Cross-Sectional Morphologies of Early from Late Acheulian Handaxes" which you submitted to Archaeometry, has been reviewed. The comments of the referee(s) are included at the bottom of this letter.
The referee(s) have recommended some minor revisions to your manuscript. Therefore, I invite you to respond to the referee(s)' comments and revise your manuscript.
To revise your manuscript, log into https://mc.manuscriptcentral.com/arch and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions." Under "Actions," click on "Create a Revision." Your manuscript number has been appended to denote a revision.
You will be unable to make your revisions on the originally submitted version of the manuscript. Instead, revise your manuscript using a word processing programmeme and save it on your computer. Please also highlight the changes to your manuscript within the document by using the track changes mode in MS Word or by using bold or coloured text.
Once the revised manuscript is prepared, you can upload it and submit it through your Author Centre.
When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided. You can use this space to document any changes you make to the original manuscript. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response to the referee(s).
IMPORTANT: Your original files are available to you when you upload your revised manuscript. Please delete any redundant files before completing the submission.
Because we are trying to facilitate timely publication of manuscripts submitted to Archaeometry, your revised manuscript should be uploaded as soon as possible. If it is not possible for you to submit your revision in a reasonable amount of time, we may have to consider your paper as a new submission.
Once again, thank you for submitting your manuscript to Archaeometry and I look forward to receiving your revision.
Yours sincerely,
Prof. Mark Pollard
Managing Editor, Archaeometry
mark.pollard@rlaha.ox.ac.ukThe reviewer has raised a further set of comments on the revised manuscript. Please consider the points raised, but please treat them as advisory.
Referee(s)' Comments to Author:
Comments to the Author
I would like to thank the author to take int consideration the most part of the corrections and suggestions. He has done a great effort to introduced them into the main text or in an appropriated Supplementary Information section.Nevertheless, I would like to suggest a few more thing to be consider by the author:
Page 6, first paragraph, please refer to Figure 1 when explain the cross-sections. There is no a clear definition of this term. Including this call to Figure 1, the reader will find the correct explanation in the Figure caption.
Materials section: Due the small size of the sample and the quantity of previous studies, it is mandatory to make at least on paragraph explaining the main technological features. These must be compiled again in the discussion section. For example, the higher variation on tips are related to longer shaping sequences on Late Acheulean handaxes? Or is there more thinning invasive work or by contrast, the final retouch is marginal and concentrated on edges? Or perhaps the final retouch is only located tips? In other words, the morphometry requires a clear technological background behind. It is not necessary to be so extent because, as the author stated, everything is published. But it is necessary to include the main technological aspects.
Together with the technological explanation, it is also mandatory to include one or two Figures with tools. The COVID has altered our routines and limited our capabilities, I understand the author’s explanation. Nevertheless, as the technology is published, I suggest to take some of the published pictures and create one/two new figure(s). Then, in the Figure caption, include: Modified from + reference.
Pages 19-20, Additional Analysis (SOM). This information is really useful but I think it must be integrated, form my point of view, in the Result section, as well as the fisrt paragraph of page 21.
"Before the cross-sectional results are discussed, it should be noted that the analyses of planview and profile shapes in the SOM confirm that they have a relatively low discriminatory power for distinguishing Early from Late Acheulian handaxes. This corroborates previous studies suggesting that planview and profile perspectives do not predict the affiliation of Acheulian handaxes from South Africa to either the Early or Late periods of this industry (Li et al. 2018; Caruana and Herries, 2020)."
If the author does not readapt this part to integrate it in the main test, it must be moved to Supplem. Information.
The Figure 2 is so clear. The main differences between all sites comes from tips. Nevertheless, the author could complement the PCA analysis from mid and base cross-sections (Figures 3 and 4) with some statistical analysis, to detect the intra-site variability: 1) Distance and significant test, to check the differences between the different groups; 2) MANOVA analysis with the first 10 PC to check how similar the scatters are.
Page 24, Reduction intensity comparisons: in my previous review, I pointed out the low values of r2. In this case, I have to repeat the same idea. In spite of the p values, there is not a significant correlation between PC1 scores and SDI values. In fact, the aouthor say it in the sentence before.
Decision letter by
Cite this decision letter
Reviewer report
2020/10/06I would like to thank the author to take int consideration the most part of the corrections and suggestions. He has done a great effort to introduced them into the main text or in an appropriated Supplementary Information section.
Nevertheless, I would like to suggest a few more thing to be consider by the author:
Page 6, first paragraph, please refer to Figure 1 when explain the cross-sections. There is no a clear definition of this term. Including this call to Figure 1, the reader will find the correct explanation in the Figure caption.
Materials section: Due the small size of the sample and the quantity of previous studies, it is mandatory to make at least on paragraph explaining the main technological features. These must be compiled again in the discussion section. For example, the higher variation on tips are related to longer shaping sequences on Late Acheulean handaxes? Or is there more thinning invasive work or by contrast, the final retouch is marginal and concentrated on edges? Or perhaps the final retouch is only located tips? In other words, the morphometry requires a clear technological background behind. It is not necessary to be so extent because, as the author stated, everything is published. But it is necessary to include the main technological aspects.
Together with the technological explanation, it is also mandatory to include one or two Figures with tools. The COVID has altered our routines and limited our capabilities, I understand the author’s explanation. Nevertheless, as the technology is published, I suggest to take some of the published pictures and create one/two new figure(s). Then, in the Figure caption, include: Modified from + reference.
Pages 19-20, Additional Analysis (SOM). This information is really useful but I think it must be integrated, form my point of view, in the Result section, as well as the fisrt paragraph of page 21.
"Before the cross-sectional results are discussed, it should be noted that the analyses of planview and profile shapes in the SOM confirm that they have a relatively low discriminatory power for distinguishing Early from Late Acheulian handaxes. This corroborates previous studies suggesting that planview and profile perspectives do not predict the affiliation of Acheulian handaxes from South Africa to either the Early or Late periods of this industry (Li et al. 2018; Caruana and Herries, 2020)."
If the author does not readapt this part to integrate it in the main test, it must be moved to Supplem. Information.
The Figure 2 is so clear. The main differences between all sites comes from tips. Nevertheless, the author could complement the PCA analysis from mid and base cross-sections (Figures 3 and 4) with some statistical analysis, to detect the intra-site variability: 1) Distance and significant test, to check the differences between the different groups; 2) MANOVA analysis with the first 10 PC to check how similar the scatters are.
Page 24, Reduction intensity comparisons: in my previous review, I pointed out the low values of r2. In this case, I have to repeat the same idea. In spite of the p values, there is not a significant correlation between PC1 scores and SDI values. In fact, the aouthor say it in the sentence before.
Reviewed by
Cite this review
Author Response
2020/08/30I would like to thank the managing editor, Prof. Mark Pollard for facilitating the review of my manuscript and invitation to reply to the two anonymous referees, who both provided helpful and encouraging comments. I note that both referees have recommended this manuscript for publication once their concerns have been addressed. Below, I list my responses to the editor and referees, which lists comments in numerical order followed by my response. I believe that that the revisions have adequately addressed these comments and feel that the quality of the manuscript has been greatly improved by the thoughtful discussion provided by the two referees.
I. Managing Editor's Comments to Author:
COMMENT 1: “The comments are rather long but basically encouraging. I suggest you do what is reasonable, and bear in mind that you can use online supplemental material for additional figures, as requested by one of the reviewers.”
RESPONSE 1: Thank you for inviting me to respond to the referees’ comments. I have found these reviews and comments to be extremely helpful in revising my manuscript, which I now find to be significantly improved in its argument. I feel the revisions sufficiently address the referee’s concerns, and I hope this will warrant publication, which both referees support.
I have created a supplementary online supplementary file (SOM) to accompany this manuscript, as the word count will exceed limits upon addressing some of the reviewer comments. If there are any issues with the number of figures and tables I’ve included in the SOM, I will happily edit this section to adhere to Archaeometry’s regulations.II. Referee 1 Major Concerns:
I am very grateful to Referee 1 for their in-depth review of my work and constructive comments. I would like to thank them for recommending my work for publication as well. Referee 1 begins their review with a summary of the manuscript. It is rare for a reviewer to go to such lengths, but extremely insightful. This summary provides valuable insight into what the reader takes away from the work, and I can see that the main points I was attempting to communicate have come across adequately. I thank Referee 1 for their time and effort in putting this summary together. I note that Referee 1 does not pose any questions in the summary, thus I jump straight to the section highlighting their concerns:
COMMENT 1. “The author remarks that the four sites analysed were chosen because they exhibited ‘no significant differences in planview or profile morphologies’ and this would allow for the newly applied methodology detailed in this paper to be tested with regard to discriminating between them. However, apart from referencing papers to which this is discussed, there are no graphs or data that discuss this lack of variance. It would be useful to see how the planform and profile morphological variance compares to the variation demonstrated in the PCA’s of tip, midsection, and base variance between the four sites.”
RESPONSE 1. The referee is correct in stating that the manuscript lacks any graphical display of shape variance relating to planview or profile outlines, which would further provide support for the argument that cross-sectional shapes better discriminate Early from Late Acheulian handaxes. I am grateful to the referee for bringing this to my attention. I have added a SOM section which provides an assessment of both planview and profile shapes of the handaxe samples, following the methods outlined in the main text. Both PCA and CVA plots for planview and profile outlines confer with previous research that these ‘gross morphological’ perspectives are relatively similar across the Early and Late Acheulian assemblages tested here. The CVA results for planview and profile shapes show lower discriminatory power when compared to the results for cross-sectional shape. I believe this strengthens the argument of the manuscript and provides additional ‘proof of concept’ for the ability of cross-sectional shapes to distinguish Early from Late Acheulian handaxes.
COMMENT 2. “Moreover, the papers (Li et al. 2018; Caruana and Herries 2020) referencing the lack of variation among planview and profile morphologies from the four sites use a different methodology for calculating shape variance to the one used by the author here. The papers referenced do not calculate shape variance in a ‘size-free’ way (Caruana and Herries 2020: 93) by using semi-landmarks and performing Procrustes analysis to normalise allometric effects as is presented in the submitted paper. Li et al. (2018: 32) ‘use measurements and indices developed by Bordes (1961) and Roe (1964, 1968)’ which, although designed to look at shape, are not equivalent to the ‘Fifty evenly-space semilandmarks’ used in this submission to measure the morphology of cross-sections. In reference to my first point, the reasons why different methods are used here need to be addressed and/or explained.”
RESPONSE 2. I thank the reviewer for bringing this issue to my attention. I do agree that the research discussed in the paper, which alludes to a lack of difference in planview and profile shape variance between Early and Late Acheulian handaxes, are based on different methods than those presented in the manuscript. I have added a SOM section that analyses both planview and profile shapes of the sample analyzed in the main text, which corroborates previous findings that both planview and profile shapes overlap between Early from Late Acheulian handaxes. I also have added the following edits to the manuscript:
Pg. 20. In the Results section I have added the following sentence:
“Before the cross-sectional results are discussed, it should be noted that the analyses of planview and profile shapes in the SOM confirm that they have a relatively low discriminatory power for distinguishing Early from Late Acheulian handaxes. This corroborates previous studies suggesting that planview and profile perspectives do not predict the affiliation of Acheulian handaxes from South Africa to either the Early or Late periods of this industry (Li et al. 2018; Caruana and Herries 2020).”
COMMENT 3. “Linked to this issue of shape variance is the repeated conflation of ‘shape variability’ and ‘asymmetry’. Although I suspect that this has not impacted the purpose or results of the study submitted, it is something that should be revised and/or explained as on my first reading of the paper it was confusing as to what was being tested. An example of this is (Page 7, Lines 14 to 15): ‘the Tabun sample displayed a marked increase in asymmetry, i.e. shape variability’. Shape variability is not the same as asymmetry – theoretically, you can have a pointed handaxe with the same index of asymmetry as an ovate handaxe but the two of them have different shapes. Moreover, the author references Shipton and Clarkson (2015) who the author uses (Page 4, Lines 5 to 6) as an example of a paper that is ‘focused on planview symmetry in handaxes’ and which finds that ‘shape variability and reduction intensity are independent variables’. Shipton and Clarkson (2015) do find that shape variability and reduction intensity are independent variables, as they also do with asymmetry when testing handaxes from Broom, however, Shipton and Clarkson (2015) and Shipton et al. (2019), who are both referenced by the author, separate shape/morphological variability and asymmetry and do not use them interchangeably. However valid the author’s comments with regard to shaping processes having little effect on producing symmetrical forms are (Page 4, Lines 7 to 8), using planview shape and planview asymmetry interchangeably should be amended.”
RESPONSE 3. I understand the referees concern here and acknowledge that the concepts of ‘asymmetry’ and ‘shape variability’ should be disentangled. An Acheulian assemblage with numerous different handaxe forms (i.e. cordiform vs. ovate) might be described as ‘high shape variability,’ although if those handaxes are generally symmetrical, the assemblage would be described as ‘low asymmetry’. I appreciate the referee’s recognition that despite the confusion of these terms, the results of this study remain unaffected. To correct this confusion, I have made the following revisions:
Pg. 4. I have specifically mentioned the difference between ‘symmetry/asymmetry’ and ‘shape variability’ on Page 4 in the following sentences:
“The analysis of handaxe shape has been predominantly focuses on issues of symmetry and shape variation, the latter referring to differences in handaxe forms (i.e. pointed vs. ovate) (Iovita and McPherron 2011; Shipton and Clarkson 2015; Shipton et al. 2019). However, relating these factors tospecific manufacturing procedures remains challenging. For example, studies focused on how planview symmetry and shape variability relate to reduction intensity have found these variables to be independent (Shipton and Clarkson 2015; Shipton et al. 2019; García-Medrano et al. 2019). This suggests the shaping processes alone have little effect on producing symmetrical handaxes or differences in plan forms, which instead results from the knapper’s intentions in the placement of flake detachments.”
Pg. 6. I have further referred to differences between ‘symmetry/asymmetry’ and ‘shape variability’ in the following sentence:
“Examining shape variance, both in terms of symmetry and plan form in Acheulian handaxes has provided important insights into the evolution of technological skill and cognition of Early and Middle Pleistocene hominins…”
Pg. 7. I have changed the following sentence to clarify terminologies in the following sentence:
“This is specifically true for planview symmetry, where studies have demonstrated an increase in variability relating to planview outlines in Late Acheulian assemblages when compared to older assemblages…”
Pg. 8. I have removed the phrase “…i.e. shape variability.” From the reference to the Saragusti et al. (2005) study, which now only refers to asymmetry in accordance with the wording used in this study.
Pg. 9. In reference to Li et al.’s (2018) study of symmetry, I have added:
“They segmented 3D scans and compared the volumes of halves to assess symmetry, which showed no significant differences between assemblages.”
Pg. 15. I have clarified the relationship between ‘refinement’ and planview ‘symmetry’ in the following sentence:
“The lack of correlation between shape refinement (i.e. increasing planview symmetry and profile thinness) and time in these assemblages thus provides a means of testing the discriminatory strength of cross-sectional shapes in identifying manufacturing trends that distinguish Early from Late Acheulian handaxes.”
COMMENT 4. “The author uses experimental studies by Callahan (1979) and Shelley (1990) which suggest that certain cross-sectional morphologies are indicative of reduced thinning success and can correlate with signatures of variably skilled action such as step/hinge fractures (overly concave/convex) or mismanagement of flaking surfaces (rhomboidal shapes) that hinder thinning. The author uses these general shapes in conjunction with the knowledge of advancements in thinning procedures highlighted and attributed to the Late Acheulean at other sites (Stout et al. 2014 etc.) to make the prediction that thinner and more elongated ovular shapes are the result of successful thinning and as such should be demonstrated more in Late Acheulean assemblages. I am happy with this initial prediction but the author uses these shapes (reproduced from Callahan) as a high-fidelity universal method for identifying thinning success without corroborating with qualitative observations of whether any of the identifiers of unsuccessful thinning that Callahan (1979) refers to are visible on the handaxes that are more rhomboidal, concave or convex in the author’s dataset. The author concludes on page 24 that a distinguishing factor between handaxe assemblages is ‘knapping mishaps’ and that ‘most of the variance detected in the PCA analyses related to shapes resulting from either poor management of bifacial flaking surfaces or step/hinge fractures’, however, there is no discussion prior to this regarding whether observations of step/hinge fractures are more prevalent amongst those with more rhomboidal, convex or concave cross-sectional morphologies. Those aspects of the handaxes are assumed based on patterns from experimental observations. It would be useful if there was some discussion of this and/or more explicit acknowledgment of the assumption that certain shapes are being used as proxies for said ‘less skilled’ behavirous.”
RESPONSE 4. I thank Referee 1 for this insightful comment, and I do agree that my observations require justification. I respond in two parts below. I first make a clarifying statement for the editor and then respond to the referee’s concern. Before, I would like to point out though that the text contains a variety of explanations, which are repeated throughout the Results section in assessing variance in cross-sectional shapes between samples. For instance, on pg. 22 I state:
“Cave of Hearths is skewed more towards the positive PC1 axis, albeit shows strong variance along the PC2 axis. This further supports the notion that handaxe tips were potentially the focus of reduction processes, whereas midsections and bases were either not as intensively reduced or prone to more knapping mishaps.”
I believe this accounts for a variety of possibilities and that step and hinge fractures are not the sole reason why rhomboidal, convex and concave cross-sectional shape result.
RESPONSE TO EDITOR: I would also like to clarify that the rhomboidal shapes associated with the negative end of PC1 are interpreted here as potentially resulting from the ‘mismanagement’ of bifacial cross-sections. This interpretation is supported by the experimental research referenced in the text (i.e. Shelley 1990). This manuscript is focused on identifying cross-sectional shapes that result from extensive thinning, as well as those that potentially hinder thinning processes. It is assumed that thinning played a role in the production of all handaxes, but the Late Acheulian samples is characterized by a wholesale increase in successful thinning removals. Nonetheless, in the Predictions section I state:
“It is predicted that Early Acheulian assemblage should display more variation towards cross-sectional shapes that are problematic for thinning (Fig. 1B-D). This might result from a lack of thinning in handaxe shaping, differences in technological skill or raw materials.”
As such, I make allowances for more rhomboidal shapes resulting from an overall lack of thinning, technological skill and/or raw material constraints. While the referee does not insinuate that all cross-sectional shapes resulting from potential ‘knapping mishaps’, I would like to clarify that I have not put forth an argument stating that all shapes ‘problematic’ for thinning have resulted from step/hinge fractures. One has to be aware that the shape of Early Acheulian handaxes may have been affected by biomechanical constraints on toolmaking affiliated with Homo erectus/ergaster. The evolution of archaic Homo sapiens (i.e. H. heidelbergensis) may have seen an increase in toolmaking ‘skill’, which was accompanied by the production of Late Acheulian handaxes. I have stated this possibility in my Conclusions section.
RESPONSE TO REFEREE: In the SOM I have provided an additional assessment of the abundance of step and hinge fractures within the samples assessed here. I have calculated the surface area of step and hinge fractures on 3D meshes of handaxes in Geomagic Warp. These values were log transformed to normalize variance between assemblages, and to assess the relationship between flaws and cross-sectional shapes, I have included a correlation table (Kendall’s tau-b) comparing the surface area of flaws to principal component scores. Similar to SDI comparisons, PC1 and flaw rates show significant, yet weak correlation, suggesting a relationship between flaws and cross-sectional shapes. No correlation between flaws and PC2 was detected, again similar to SDI comparisons. However, I suspect that the lack of correlation between PC2 and SDI and flaw ratios is not a reality, but rather a product of how PC2 scores are calculated. The major access of variation in the PCAs is PC1 (unsurprisingly), while PC2 scores pick up less than 20% of shape variation across the board. Thus, I suspect the problem with correlations is due to the little amount of shape variance that PC2 scores actually captures. This is an interesting issue onto itself and I will perhaps look into this in future.
On a personal note, I was planning on using a similar analysis, based on flaw rates on handaxe surface, in an upcoming paper focused on handaxes of Amanzi Springs, which will expand on this concept. I have include an analysis of knapping flaws here to present a quantitative assessment of the abundance of step and hinge fractures represented within each sample and how this potentially relates to the outcomes of cross-sectional shape variation.
I have discussed this assessment as a way of providing some additional insight into the potential effects of flaw rates on cross-sectional shape variance, albeit this is not a conclusive analysis. In the main text I have made the following corrections to tone down the importance of step and hinge fractures in the interpretations of the geometric morphometric results:
Pg. 20. I have added a phrase to the interpretations of the PCAs, which broadens the possibilities of why shapes along the PC2 access may have resulted. This sentence now reads which now reads:
“PC2 axes represented shapes that were either overly convex or concave, which likely resulted from raw material constraints, a lack of thinning, or knapping mishaps including poor management of surface convexity and/or step fractures.”
Pg. 21. I have added the word “possibly” to the following sentence, which softens the association between convex/concave cross sections and step/hinge fractures:
“The Rietputs 15 sample is also spread across the PC2 axis, suggesting that some tip shapes are overly concave/convex or possibly hindered by step fractures.”
Pg. 23. I have removed the following sentence:
“This may be due to the fact that PC2 shapes relate to significant step fracturing that potentially halted reduction processes.”
Pg. 26. To tone down the importance of knapping mishaps, in the Conclusions section I add the phrase, “…a lack of thinning, and potentially the…” to the following sentence:
“Most of the variance detected in the PCA analyses related to shapes resulting from either a lack of thinning, and potentially the poor management of bifacial flaking surfaces or step/hinge fractures.”
COMMENT 5. “The author’s conclusion that assessments of cross-sectional shape can distinguish handaxes from different periods of Acheulian industry using the broad distinction that Late Acheulean handaxes are thinner, more elongated, and ovular than Early Acheulean handaxes is somewhat confused by the results from Amanzi Springs. The Amanzi springs shape variance overlaps Early and Late Acheulean patterns and is explained by the author through analysis performed in a separate paper which suggests handaxes from Amanzi Springs were ‘abandoned before reaching the final stages of manufacturing due to production flaws’ (Page 23, Lines 1 to 2). Therefore, on the one hand, it serves to bolster the author’s argument for using observations from Callahan (1979) to distinguish between handaxes with more or fewer flaws related to greater/lesser success of thinning but, on the other hand, suggests that using cross-sectional morphological analysis to discriminate between Early and Late Acheulean handaxes is improved through qualitative observation and cannot be used universally with great confidence without such observations. This will be made clearer when the author increases the sample size of their study but this may be something to comment on (Page 24, lines 1 to 2) when the author notes that the ‘results of the study remain preliminary and require further testing with expanded sample sizes’. However, the author (Page 21, Line 10) states that cross-sectional shapes ‘strongly discriminate handaxes from Early and Late Acheulean assemblages’, something that arguably needs to be tempered given the importance of qualitative observations of Amanzi Springs to the explanation of the morphological variance it exhibits. The point regarding the strength of the discriminatory power of cross-sectional morphological analysis would be made more powerfully if, as suggested above, there were graphs/data outlining the comparative planview shape variability.”
RESPONSE 5. I am grateful to Referee 1 for their insightful assessment of this manuscript and I do agree that the ‘strength’ of cross-sectional shapes to discriminate Early from Late Acheulian handaxes is equally related to the ‘weakness’ of planview and profile shapes to do so. As stated above in RESPONSE 2, I have created a SOM file that contains an analysis of both planview and profile shapes using the same methods outlined in this study. The results of comparing planview and profile outlines confirms that cross-sectional shapes indeed have stronger discriminatory power in a CVA. The amendments to the main text outlined above also support these results. While I would have liked to expand on the comparisons of the discriminatory power between planview/profile and cross-sectional shapes, I am already over the recommended word count for the journal. Thus, I have provided some discussion in the SOM.
Referee 1 Minor Concerns:
COMMENT 1. “Abstract exceeds word-limit. It is currently 118 words long.”
RESPONSE 1. I have shortened the Abstract section to 100 words.
COMMENT 2. “I would recommend that the term ‘standardization’, used several times, be explained more. The author notes in the abstract that ‘standardization of planview and profile morphologies has been used to distinguish Early and Late Acheulian assemblages’ but explains in the introduction that some research has shown that ‘increasing standardization of handaxe forms is not a ubiquitous trend characterizing the Acheulian’. The author links it with the notions of refinement, however, it would be helpful if there was a more detailed outline of what is meant by ‘standardization’ with regard to the discussion of changes in Acheulean handaxes over time.”
RESPONSE 2. I appreciate the referee’s attention to terminologies, which is very important for clarity in the manuscript. I have removed the term ‘stardardization’ from the Abstract section. Attempting to align definitions of ‘standardization’ with ‘refinement’ I have also added the following phrase, “(i.e. increased planview symmetry and thin profiles)” to the following sentence on pg. 3:
“However, focused investigations testing the correlation between handaxe refinement and time have yielded mixed results, suggesting that increasing standardization of handaxes forms (i.e. increased planview symmetry and thin profiles) is not a ubiquitous trend characterizing the Acheulian (see McNabb and Cole 2015).
I note that the word ‘standardization’ is now only used three times in the manuscript. I have left this word in the section discussing Iovita and McPherron’s (2011) research as they use this term in a fairly ambiguous manner.
COMMENT 3. “Page 5, paragraph starting on Line 12 – when introducing the methods used regarding capturing cross-sections from handaxe tips, midsections, and bases, as well as plotting semi-landmarks, it would be useful if there was a reference to a figure depicting this. Moreover, an earlier explanation regarding the rationale of where measurements are taken from would be welcome – the first mention of what constitutes tip, midsection, and base comes on Page 16 in the ‘Outline extraction’ section. This is easily resolved by including what tip, midsection, and base refer to when the author first mentions ‘Cross-sections from handaxe tips, midsections and bases were captured from digital thin-sectioning’ (on Page 5, from Line 12). Just add in brackets.”
RESPONSE 3. I have added the phrase, “…(1/4, 1/2, and 3/4 of maximal length, respectively)…” to the sentence identified by the referee. I have added a short section in the SOM that provides a figure of the digital sectioning and GM processes. I would note that the definition of tip, midsection and base proportions of handaxes are well defined within the works cited in the following sentence on pg. 16, and I feel referencing these studies provides a means of referencing why I chose to measure tips, midsections and bases at 1/4, 1/2, and 3/4 of maximal length, respectively.
“Scan meshes were then imported into CloudCompare and digitally thin sectioned in tip (1/4 of maximal length), midsection (1/2 of maximal length) and base (3/4 of maximal length) regions corresponding to standard metric measurement procedures for Acheulian large cutting tools (i.e. handaxes and cleavers) (Bordes 1961; Roe 1968; Crompton and Gowlett 1993; García-Medrano et al. 2020).”
COMMENT 4. “Page 8, last sentence of the first paragraph from Lines 8-10 – make clearer the fact that the ‘maxim’ you refer to is one that relates to change over time.”
RESPONSE 5. I have added the phrase, “…over time…” to this sentence.
COMMENT 5. Page 16 – ‘Outline extraction’ and ‘Geometric morphometric methods’ sections. To repeat, key aspects of process and methodology could be demonstrated in a Figure. I understand that there is a limit of six figures, but it would be useful visually to show where measurements were taken from and how semi-landmarks were spaced around each cross-sectional profile.
RESPONSE 5. I have added a figure to the SOM.
COMMENT 6. Hypotheses would be more clearly presented by using bullet points. It would allow the reader to easily find them again in the text.
RESPONSE 6. I appreciate the referee’s concern here and I do agree that bullet points would perhaps make hypotheses easy to find, but I feel that the core hypotheses of the manuscript are succinctly contained within the Predictions section.
COMMENT 7. In the ‘Discussion’ section on page 22 (lines 13-15), the author notes that one of their own recently joint-authored papers characterizes handaxes from Amanzi Springs by their ‘high morphological variance’. This goes counter to their reasoning for selecting the four sites to test the discriminatory power of cross-sectional morphological variance because ‘no significant differences in planview or profile morphologies have been previously detected’ in the four sites chosen. The author does say that this specifically relates to variance in width, thickness, and mass, however, it would be useful if this was made clearer that they are not talking about shape variance. The word ‘morphology’ should be amended perhaps.
RESPONSE 7. I appreciate the referee’s concern here, but the reasoning for the sample selection is based on the fact that ‘shape does not predict affiliation to the Early or Late Acheulian period’. See pg. 15, which states:
“This sample was chosen because of recently studies showing that the shape of handaxes from Rietputs 15, Cave of Hearths and Amanzi Springs does not predict affiliation to specific periods of the Acheulian (Li et al. 2018; Caruana and Herries 2020). Furthermore, morphological variability in planview and profile shape has been shown to increase the Late Acheulian assemblages. This is specifically highlighted in the study focused on Amanzi Springs, which demonstrated significant differences in geometric proportions and allometric trends when compared to Rietputs 15 and Cave of Hearths (Caruana and Herries 2020).”
COMMENT 8. The difference between ‘overly concave’ and ‘overly convex’ cross-sections as demonstrated in Figure 1 could be clearer. I understand there are dashed lines that aim to demonstrate the relative concavities/convexities, but it could be pointed out more clearly.
RESPONSE 8. I do appreciate the referee’s concern, and I have added a clarifying sentence on Pg. 13, which I hope will improve definitions. I have also added some detail in Figure 1.
Referee 1 Correction suggestions:
I am very grateful for Referee 1 in pointing out these extensive mistakes. I do apologize for so many spelling and grammatical errors.
- Page 1, Line 7 of Abstract – change ‘preliminarily’ to ‘preliminary’
RESPONSE: I have removed this word from the abstract.
- Page 2, Line 14 – add an ‘s’ to the end of ‘connotation’
RESPONSE: Done.
- Page 4, Line 3 – change ‘focuses’ to ‘focused’
RESPONSE: Done.
- Page 5, Line 9 – change ‘planivew’ to ‘planview’
RESPONSE: Done.
- Page 5, Line 10 – remove ‘the’ before ‘distinguishing’
RESPONSE: Done.
- Page 7, Line 3 – the word ‘found’ is used twice in one sentence. Change one of them.
RESPONSE: I have changed “found” to “concluded”.
- Page 8, Line 3 – change ‘then’ to ‘than’
RESPONSE: Done.
- Page 8, Line 14 – change ‘compare’ to ‘compared’
RESPONSE: Done.
- Page 8, Line 16 – change ‘plaview’ to ‘planview’
RESPONSE: Done.
- Page 9, Line 10 – with regard to final thinning phases, check whether they mean ‘effected’ in relation to their impact on the final shape of handaxes. Do they mean that thinning phases brought about the final shape, or that they had a significant impact? It doesn’t read too well. I suggest changing to ‘affected’.
RESPONSE: I have changed this the word to ‘affected’.
- Page 9, Line 14 – change ‘distinguish’ to ‘distinguishes’
RESPONSE: Done.
- Page 10, Lines 1 to 3 – first sentence reads oddly. I would remove the word ‘that’ on Line 1.
RESPONSE: Done.
- Page 10, Line 4 – add ‘the’ before ‘midsection’ and add ‘an’ before ‘effort’
RESPONSE: Done.
- Page 10, Line 9 – change ‘were’ to ‘where’
RESPONSE: Done.
- Page 11, Line 8 – change ‘were’ to ‘where’
RESPONSE: Done.
- Page 11, Line 11 – change ‘important’ to ‘importance’
RESPONSE: Done.
- Page 12, Line 4 – add an ‘s’ to the end of ‘perspective’
RESPONSE: Done.
- Page 12, Line 7 – add ‘of’ before ‘cross-sectional’
RESPONSE: Done.
- Page 13, Line 4 – remove “ ‘s “ from “Callahan’s”
RESPONSE: Done.
- Page 14, Line 4 – add ‘s’ to the end of ‘Hearth’
RESPONSE: Done.
- Page 15, Line 2 – end of sentence doesn’t read well. Seems to be missing a word like ‘insights’ after the word ‘useful’
RESPONSE: I have modified this sentence accordingly.
- Page 15, Line 3 – change ‘recently’ to ‘recent’
RESPONSE: Done.
- Page 15, Line 6 – change ‘the’ to ‘in’
RESPONSE: Done.
- Page 18, Line 2 – repetition of ‘are discussed’ – remove.
RESPONSE: Done.
- Page 18, Line 7 – change ‘The’ to ‘This’
RESPONSE: Done.
- Page 18, Line 9 – change ‘planivew’ to ‘planview’
RESPONSE: Done.
- Page 19, Line 11 – repetition of ‘patterns of’ – remove
RESPONSE: Done.
- Page 19, Line 13 – add ‘is’ before ‘loaded’
RESPONSE: Done.
- Page 20, Line 9 – change ‘lesser of an’ to ‘a lesser’
RESPONSE: Done.
- Page 21, Line 17 – change ‘Figure 6’ to ‘Figure 6A’ to reflect that Figure 6 is made up of one boxplot and three regression plots.
RESPONSE: Done.
- Page 22, Line 4 – author refers to ‘regression graphs presented in Figure 3’ – this needs to be changed to reflect that they are actually in Figure 6.
RESPONSE: Done.
- Page 25, Line 13 – change ‘the demonstrates’ to ‘that demonstrate’
RESPONSE: Done.
- Page 25, Line 15 – add ‘of’ after ‘most’
RESPONSE: Done.
- Page 26, Line 1 – at the end of the sentence that began on the previous page, there is ‘(refs).’ – this needs to be amended to reflect the references the author intends to cite here.
RESPONSE: Done.
III. Referee 2 Major Concerns:
I am grateful to Referee 2 for their review of my work and thank them for their time and effort. I am also very thankful for their supportive comments and recommendation to publish my manuscript following their recommended revisions. Below I respond to their comments in numerical order.
COMMENT 1. In spite of the structure is adeccuate, I suggest to devide the proposed “Materials and Method” into two sections: Materials (including general info about sites and assemblages) and Methods.
RESPONSE 1. I thank Referee 2 for this suggestion and have separated ‘Materials’ and ‘Methods’ sections.
COMMENT 2. The method is innovative and appropriated. Nevertheless, the morphometry applies to the lithic technology requires a clear technological characterization of the instruments. This is the only way to understand the technological process beyond the shape of tools. In this case, the author only assume the number of scars as the sole variable which characterize the type of shaping. So, it is mandatory to explain: the type of blank used (cobble or flake), the % of cortex and its localization, the removal series, the presence or absence of final retouch or the edge angle (average or more than one angle per edge, at least one per each part). If possible, characterize each one of the 3 parts considered, to understand the differences between them. The method of analysis must be recorded in the Method section, together with GM.
RESPONSE 2: I appreciate the referee’s concern here and they are correct in stating that description of the technological aspects of handaxes analyzed here is critical for understanding how the outcomes of this analysis relate to handaxe production at these sites. However, I argue that any focus on individual assemblages is beyond the scope of this study and the limited word count prevents me from describing all four assemblages in such detail. Ultimately, this paper aims to describe and test a novel GM method for examining the shape of handaxes, with the aim of testing the discriminatory power of cross-sectional shapes to distinguish Early from Late Acheulian handaxes. As such, comparing the results of the GM methods outlined in the text and the descriptive/metric data of the assemblages shifts the focus of this study to understanding handaxe production from the mentioned sites.
I am limited to 8000 words and have already exceed this limit to account for changes I consider imperative to describing and supporting this new GM method. Because this is a novel means of examining handaxe shape, I have opted to focus on providing as much detail as possible to describe the process of data collection and outcomes. Further, the technological descriptions mentioned in the referee’s comment above have already been published in other works: Amanzi Springs (Deacon 1970; Caruana and Herries 2020); Cave of Hearth (McNabb 2009; Li et al. 2018); Rieptputs 15 (Kuman and Gibbon 2017; Li et al. 2018); and Sterkfontein (Kuman and Clarke 2000; Kuman and Gibbon 2017). I have specifically stated this in the Materials section, and I would rather refer readers to these studies, which focus on these assemblages in detail.
While I would like to include all the details of the assemblages in the main text as suggested by the referee, in reality, this would result in a separate, stand-alone manuscript with a different focus and goal. I would like to mention that I am currently working on two manuscripts that apply this method to assessing LCTs from Amanzi Springs, Cave of Hearth and Rietputs 15, which will use the assessment of cross-sectional shape to compare with metric data form these sites to examine the production of these tools on a more refined level. I wanted to submit this manuscript to ensure that the method first passes a peer-review process to validate its use and conclusions.
However, I have added two tables in a SOM file, one displaying metric data (16 variables), and the other displaying a summary of edge angles for tip, midsection and base measured from cross sections used in this study. These descriptive statistical data will at least allow readers to compare variables with the results of the GM analysis.COMMENT 3. By other side, it is required to characterize metrically the samples (mean, CV and SD), elongation and refinement indexes and if possible, volume information.
RESPONSE 3: As outlined in the response above, I appreciate the referees concern and agree that this information is very important for the reader’s assessment of individual assemblages. I have added a SOM with some metric data for these assemblages.
COMMENT 4. Please, include at least a Figure with instruments to illustrate the assumptions/conclusions of the research done: oval vs. triangular shapes, presence or absence of thinning works, etc,
RESPONSE 4: Due to the Covid-19 pandemic, South Africa has been under a mandatory lockdown, which restricts travel, and subsequently access to universities. As such, I cannot access the handaxes themselves for high-resolution photography. I do note that high quality figures have been produced and published in the studies I refer readers to for the individual sites, which provides detail on LCT shaping processes. In Figure 1, I use cross-sections from 3D scans of handaxes. In the SOM, I also provide a figure of the data acquisition process in the SOM, which presents an example of a cross-sectional outline derived from a handaxe scan.
COMMENT 5. Page 20. Figure 1 call. Please, include Callahan 1979 reference. Nevertheless, the differences between shapes C and D ar not clear. A good solution could be to include a real example per each shape (A, B, C and D).
RESPONSE 5: I have added the Callahan reference to Pg. 20, where Figure 1 is referenced. I do note that in Figure 1, I use examples of cross-sectional scans included in this study (right-hand column), which is explained in the figure caption. I have also added a sentence on Pg. 13 to clarify the definition of overly concave and convex surfaces, as well as detail to Figure 1.
COMMENT 6. Page 20. Line 38. “pattern” word is duplicated
RESPONSE 6: I have deleted the second use of this word.
COMMENT 7. Page 21. Line 21. Please, include here that Fig.3 refers to mid sections
RESPONSE 7: I have added in the phrase “…in midsections…” to this sentence.
COMMENT 8. Page 23. Line 16. The values of r2 are low to consider a clear relation between reduction intensity and manufacturing processes. But in case of a positive relation this occurs mainly on tips (r2=0.213), as you declare on page 21, lines 35-43.
RESPONSE 8: I appreciate the referee’s comment here and I do acknowledge the weak relationship of the regression figures. However, the p-values support that SDI and PC scores are significantly related, and as such are dependent variables.
COMMENT 9. Please, at the end of the Result section, take into account that SDI could be affected by several aspects of the knapping process. So, it is strongly recommended to define the relation between SDI and other variables, such as volume information, to get a deeper understanding of how the reduction intensity affects shape. Besides, I suggest to contrast this PCA results with any metrical information (e.g. width and thickness on tips variation). This will be very useful to, for example, reinforce the differences between the shaping strategy on tips and the rest of tools, or even between the different assemblages.
RESPONSE 9: I appreciate the referee’s comment and agree that comparisons between the SDI and other variables, and also considering how the GM results might relate to other variables of reduction is highly useful information. However, I argue that this is beyond the scope if this study. This manuscript presents a novel method for analyzing handaxe shapes questioning how the GM analysis discriminates Early from Late Acheulian handaxes. The word count of this paper has to be devoted to outlining this method and describing its outcomes. As such, information on the assemblages themselves, and the influence of thinning on the overall shape of handaxes from these sites is the subject of separate studies. Please note, I am currently working on manuscripts using this method to assess the individual assemblages in a more focused level. While I am very sympathetic to the referee’s interests in what the results say about shaping processes at these sites, this study simply aims to introduce a novel method for detecting the influence of thinning processes in handaxes via GM techniques. The goal of this study is to test the discriminatory power of cross-sectional shapes to distinguish handaxes from differing time periods. My interpretations also suggest that this method can pick up on knapping mishaps that might affect handaxe shape, but further testing is needed to confirm this hypothesis.
I am also aware that the SDI is affected by other variables and can be confounded by a lack of estimating the original blank size. In truth, this index compares reduction intensity at the point of discard, rather than measures reduction throughout the life-history of the tools. I simply use the SDI as an independent measure to relate cross-sectional shapes to reduction processes. Although the correlations are weak, the p-value results confirm that variance in cross-sectional shapes is in part related to reduction intensity.COMMENT 10. Page 24. Line 40. Please, remember that there is no any objective information about how the edges are, e.g. angle distribution along the edge or specific angle pattern, such us on tips to compare between the different sites.
RESPONSE 10: I thank the referee for this comment, and I have added a table in the SOM that provides descriptive statistical data on edge angles calculated from the cross-section scans used in this study.
COMMENT 11. The conclusion section must record all the results of this work, not only general assumptions. The author must explain which are the contributions that cross-section analysis represent to the study of technological evolution from Early to Late Acheulean.
RESPONSE 11: I thank the reviewer for their comment and appreciate the need to expand the discussion on how the methods used here can contribute to a wider evolutionary perspective. However, the last paragraph of the conclusion does attempt to touch on this subject. I state that:
“Lastly, insights from Shelley’s (1990) study relating cross-sectional shapes in handaxes to knapping skill provide perspective on tracing potential differences in motor-cognitive development across the Early to Late Acheulian boundary. If the complexity and extent of handaxe thinning is characteristic of Late Acheulian toolmaking, corresponding to hierarchical actions and advanced technological skill, cross-sectional shape variance may provide further insight into the effects of these phenomena on handaxe morphologies.”
I have added some text to the end of this section to further clarify the role of cross-sectional shape analysis in understanding evolutionary trends on pg. 27:
“As mentioned above, archaeological analysis of Late Acheulian handaxes suggest that the organization of thinning procedures potentially distinguish them from earlier forms (Shipton et al. 2013; Stout et al. 2014). The analysis of cross-sectional shapes in handaxes provides a method of assessing the extent of thinning procedures within and between assemblages. Further development of this method can potentially provide insight into the extent of thinning on both synchronic and diachronic scales, which can directly test the hypothesis that the Late Acheulian handaxes are characterised by extensive thinning processes.”
Author response by
Cite this author response
- pre-publication peer review (ROUND 1)
Decision Letter
2020/08/2424-Aug-2020
Dear Dr. Caruana:
Manuscript ID ARCH-06-0288 entitled "A Pilot Study Comparing the Effects of Thinning Processes on the Cross-Sectional Morphologies of Early from Late Acheulian Handaxes" which you submitted to Archaeometry, has been reviewed. The comments of the referee(s) are included at the bottom of this letter.
The referee(s) have recommended major revisions to your manuscript. Therefore, I invite you to respond to the referee(s)' comments and revise your manuscript.
If you would like help with English language editing, or other article preparation support, Wiley Editing Services offers expert help with English Language Editing, as well as translation, manuscript formatting, and figure formatting at www.wileyauthors.com/eeo/preparation. You can also check out our resources for Preparing Your Article for general guidance about writing and preparing your manuscript at www.wileyauthors.com/eeo/prepresources.
To revise your manuscript, log into https://mc.manuscriptcentral.com/arch and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions." Under "Actions," click on "Create a Revision." Your manuscript number has been appended to denote a revision.
You will be unable to make your revisions on the originally submitted version of the manuscript. Instead, revise your manuscript using a word processing programmeme and save it on your computer. Please also highlight the changes to your manuscript within the document by using the track changes mode in MS Word or by using bold or coloured text.
Once the revised manuscript is prepared, you can upload it and submit it through your Author Centre.
When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided. You can use this space to document any changes you make to the original manuscript. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response to the referee(s).
IMPORTANT: Your original files are available to you when you upload your revised manuscript. Please delete any redundant files before completing the submission.
Because we are trying to facilitate timely publication of manuscripts submitted to Archaeometry, your revised manuscript should be uploaded as soon as possible. If it is not possible for you to submit your revision in a reasonable amount of time, we may have to consider your paper as a new submission.
Once again, thank you for submitting your manuscript to Archaeometry and I look forward to receiving your revision.
Yours sincerely,
Prof. Mark Pollard
Managing Editor, Archaeometry
mark.pollard@rlaha.ox.ac.ukManaging Editor's Comments to Author:
Managing Editor: Pollard, Mark
The comments are rather long but basically encouraging. Isuggest you do what is reasonable, and bear in mind that you can use online supplemental material for additional figures, as requested by one of the reviewers,Referee(s)' Comments to Author:
Referee: 1
Comments to the Author
SummaryThis pilot study represents a valid attempt to outline a new method for quantitatively distinguishing Early Acheulean handaxes from Late Acheulean handaxes by looking at the variance in their cross-sectional morphologies. Principal components analysis and canonical variate analysis of cross-sectional morphological variance of tip, midsection, and base of handaxes from two Early Acheulean sites (Rietputs 15 and Sterkfontein Member 5 West) and two Late Acheulean sites (Cave of Hearths and Amanzi Springs), as well as regression analysis of cross-sectional shape and reduction intensity, brings the author to conclude that later Acheulean handaxes are characterised by increased thinning as a result of advancements in flaking mechanics and thinning procedures as has been suggested by previous research. Therefore, the methodology presented is argued by the author to be successful in discriminating between handaxe assemblages of different periods.
The author initially discusses how planview and profile morphologies of handaxes have been predominantly used in attempts to distinguish Early from Late Acheulean handaxes, with the general hypothesis being that handaxes become more refined (characterised by an increase in symmetry and thinness) over time. However, the author rightly references research which suggests that this trend towards symmetry and refinement is not a universal one (McNabb and Cole 2015), that biface shape does not correlate with thinning reduction intensity (Shipton and Clarkson 2015) and that analysis of profile morphologies only goes so far in providing insight regarding the success of thinning procedures. The author discusses research (Stout et al. 2014; Shipton 2018; Garcia-Medrano et al. 2019) that identifies advancements in flaking mechanics and thinning procedures as broadly characteristic of Late Acheulean handaxes and focuses on this aspect (thinning procedures) to test whether the extent of thinning can reliably differentiate Early from Late Acheulean handaxes.
The method proposed involves 3D-scanning handaxes from four South African handaxe assemblages (Rietputs 15, Sterkfontein Member 5 West, Cave of Hearths, and Amanzi Springs) and investigating 2D cross-sectional morphological variance at the tip, midsection, and base. This methodology is influenced by experimental research by Callahan (1979) which suggests that cross-sectional morphologies can highlight the success of thinning procedures, as well as Shelley (1990) who suggests cross-sectional morphologies can relate to degrees of knapping skill. These experimental insights, as well as previous research (Li et al. 2018; Caruana and Herries 2020) reporting no significant differences in planview or profile variation from the four sites above, constitute the main rationale for the adoption of cross-sectional morphological analysis in this pilot study. The author uses cross-sectional shape variance analysis to test its ‘discriminatory power’, with the assumption being that experimental understanding of what delineates successful, skilled thinning will be mirrored in the difference between Early Acheulean handaxes and Late Acheulean handaxes in this pilot’s sample of sites.
The cross-sectional shape of tip, midsection, and base are captured through semi-landmarks that provide the outline morphology of each cross-section. The author normalised the shapes via a Procrustes analysis to remove size influences when analysing morphological variation via Principal Components Analysis. Using insights from Callahan’s (1979) experimental work, the author analysed deformation grids of shape variance to identify cross-sectional shapes that demonstrate successful thinning or flaws. Then, using the principal components captured, the author performs a canonical variate analysis in order to test the power of this method to discriminate between handaxe assemblages. The author further tests the ability of cross-sectional morphological variance to discriminate between Early and Late Acheulean assemblages by modifying Shipton and Clarkson’s (2015) correlation of planview shape morphology and Scar Density Index (which tests the ratio of flake scars to surface area and is used as a proxy for reduction intensity) to correlating Scar Density Index with cross-sectional morphologies – the assumption being that cross-sectional shapes indicating increased thinning will correlate with Scar Density Index where planview shape morphology did not.
The author finds that cross-sectional morphologies of handaxe tips are more reliable than midsections and bases at discriminating between handaxe assemblages where ‘no significant differences in planview or profile morphologies have been previously detected’. However, CVA analysis was able to confidently attribute a significant amount of handaxes to the correct assemblage based on a combination of all cross-sectional morphologies (tip, midsection, and base). Additionally, regression analysis of principal components finds that only PC1 correlates in a statistically significant manner with Scar Density Index but only in a weak correlation. The author argues that the significant correlation of PC1 scores and SDI values is demonstrative of cross-sectional shape variance as being related, in some part, to reduction/thinning intensity. Although the regression analysis is left somewhat separate from the PCA and CVA analysis, the author persuasively argues for the use of cross-sectional morphologies as a method for distinguishing between assemblages and proposes that it should be tested further with larger sample size.
This is a thought-provoking proof of concept paper which will equally interest researchers focusing on the Acheulean in South Africa as it will interest those researchers focusing on the Acheulean in the rest of Africa, Europe, and elsewhere. The methodology outlined by the author has the potential to influence other researchers to replicate it towards their own research interests. However, there are a number of assumptions and areas that are not adequately explained or sufficiently demonstrated. With revisions, this paper is worthy of publication.
The main issues I have with this paper are:
- The author remarks that the four sites analysed were chosen because they exhibited ‘no significant differences in planview or profile morphologies’ and this would allow for the newly applied methodology detailed in this paper to be tested with regard to discriminating between them. However, apart from referencing papers to which this is discussed, there are no graphs or data that discuss this lack of variance. It would be useful to see how the planform and profile morphological variance compares to the variation demonstrated in the PCA’s of tip, midsection, and base variance between the four sites.
- Moreover, the papers (Li et al. 2018; Caruana and Herries 2020) referencing the lack of variation among planview and profile morphologies from the four sites use a different methodology for calculating shape variance to the one used by the author here. The papers referenced do not calculate shape variance in a ‘size-free’ way (Caruana and Herries 2020: 93) by using semi-landmarks and performing Procrustes analysis to normalise allometric effects as is presented in the submitted paper. Li et al. (2018: 32) ‘use measurements and indices developed by Bordes (1961) and Roe (1964, 1968)’ which, although designed to look at shape, are not equivalent to the ‘Fifty evenly-space semilandmarks’ used in this submission to measure the morphology of cross-sections. In reference to my first point, the reasons why different methods are used here need to be addressed and/or explained.
- Linked to this issue of shape variance is the repeated conflation of ‘shape variability’ and ‘asymmetry’. Although I suspect that this has not impacted the purpose or results of the study submitted, it is something that should be revised and/or explained as on my first reading of the paper it was confusing as to what was being tested. An example of this is (Page 7, Lines 14 to 15): ‘the Tabun sample displayed a marked increase in asymmetry, i.e. shape variability’. Shape variability is not the same as asymmetry – theoretically, you can have a pointed handaxe with the same index of asymmetry as an ovate handaxe but the two of them have different shapes. Moreover, the author references Shipton and Clarkson (2015) who the author uses (Page 4, Lines 5 to 6) as an example of a paper that is ‘focused on planview symmetry in handaxes’ and which finds that ‘shape variability and reduction intensity are independent variables’. Shipton and Clarkson (2015) do find that shape variability and reduction intensity are independent variables, as they also do with asymmetry when testing handaxes from Broom, however, Shipton and Clarkson (2015) and Shipton et al. (2019), who are both referenced by the author, separate shape/morphological variability and asymmetry and do not use them interchangeably. However valid the author’s comments with regard to shaping processes having little effect on producing symmetrical forms are (Page 4, Lines 7 to 8), using planview shape and planview asymmetry interchangeably should be amended.
- The author uses experimental studies by Callahan (1979) and Shelley (1990) which suggest that certain cross-sectional morphologies are indicative of reduced thinning success and can correlate with signatures of variably skilled action such as step/hinge fractures (overly concave/convex) or mismanagement of flaking surfaces (rhomboidal shapes) that hinder thinning. The author uses these general shapes in conjunction with the knowledge of advancements in thinning procedures highlighted and attributed to the Late Acheulean at other sites (Stout et al. 2014 etc.) to make the prediction that thinner and more elongated ovular shapes are the result of successful thinning and as such should be demonstrated more in Late Acheulean assemblages. I am happy with this initial prediction but the author uses these shapes (reproduced from Callahan) as a high-fidelity universal method for identifying thinning success without corroborating with qualitative observations of whether any of the identifiers of unsuccessful thinning that Callahan (1979) refers to are visible on the handaxes that are more rhomboidal, concave or convex in the author’s dataset. The author concludes on page 24 that a distinguishing factor between handaxe assemblages is ‘knapping mishaps’ and that ‘most of the variance detected in the PCA analyses related to shapes resulting from either poor management of bifacial flaking surfaces or step/hinge fractures’, however, there is no discussion prior to this regarding whether observations of step/hinge fractures are more prevalent amongst those with more rhomboidal, convex or concave cross-sectional morphologies. Those aspects of the handaxes are assumed based on patterns from experimental observations. It would be useful if there was some discussion of this and/or more explicit acknowledgment of the assumption that certain shapes are being used as proxies for said ‘less skilled’ behaviours.
- The author’s conclusion that assessments of cross-sectional shape can distinguish handaxes from different periods of Acheulian industry using the broad distinction that Late Acheulean handaxes are thinner, more elongated, and ovular than Early Acheulean handaxes is somewhat confused by the results from Amanzi Springs. The Amanzi springs shape variance overlaps Early and Late Acheulean patterns and is explained by the author through analysis performed in a separate paper which suggests handaxes from Amanzi Springs were ‘abandoned before reaching the final stages of manufacturing due to production flaws’ (Page 23, Lines 1 to 2). Therefore, on the one hand, it serves to bolster the author’s argument for using observations from Callahan (1979) to distinguish between handaxes with more or fewer flaws related to greater/lesser success of thinning but, on the other hand, suggests that using cross-sectional morphological analysis to discriminate between Early and Late Acheulean handaxes is improved through qualitative observation and cannot be used universally with great confidence without such observations. This will be made clearer when the author increases the sample size of their study but this may be something to comment on (Page 24, lines 1 to 2) when the author notes that the ‘results of the study remain preliminary and require further testing with expanded sample sizes’. However, the author (Page 21, Line 10) states that cross-sectional shapes ‘strongly discriminate handaxes from Early and Late Acheulean assemblages’, something that arguably needs to be tempered given the importance of qualitative observations of Amanzi Springs to the explanation of the morphological variance it exhibits. The point regarding the strength of the discriminatory power of cross-sectional morphological analysis would be made more powerfully if, as suggested above, there were graphs/data outlining the comparative planview shape variability.
Minor Issues
- Abstract exceeds word-limit. It is currently 118 words long.
- I would recommend that the term ‘standardization’, used several times, be explained more. The author notes in the abstract that ‘standardization of planview and profile morphologies has been used to distinguish Early and Late Acheulian assemblages’ but explains in the introduction that some research has shown that ‘increasing standardization of handaxe forms is not a ubiquitous trend characterizing the Acheulian’. The author links it with the notions of refinement, however, it would be helpful if there was a more detailed outline of what is meant by ‘standardization’ with regard to the discussion of changes in Acheulean handaxes over time.
- Page 5, paragraph starting on Line 12 – when introducing the methods used regarding capturing cross-sections from handaxe tips, midsections, and bases, as well as plotting semi-landmarks, it would be useful if there was a reference to a figure depicting this. Moreover, an earlier explanation regarding the rationale of where measurements are taken from would be welcome – the first mention of what constitutes tip, midsection, and base comes on Page 16 in the ‘Outline extraction’ section. This is easily resolved by including what tip, midsection, and base refer to when the author first mentions ‘Cross-sections from handaxe tips, midsections and bases were captured from digital thin-sectioning’ (on Page 5, from Line 12). Just add in brackets.
- Page 8, last sentence of the first paragraph from Lines 8-10 – make clearer the fact that the ‘maxim’ you refer to is one that relates to change over time.
- Page 16 – ‘Outline extraction’ and ‘Geometric morphometric methods’ sections. To repeat, key aspects of process and methodology could be demonstrated in a Figure. I understand that there is a limit of six Figures but it would be useful visually to show where measurements were taken from and how semi-landmarks were spaced around each cross-sectional profile.
- Hypotheses would be more clearly presented by using bullet points. It would allow the reader to easily find them again in the text.
- In the ‘Discussion’ section on page 22 (lines 13-15), the author notes that one of their own recently joint-authored papers characterises handaxes from Amanzi Springs by their ‘high morphological variance’. This goes counter to their reasoning for selecting the four sites to test the discriminatory power of cross-sectional morphological variance because ‘no significant differences in planview or profile morphologies have been previously detected’ in the four sites chosen. The author does say that this specifically relates to variance in width, thickness, and mass, however, it would be useful if this was made clearer that they are not talking about shape variance. The word ‘morphology’ should be amended perhaps.
- The difference between ‘overly concave’ and ‘overly convex’ cross-sections as demonstrated in Figure 1 could be clearer. I understand there are dashed lines that aim to demonstrate the relative concavities/convexities but it could be pointed out more clearly.
Correction suggestions
- Page 1, Line 7 of Abstract – change ‘preliminarily’ to ‘preliminary’
- Page 2, Line 14 – add an ‘s’ to the end of ‘connotation’
- Page 4, Line 3 – change ‘focuses’ to ‘focused’
- Page 5, Line 9 – change ‘planivew’ to ‘planview’
- Page 5, Line 10 – remove ‘the’ before ‘distinguishing’
- Page 7, Line 3 – the word ‘found’ is used twice in one sentence. Change one of them.
- Page 8, Line 3 – change ‘then’ to ‘than’
- Page 8, Line 14 – change ‘compare’ to ‘compared’
- Page 8, Line 16 – change ‘plaview’ to ‘planview’
- Page 9, Line 10 – with regard to final thinning phases, check whether they mean ‘effected’ in relation to their impact on the final shape of handaxes. Do they mean that thinning phases brought about the final shape, or that they had a significant impact? It doesn’t read too well. I suggest changing to ‘affected’.
- Page 9, Line 14 – change ‘distinguish’ to ‘distinguishes’
- Page 10, Lines 1 to 3 – first sentence reads oddly. I would remove the word ‘that’ on Line 1.
- Page 10, Line 4 – add ‘the’ before ‘midsection’ and add ‘an’ before ‘effort’
- Page 10, Line 9 – change ‘were’ to ‘where’
- Page 11, Line 8 – change ‘were’ to ‘where’
- Page 11, Line 11 – change ‘important’ to ‘importance’
- Page 12, Line 4 – add an ‘s’ to the end of ‘perspective’
- Page 12, Line 7 – add ‘of’ before ‘cross-sectional’
- Page 13, Line 4 – remove “ ‘s “ from “Callahan’s”
- Page 14, Line 4 – add ‘s’ to the end of ‘Hearth’
- Page 15, Line 2 – end of sentence doesn’t read well. Seems to be missing a word like ‘insights’ after the word ‘useful’
- Page 15, Line 3 – change ‘recently’ to ‘recent’
- Page 15, Line 6 – change ‘the’ to ‘in’
- Page 18, Line 2 – repetition of ‘are discussed’ – remove.
- Page 18, Line 7 – change ‘The’ to ‘This’
- Page 18, Line 9 – change ‘planivew’ to ‘planview’
- Page 19, Line 11 – repetition of ‘patterns of’ – remove
- Page 19, Line 13 – add ‘is’ before ‘loaded’
- Page 20, Line 9 – change ‘lesser of an’ to ‘a lesser’
- Page 21, Line 17 – change ‘Figure 6’ to ‘Figure 6A’ to reflect that Figure 6 is made up of one boxplot and three regression plots.
- Page 22, Line 4 – author refers to ‘regression graphs presented in Figure 3’ – this needs to be changed to reflect that they are actually in Figure 6.
- Page 25, Line 13 – change ‘the demonstrates’ to ‘that demonstrate’
- Page 25, Line 15 – add ‘of’ after ‘most’
- Page 26, Line 1 – at the end of the sentence that began on the previous page, there is ‘(refs).’ – this needs to be amended to reflect the references the author intends to cite here.
Referee: 2
Comments to the Author
One of the most interesting discussion topics on Middle Pleistocene technologies is the tools’ morphometrical variation. The author explores this aspect, paying special attention to the cross-section aspect. The original premise is a classical assumption: “Late Acheulean handaxes differ from earlier in terms of complexity and extent of thinning procedures”. Nevertheless, the author apply an innovative approach. By one side, each tool is divided in three techno-functional parts: tip, mid and butt. Besides, the geometric morphometrics analysis is combined with the Scar Density Index, to assess if the reduction process (number of scars per surface) affects the final morphology of tools.
The author compares handaxes from two Early Acheulean (Rietpots 15 and Sterkfontein Member 5 West) and two Late Acheulean sites (Cave of Hearts and Amazi Springs).
I think the structure is adequate, the research questions are clearly stated and the conclusions are good supported by illustrations, tables and supplementary information. So, I strongly recommend its publication. Nevertheless, I think it is mandatory to consider the next suggestions and corrections.
1. In spite of the structure is adeccuate, I suggest to devide the proposed “Materials and Method” into two sections: Materials (including general info about sites and assemblages) and Methods.
2. The method is innovative and appropriated. Nevertheless, the morphometry applies to the lithic technology requires a clear technological characterization of the instruments. This is the only way to understand the technological process beyond the shape of tools. In this case, the author only assume the number of scars as the sole variable which characterize the type of shaping. So, it is mandatory to explain: the type of blank used (cobble or flake), the % of cortex and its localization, the removal series, the presence or absence of final retouch or the edge angle (average or more than one angle per edge, at least one per each part). If possible, characterize each one of the 3 parts considered, to understand the differences between them. The method of analysis must be recorded in the Method section, together with GM.
3. By other side, it is required to characterize metrically the samples (mean, CV and SD), elongation and refinement indexes and if possible, volume information.
4. Please, include at least a Figure with instruments to illustrate the assumptions/conclusions of the research done: oval vs. triangular shapes, presence or absence of thinning works, etc,
5. Page 20. Figure 1 call. Please, include Callahan 1979 reference. Nevertheless, the differences between shapes C and D ar not clear. A good solution could be to include a real example per each shape (A, B, C and D).
6. Page 20. Line 38. “pattern” word is duplicated
7. Page 21. Line 21. Please, include here that Fig.3 refers to mid sections
8. Page 23. Line 16. The values of r2 are low to consider a clear relation between reduction intensity and manufacturing processes. But in case of a positive relation this occurs mainly on tips (r2=0.213), as you declare on page 21, lines 35-43.
9. Please, at the end of the Result section, take into account that SDI could be affected by several aspects of the knapping process. So, it is strongly recommended to define the relation between SDI and other variables, such as volume information, to get a deeper understanding of how the reduction intensity affects shape. Besides, I suggest to contrast this PCA results with any metrical information (e.g. width and thickness on tips variation). This will be very useful to, for example, reinforce the differences between the shaping strategy on tips and the rest of tools, or even between the different assemblages.
10. Page 24. Line 40. Please, remember that there is no any objective information about how the edges are, e.g. angle distribution along the edge or specific angle pattern, such us on tips to compare between the different sites.
11. The conclusion section must record all the results of this work, not only general assumptions. The author must explain which are the contributions that cross-section analysis represent to the study of technological evolution from Early to Late Acheulean.Decision letter by
Cite this decision letter
Reviewer report
2020/07/12One of the most interesting discussion topics on Middle Pleistocene technologies is the tools’ morphometrical variation. The author explores this aspect, paying special attention to the cross-section aspect. The original premise is a classical assumption: “Late Acheulean handaxes differ from earlier in terms of complexity and extent of thinning procedures”. Nevertheless, the author apply an innovative approach. By one side, each tool is divided in three techno-functional parts: tip, mid and butt. Besides, the geometric morphometrics analysis is combined with the Scar Density Index, to assess if the reduction process (number of scars per surface) affects the final morphology of tools.
The author compares handaxes from two Early Acheulean (Rietpots 15 and Sterkfontein Member 5 West) and two Late Acheulean sites (Cave of Hearts and Amazi Springs).
I think the structure is adequate, the research questions are clearly stated and the conclusions are good supported by illustrations, tables and supplementary information. So, I strongly recommend its publication. Nevertheless, I think it is mandatory to consider the next suggestions and corrections.
1. In spite of the structure is adeccuate, I suggest to devide the proposed “Materials and Method” into two sections: Materials (including general info about sites and assemblages) and Methods.
2. The method is innovative and appropriated. Nevertheless, the morphometry applies to the lithic technology requires a clear technological characterization of the instruments. This is the only way to understand the technological process beyond the shape of tools. In this case, the author only assume the number of scars as the sole variable which characterize the type of shaping. So, it is mandatory to explain: the type of blank used (cobble or flake), the % of cortex and its localization, the removal series, the presence or absence of final retouch or the edge angle (average or more than one angle per edge, at least one per each part). If possible, characterize each one of the 3 parts considered, to understand the differences between them. The method of analysis must be recorded in the Method section, together with GM.
3. By other side, it is required to characterize metrically the samples (mean, CV and SD), elongation and refinement indexes and if possible, volume information.
4. Please, include at least a Figure with instruments to illustrate the assumptions/conclusions of the research done: oval vs. triangular shapes, presence or absence of thinning works, etc,
5. Page 20. Figure 1 call. Please, include Callahan 1979 reference. Nevertheless, the differences between shapes C and D ar not clear. A good solution could be to include a real example per each shape (A, B, C and D).
6. Page 20. Line 38. “pattern” word is duplicated
7. Page 21. Line 21. Please, include here that Fig.3 refers to mid sections
8. Page 23. Line 16. The values of r2 are low to consider a clear relation between reduction intensity and manufacturing processes. But in case of a positive relation this occurs mainly on tips (r2=0.213), as you declare on page 21, lines 35-43.
9. Please, at the end of the Result section, take into account that SDI could be affected by several aspects of the knapping process. So, it is strongly recommended to define the relation between SDI and other variables, such as volume information, to get a deeper understanding of how the reduction intensity affects shape. Besides, I suggest to contrast this PCA results with any metrical information (e.g. width and thickness on tips variation). This will be very useful to, for example, reinforce the differences between the shaping strategy on tips and the rest of tools, or even between the different assemblages.
10. Page 24. Line 40. Please, remember that there is no any objective information about how the edges are, e.g. angle distribution along the edge or specific angle pattern, such us on tips to compare between the different sites.
11. The conclusion section must record all the results of this work, not only general assumptions. The author must explain which are the contributions that cross-section analysis represent to the study of technological evolution from Early to Late Acheulean.Reviewed by
Cite this review
Reviewer report
2020/07/10Summary
This pilot study represents a valid attempt to outline a new method for quantitatively distinguishing Early Acheulean handaxes from Late Acheulean handaxes by looking at the variance in their cross-sectional morphologies. Principal components analysis and canonical variate analysis of cross-sectional morphological variance of tip, midsection, and base of handaxes from two Early Acheulean sites (Rietputs 15 and Sterkfontein Member 5 West) and two Late Acheulean sites (Cave of Hearths and Amanzi Springs), as well as regression analysis of cross-sectional shape and reduction intensity, brings the author to conclude that later Acheulean handaxes are characterised by increased thinning as a result of advancements in flaking mechanics and thinning procedures as has been suggested by previous research. Therefore, the methodology presented is argued by the author to be successful in discriminating between handaxe assemblages of different periods.
The author initially discusses how planview and profile morphologies of handaxes have been predominantly used in attempts to distinguish Early from Late Acheulean handaxes, with the general hypothesis being that handaxes become more refined (characterised by an increase in symmetry and thinness) over time. However, the author rightly references research which suggests that this trend towards symmetry and refinement is not a universal one (McNabb and Cole 2015), that biface shape does not correlate with thinning reduction intensity (Shipton and Clarkson 2015) and that analysis of profile morphologies only goes so far in providing insight regarding the success of thinning procedures. The author discusses research (Stout et al. 2014; Shipton 2018; Garcia-Medrano et al. 2019) that identifies advancements in flaking mechanics and thinning procedures as broadly characteristic of Late Acheulean handaxes and focuses on this aspect (thinning procedures) to test whether the extent of thinning can reliably differentiate Early from Late Acheulean handaxes.
The method proposed involves 3D-scanning handaxes from four South African handaxe assemblages (Rietputs 15, Sterkfontein Member 5 West, Cave of Hearths, and Amanzi Springs) and investigating 2D cross-sectional morphological variance at the tip, midsection, and base. This methodology is influenced by experimental research by Callahan (1979) which suggests that cross-sectional morphologies can highlight the success of thinning procedures, as well as Shelley (1990) who suggests cross-sectional morphologies can relate to degrees of knapping skill. These experimental insights, as well as previous research (Li et al. 2018; Caruana and Herries 2020) reporting no significant differences in planview or profile variation from the four sites above, constitute the main rationale for the adoption of cross-sectional morphological analysis in this pilot study. The author uses cross-sectional shape variance analysis to test its ‘discriminatory power’, with the assumption being that experimental understanding of what delineates successful, skilled thinning will be mirrored in the difference between Early Acheulean handaxes and Late Acheulean handaxes in this pilot’s sample of sites.
The cross-sectional shape of tip, midsection, and base are captured through semi-landmarks that provide the outline morphology of each cross-section. The author normalised the shapes via a Procrustes analysis to remove size influences when analysing morphological variation via Principal Components Analysis. Using insights from Callahan’s (1979) experimental work, the author analysed deformation grids of shape variance to identify cross-sectional shapes that demonstrate successful thinning or flaws. Then, using the principal components captured, the author performs a canonical variate analysis in order to test the power of this method to discriminate between handaxe assemblages. The author further tests the ability of cross-sectional morphological variance to discriminate between Early and Late Acheulean assemblages by modifying Shipton and Clarkson’s (2015) correlation of planview shape morphology and Scar Density Index (which tests the ratio of flake scars to surface area and is used as a proxy for reduction intensity) to correlating Scar Density Index with cross-sectional morphologies – the assumption being that cross-sectional shapes indicating increased thinning will correlate with Scar Density Index where planview shape morphology did not.
The author finds that cross-sectional morphologies of handaxe tips are more reliable than midsections and bases at discriminating between handaxe assemblages where ‘no significant differences in planview or profile morphologies have been previously detected’. However, CVA analysis was able to confidently attribute a significant amount of handaxes to the correct assemblage based on a combination of all cross-sectional morphologies (tip, midsection, and base). Additionally, regression analysis of principal components finds that only PC1 correlates in a statistically significant manner with Scar Density Index but only in a weak correlation. The author argues that the significant correlation of PC1 scores and SDI values is demonstrative of cross-sectional shape variance as being related, in some part, to reduction/thinning intensity. Although the regression analysis is left somewhat separate from the PCA and CVA analysis, the author persuasively argues for the use of cross-sectional morphologies as a method for distinguishing between assemblages and proposes that it should be tested further with larger sample size.
This is a thought-provoking proof of concept paper which will equally interest researchers focusing on the Acheulean in South Africa as it will interest those researchers focusing on the Acheulean in the rest of Africa, Europe, and elsewhere. The methodology outlined by the author has the potential to influence other researchers to replicate it towards their own research interests. However, there are a number of assumptions and areas that are not adequately explained or sufficiently demonstrated. With revisions, this paper is worthy of publication.
The main issues I have with this paper are:
- The author remarks that the four sites analysed were chosen because they exhibited ‘no significant differences in planview or profile morphologies’ and this would allow for the newly applied methodology detailed in this paper to be tested with regard to discriminating between them. However, apart from referencing papers to which this is discussed, there are no graphs or data that discuss this lack of variance. It would be useful to see how the planform and profile morphological variance compares to the variation demonstrated in the PCA’s of tip, midsection, and base variance between the four sites.
- Moreover, the papers (Li et al. 2018; Caruana and Herries 2020) referencing the lack of variation among planview and profile morphologies from the four sites use a different methodology for calculating shape variance to the one used by the author here. The papers referenced do not calculate shape variance in a ‘size-free’ way (Caruana and Herries 2020: 93) by using semi-landmarks and performing Procrustes analysis to normalise allometric effects as is presented in the submitted paper. Li et al. (2018: 32) ‘use measurements and indices developed by Bordes (1961) and Roe (1964, 1968)’ which, although designed to look at shape, are not equivalent to the ‘Fifty evenly-space semilandmarks’ used in this submission to measure the morphology of cross-sections. In reference to my first point, the reasons why different methods are used here need to be addressed and/or explained.
- Linked to this issue of shape variance is the repeated conflation of ‘shape variability’ and ‘asymmetry’. Although I suspect that this has not impacted the purpose or results of the study submitted, it is something that should be revised and/or explained as on my first reading of the paper it was confusing as to what was being tested. An example of this is (Page 7, Lines 14 to 15): ‘the Tabun sample displayed a marked increase in asymmetry, i.e. shape variability’. Shape variability is not the same as asymmetry – theoretically, you can have a pointed handaxe with the same index of asymmetry as an ovate handaxe but the two of them have different shapes. Moreover, the author references Shipton and Clarkson (2015) who the author uses (Page 4, Lines 5 to 6) as an example of a paper that is ‘focused on planview symmetry in handaxes’ and which finds that ‘shape variability and reduction intensity are independent variables’. Shipton and Clarkson (2015) do find that shape variability and reduction intensity are independent variables, as they also do with asymmetry when testing handaxes from Broom, however, Shipton and Clarkson (2015) and Shipton et al. (2019), who are both referenced by the author, separate shape/morphological variability and asymmetry and do not use them interchangeably. However valid the author’s comments with regard to shaping processes having little effect on producing symmetrical forms are (Page 4, Lines 7 to 8), using planview shape and planview asymmetry interchangeably should be amended.
- The author uses experimental studies by Callahan (1979) and Shelley (1990) which suggest that certain cross-sectional morphologies are indicative of reduced thinning success and can correlate with signatures of variably skilled action such as step/hinge fractures (overly concave/convex) or mismanagement of flaking surfaces (rhomboidal shapes) that hinder thinning. The author uses these general shapes in conjunction with the knowledge of advancements in thinning procedures highlighted and attributed to the Late Acheulean at other sites (Stout et al. 2014 etc.) to make the prediction that thinner and more elongated ovular shapes are the result of successful thinning and as such should be demonstrated more in Late Acheulean assemblages. I am happy with this initial prediction but the author uses these shapes (reproduced from Callahan) as a high-fidelity universal method for identifying thinning success without corroborating with qualitative observations of whether any of the identifiers of unsuccessful thinning that Callahan (1979) refers to are visible on the handaxes that are more rhomboidal, concave or convex in the author’s dataset. The author concludes on page 24 that a distinguishing factor between handaxe assemblages is ‘knapping mishaps’ and that ‘most of the variance detected in the PCA analyses related to shapes resulting from either poor management of bifacial flaking surfaces or step/hinge fractures’, however, there is no discussion prior to this regarding whether observations of step/hinge fractures are more prevalent amongst those with more rhomboidal, convex or concave cross-sectional morphologies. Those aspects of the handaxes are assumed based on patterns from experimental observations. It would be useful if there was some discussion of this and/or more explicit acknowledgment of the assumption that certain shapes are being used as proxies for said ‘less skilled’ behaviours.
- The author’s conclusion that assessments of cross-sectional shape can distinguish handaxes from different periods of Acheulian industry using the broad distinction that Late Acheulean handaxes are thinner, more elongated, and ovular than Early Acheulean handaxes is somewhat confused by the results from Amanzi Springs. The Amanzi springs shape variance overlaps Early and Late Acheulean patterns and is explained by the author through analysis performed in a separate paper which suggests handaxes from Amanzi Springs were ‘abandoned before reaching the final stages of manufacturing due to production flaws’ (Page 23, Lines 1 to 2). Therefore, on the one hand, it serves to bolster the author’s argument for using observations from Callahan (1979) to distinguish between handaxes with more or fewer flaws related to greater/lesser success of thinning but, on the other hand, suggests that using cross-sectional morphological analysis to discriminate between Early and Late Acheulean handaxes is improved through qualitative observation and cannot be used universally with great confidence without such observations. This will be made clearer when the author increases the sample size of their study but this may be something to comment on (Page 24, lines 1 to 2) when the author notes that the ‘results of the study remain preliminary and require further testing with expanded sample sizes’. However, the author (Page 21, Line 10) states that cross-sectional shapes ‘strongly discriminate handaxes from Early and Late Acheulean assemblages’, something that arguably needs to be tempered given the importance of qualitative observations of Amanzi Springs to the explanation of the morphological variance it exhibits. The point regarding the strength of the discriminatory power of cross-sectional morphological analysis would be made more powerfully if, as suggested above, there were graphs/data outlining the comparative planview shape variability.
Minor Issues
- Abstract exceeds word-limit. It is currently 118 words long.
- I would recommend that the term ‘standardization’, used several times, be explained more. The author notes in the abstract that ‘standardization of planview and profile morphologies has been used to distinguish Early and Late Acheulian assemblages’ but explains in the introduction that some research has shown that ‘increasing standardization of handaxe forms is not a ubiquitous trend characterizing the Acheulian’. The author links it with the notions of refinement, however, it would be helpful if there was a more detailed outline of what is meant by ‘standardization’ with regard to the discussion of changes in Acheulean handaxes over time.
- Page 5, paragraph starting on Line 12 – when introducing the methods used regarding capturing cross-sections from handaxe tips, midsections, and bases, as well as plotting semi-landmarks, it would be useful if there was a reference to a figure depicting this. Moreover, an earlier explanation regarding the rationale of where measurements are taken from would be welcome – the first mention of what constitutes tip, midsection, and base comes on Page 16 in the ‘Outline extraction’ section. This is easily resolved by including what tip, midsection, and base refer to when the author first mentions ‘Cross-sections from handaxe tips, midsections and bases were captured from digital thin-sectioning’ (on Page 5, from Line 12). Just add in brackets.
- Page 8, last sentence of the first paragraph from Lines 8-10 – make clearer the fact that the ‘maxim’ you refer to is one that relates to change over time.
- Page 16 – ‘Outline extraction’ and ‘Geometric morphometric methods’ sections. To repeat, key aspects of process and methodology could be demonstrated in a Figure. I understand that there is a limit of six Figures but it would be useful visually to show where measurements were taken from and how semi-landmarks were spaced around each cross-sectional profile.
- Hypotheses would be more clearly presented by using bullet points. It would allow the reader to easily find them again in the text.
- In the ‘Discussion’ section on page 22 (lines 13-15), the author notes that one of their own recently joint-authored papers characterises handaxes from Amanzi Springs by their ‘high morphological variance’. This goes counter to their reasoning for selecting the four sites to test the discriminatory power of cross-sectional morphological variance because ‘no significant differences in planview or profile morphologies have been previously detected’ in the four sites chosen. The author does say that this specifically relates to variance in width, thickness, and mass, however, it would be useful if this was made clearer that they are not talking about shape variance. The word ‘morphology’ should be amended perhaps.
- The difference between ‘overly concave’ and ‘overly convex’ cross-sections as demonstrated in Figure 1 could be clearer. I understand there are dashed lines that aim to demonstrate the relative concavities/convexities but it could be pointed out more clearly.
Correction suggestions
- Page 1, Line 7 of Abstract – change ‘preliminarily’ to ‘preliminary’
- Page 2, Line 14 – add an ‘s’ to the end of ‘connotation’
- Page 4, Line 3 – change ‘focuses’ to ‘focused’
- Page 5, Line 9 – change ‘planivew’ to ‘planview’
- Page 5, Line 10 – remove ‘the’ before ‘distinguishing’
- Page 7, Line 3 – the word ‘found’ is used twice in one sentence. Change one of them.
- Page 8, Line 3 – change ‘then’ to ‘than’
- Page 8, Line 14 – change ‘compare’ to ‘compared’
- Page 8, Line 16 – change ‘plaview’ to ‘planview’
- Page 9, Line 10 – with regard to final thinning phases, check whether they mean ‘effected’ in relation to their impact on the final shape of handaxes. Do they mean that thinning phases brought about the final shape, or that they had a significant impact? It doesn’t read too well. I suggest changing to ‘affected’.
- Page 9, Line 14 – change ‘distinguish’ to ‘distinguishes’
- Page 10, Lines 1 to 3 – first sentence reads oddly. I would remove the word ‘that’ on Line 1.
- Page 10, Line 4 – add ‘the’ before ‘midsection’ and add ‘an’ before ‘effort’
- Page 10, Line 9 – change ‘were’ to ‘where’
- Page 11, Line 8 – change ‘were’ to ‘where’
- Page 11, Line 11 – change ‘important’ to ‘importance’
- Page 12, Line 4 – add an ‘s’ to the end of ‘perspective’
- Page 12, Line 7 – add ‘of’ before ‘cross-sectional’
- Page 13, Line 4 – remove “ ‘s “ from “Callahan’s”
- Page 14, Line 4 – add ‘s’ to the end of ‘Hearth’
- Page 15, Line 2 – end of sentence doesn’t read well. Seems to be missing a word like ‘insights’ after the word ‘useful’
- Page 15, Line 3 – change ‘recently’ to ‘recent’
- Page 15, Line 6 – change ‘the’ to ‘in’
- Page 18, Line 2 – repetition of ‘are discussed’ – remove.
- Page 18, Line 7 – change ‘The’ to ‘This’
- Page 18, Line 9 – change ‘planivew’ to ‘planview’
- Page 19, Line 11 – repetition of ‘patterns of’ – remove
- Page 19, Line 13 – add ‘is’ before ‘loaded’
- Page 20, Line 9 – change ‘lesser of an’ to ‘a lesser’
- Page 21, Line 17 – change ‘Figure 6’ to ‘Figure 6A’ to reflect that Figure 6 is made up of one boxplot and three regression plots.
- Page 22, Line 4 – author refers to ‘regression graphs presented in Figure 3’ – this needs to be changed to reflect that they are actually in Figure 6.
- Page 25, Line 13 – change ‘the demonstrates’ to ‘that demonstrate’
- Page 25, Line 15 – add ‘of’ after ‘most’
- Page 26, Line 1 – at the end of the sentence that began on the previous page, there is ‘(refs).’ – this needs to be amended to reflect the references the author intends to cite here.
Reviewed by
Cite this review
