Saturday, October 9, 2021

Short Draft Ancient MakerSpaces Talk

 [Page 1 - Title Page]

The Mycenaean Atlas Project is a relational database dedicated to storing accurate locations of Late Helladic find spots in continental Greece and the Aegean.   It has taken six years to reach this point; the effort is entirely self-funded.  The number of BA sites is now more than 4400; the number of features is more than 6500.


It is now a tool that allows quick reference to specific individual sites and which also supplies information relevant to that site.  It should allow researchers at whatever professional level to quickly investigate the several sites which characterize Bronze Age Greece.  These goals include allowing users to have accurate coordinates for sites and to use the DB in follow-on uses.

It features an intuitive user-interface that allows for easy exploration of the BA landscape.  There is a dedicated page for each of the 4400+ sites.  The DB is accessible in several ways which include a nearly-completed API.  Those researchers who would like to obtain the full DB should contact me through e-mail.

[Splash Page Slide]

Shortly after beginning the DB work I began to develop the software to put the DB online.  This is the site; there is open access to this site - anyone may use it.  It currently hosts 4400+ site pages as well as 6500+ sites - not Bronze Age - that I call Features.  These include things such as towns, bridges, churches, etc.

The use of the site is straight-forward.  The easiest way is to enter a site name such as 'Mycenae' or 'Tiryns' in the search box and then click on the returned link.  This will bring you to the correct site page.  The site page itself has a search box so that you can, if you wish, continue with another search or you may return to the control page.

[Slide of data sources: BIBLIO]

I began this work with a copy of Simpson's Mycenaean Greece from 1981.  At the start I supposed that I had no need for a bibliography table in the DB.  Simpson would be enough.  By the time I reached my third site I had seen the error of my ways; at the present time the bibliography contains 1700+ titles and may be seen by anyone using the site.  On the splash page you may examine the Atlas' coverage of any of the 40 most significant Gazetteers used in creating the DB.  The sites were located in various ways, through gazetteers with good locational information, examination of maps such as Topoguide, Topostext, ... even user contributions such as Wikimapia.  On my blog there are many examples of the techniques I used to find specific sites.  The internet has given wide access to the scholarly literature and many a dissertation was examined for information.   Each site was, if possible, confirmed in several ways before being put in the Atlas.

What information is available for particular sites?

[Slide of a single site page]


Each site has its own page with
  • Locational information
  • Accuracy indicator
  • Elevation
  • Type of site and finds
  • Ceramic horizons associated to the site

On the right-hand side of the individual site page we have the
  • Location Notes
  • Site Bibliography

[Slide of four individual site tools]

Tools for individual sites
  • Intervisibility: This shows all the sites in the DB which are intervisible with your chosen site.  It looks out 6.5 km.
  • Aspect and Slope: This analyzes the slopes at the site and tries to show which direction the site faces
  • Nearest neighbors:   This draws a map which shows your site's nearest neighbors along with the direction to those sites.
  • Three dimensional modelling of the site environment:  There are 3D terrain models for most of the sites; these were prepared for me by Xavier Fischer of

[Slide of four types of analysis for groups of sites]

Groups of Sites

In addition to single sites you can also analyze sites as groups, e.g. all peak sanctuaries or all sites in Messenia.  You define these groups from the Control page.  If you want to group sites from certain regions then you can choose those regions from the handy thumbnail maps on the Control page.  Once your group is created you can generate group reports:
  • Group aspect: This report looks at the slopes and aspects of all sites in the selected group
  • Gazetteer: This deceptively simple list of all sites in the group turns out to be among the most useful reports.  The gazetteer has info on each site along with a link to each site in the group.
  • Elevation: The elevation report lists the elevation for each site along with a histogram and elevation graph for the group.  It also provides other statistics.
  • Bibliography: After you've selected a group the Bibliography Report will generate a list of sources that were used to create that group.
  • Chronology: This generates a chart of all the ceramic horizons that are characteristic of the group you selected.  
 [Accuracy Slide]

A primary goal of the DB was accuracy of location.  Analytical tools will only work properly when the input data is accurate and I made a commitment to expend every effort to achieve that goal.  Of course many of these sites are not clearly locatable - I will deal with that in a moment.

There is also a powerful search facility and an extended help page.

Final Ideas

   1. There are any number of online databases that can be integrated to your software.  When you design software you should identify other databases that can enhance your own product.  The elevation.api Interface is an example of this but there are other DBs such as hydrography or geology that may add to and enhance your product.

2. It's not clear to me that for a product of this type a simple lat/lon to name pairing will be adequate for undertakings of this type.  These products need a story to go along with the lat/lon pair.  We need to know more about sites than just specifically where they are.  More like Topostext.

Summary and Future

API or Application Programming Interface.  

One use of this DB may be to use the tables independently in other completely different DBs.  You may for example be developing a DB on Mycenaean weapons.  

In order to express the location of these finds you will have access to a kind of Mycenaean Site API that would 'serve' locations to online clients.  The API should be able to serve elevations, lat/lon pairs in several types, alternate names, lists of ceramic horizons, biblio names, etc.

[Thanks to contributors]

Guide to Posts that concern finding site locations in Greece

I've written a number of posts in the last few years (over two different blogs) that concern the location of various BA sites in Greece and how I found them (or didn't).  Here I present a list of those posts which should make it easier to find them all. 

 Aigialea (Achaea) (C5004, C5005, C5006, C5008)

Amarynthos: Palaiochora (Euboea) (C1223)

Chalandritsa Region of Achaea : (C653, C654, etc.)

Galatos and Stalos, Crete (C5762, C5763)

Lambaina Quarry (Messenia, C131)

Malesina, Hagios Georgios (Locris) (C5169)

Metsiphi on Euboea (C6878, etc.)

Orchomenos (Boeotia) and Lake Copais 

Tithorea, etc. (C5152)

Valta in Messenia

Sunday, August 22, 2021

A Draft of a talk to the Ancient MakerSpaces Conference in January 2022


I am very happy to have been invited to give a talk in January to the Ancient MakerSpaces conference in San Francisco. This is a workshop session of the AIA-SCS (American Institute of Archaeology and the Society for Classical Studies) and it provides a forum for individuals or groups who have been pioneering digital applications in several fields of the humanities.[1]

My talk, devoted to the Mycenaean Atlas Project, is only ten minutes and so affords just about enough time to state the goals of the M.A.P. , show some representative slides, and talk about lessons learned. And yet I took this opportunity to write a longer report about the M.A.P. which explains some of my positions in more detail. Those who would like to learn more about this project than is available in a 10-minute talk can refer to this short paper.

The Mycenaean Atlas Project
A Table of Contents for the Greek Bronze Age

'He dreamed of Thebes and Camelot,
And Priam's neighbors.'

The Mycenaean Atlas Project is a relational data base that is devoted to accurate locations for find sites of the Bronze Age in Greece. Currently it consists of 4400+ sites, 6400+ features, and nearly 4000 three-dimensional landscape models. More than six years of daily effort has gone into this project. Early on the decision was taken to make the database free and public and, to that end, the online site was created. This effort was undertaken because, to my knowledge, there was no previous effort to digitally map the entirety of the Mycenaean world.

There are now a number of digital products that try to map the worlds of ancient Greece and Rome in whole or in part.   Some of them are:
  • The excellent Aristeia Project which covers the Geometric and the Archaic for Greece.  It is  conducted under the leadership of Dr. Mazarakis-Ainian; it has some problems of accuracy but on the whole is an excellent effort.  It would be even better if it were to show signs of continuing curatorship.  No project, and definitely not my own, is error-free and so the commitment to upkeep and correction is vital. 
  • There is the Ancient Ports - Ports Antiques site which is designed and implemented by Arthur de Graauw who is an engineer specializing in harbor construction.  It is relentlessly accurate and Mr. de Graauw allows the download of his entire DB in the form of a flat-file or spreadsheet. This enables the user to convert it into a relational DB or read it into a GIS for further analysis.
  • There is the ArchaeoCosmos initiative of the Capodistria University of Athens.  This seems to be an umbrella site for housing disparate and non-integrated projects related to the geography of ancient Greece but this site appears to be still in its beginning stages. I wasn't able to reach anything useful through their online interface.
  • There is the specialized site called Topostext under the leadership of my friend Brady Kiesling.  Topostext successfully marries the landscape of Classical, Hellenistic, and Roman Greece with thousands of site names mentioned in the ancient literature; invaluable for any readers of those texts.
  • The Barrington Atlas is the basis for the Pleiades project which covers the Archaic through Roman periods.  I have some questions about their attempt to repurpose the Barrington Atlas data.
  • A very useful survey of digital projects relating to the Northern Aegean (and Greece generally) can be found in the paper by Despoina Tsiafaki and Natasa Michailidou, 'Ways to Cope with the Scientific ARENA: Taking the Results of Archaeological Research a Step Further', which came out of the Conference on Cultural Heritage and New Technologies (CHNT) no. 23, 2018.  The paper is online here.  This paper reflects the approach of the ARENA digital project which is dedicated to making information available about a large number of find-spots in Macedonia and Thrace (8th to 1st centuries BC).  Their site is here.
  • The Luwian Studies site, directed by Dr. Eberhard Zangger, is an excellent and highly accurate, source of information about western Anatolia and I have leaned on it heavily as it covers an area with which I have little familiarity.
  • The Digital Atlas of the Roman Empire seems comprehensive and well-intentioned but rather pedestrian as it tells the user little of what he or she would like to know.
  • The finest site I know of that marries specific kinds of information (Levantine ceramics) with an atlas is the Levantine Ceramics Project whose home page is here. The history of the development of this project is here. This site is proof that a committee can, on occasion, design something splendid. It provides multiple entry points to its data and a nearly faultless search system. Let me give an example: I started by searching for 'Mycenaean' and then learned that that was a synonym for 'Philistine Monochrome'. I clicked on that name and got a page describing Philistine Monochrome ware and its dates. On that page was a list of sites where such ware was found. I then clicked 'View on Map' and was taken to a leaflet map plate on which the area where such ware is found was outlined. The map page also listed the find sites for this ware. When the 'vessels of this ware were found' check box was checked a list of a dozen or so places was displayed. Then I clicked on one of those (K.6.162a) a close-up map of the site was displayed with further buttons to look at individual items.  Here is what that page looks like:

In the center, and surrounding the site is a spiral of paddle images.  Each one is clickable and supplies more information about the items found there.  This web site is a perfect marriage of function and information with an atlas.

  • The Digital Archaeological Atlas of the Holy Land, or DAAHL, is the interface to a DB that records on the order of 100,000 find sites in the Levant and Egypt.  I do not see an institutional sponsor; I believe that it is solely the work of the consortium of scholars who are listed under 'Contributors'.  It is a very valuable repository of information about the Levant but it gives distinct signs of being unloved.  Broken links and links that go nowhere, software errors, confused usability issues - all these make the experience more frustrating than it needs to be.  I hope that this resource can find the curatorship that it deserves.  This site has the potential to be as good as the Levantine Ceramics Project but right now - it isn't.

There are also many partial gazetteers of the ancient world in articles and books. A good example of this is Goodison and Guarita which covers the Mesara in Crete.  Such sources can be found in the database bibliography and scattered throughout the site itself.  The M.A.P. coverage of some forty of the most prominent can be examined directly from the Control Page of the M.A.P. itself.

What is required to identify a site?

Computer people like to think in -tuples - the smallest set of things required to accomplish a task.  Doubles, triples, and so forth.  In this case the question is what is the smallest number of elements required to uniquely identify a site?  I maintain that this is a quadruple because site identification requires four things. 

  • The first is a unique identifier that must be more than just a simple number.  It  must be unambiguous, applied to only one site, and the same denomination rule must apply to all sites.  It cannot be sub-classed; its role is not to explain but to identify.  
  • The second item is a latitude/longitude pair.  It doesn't matter what system.   UTM/WGS84/GGRS87 and even What 3 Words are all interconvertible without loss of accuracy. 
  • The third item is a rating that expresses the degree of confidence in your location.  Many sites in this field are not precisely locatable and that raises the problem of reliability.  A subjective measure of accuracy goes some way to increasing confidence in placement. 
  • The fourth and last item is actually the most important.  That is a bibliography of the site by which one may demonstrate that everyone is talking about the same thing.  This reduces, although it does not eliminate, the problem of thinking that two distinct sites are one, or of promoting one site into two based only on significantly different descriptions.  Without a bibliography these are errors that are all too easy to commit.
Notice one thing here.  The concept of a name, a string of characters or digits, uniquely designating the site is missing.  The Greek name-space is sparse and the result is a great deal of repetition.  The concept of the 'name' of a site is not essential to identifying it.  The question here is 'what can we dispense with in uniquely identifying a site?'  The name-string is one of those things that we can dispense with.

Why is accuracy important?  There are two reasons.

The first is that accuracy makes it possible to ask and answer important questions. My very strong feeling is that, in the field, and no matter what you find, whether it’s a worn, barely recognizable sherd or a palace complex, Datum One is where the object was found, exactly.    Why is accuracy of location so important when generations of archaeologists have supposed it to be unimportant?  Accuracy of location is important because only that can relate your specific find to everything else.  For example,

  • how far is it on the average from a BA habitation to a water source?  
  • What’s the standard deviation of that distance?  
  • What’s the average elevation of a BA settlement, tholos, chamber tomb?  
  • Is the average habitation above or below the average BA cemetery?  
  • What’s the average distance from a habitation to its associated cemetery (when such an association can be determined)?  
  • How many BA habitations do we know that were within 100 m of the ocean?  500 m?  1000 m?  
  • Did the Mycenaeans live in the mountains?
  • In the LH IIIC settlements were often moved to higher elevations for safety. This is probably true. Can we prove it?
  • What proportion of BA habitations were obviously maritime in orientation or were not so oriented?  
  • How many habitations with a LH IIIB2 or LH IIIC burn layer are there and how are they distributed, exactly?   
How about some accurate and useful maps of all those variables?

Second, accuracy is crucial because on that rests the potential to perform an almost unending array of analyses using already available databases.  Without accuracy of place there is no accurate elevation data, no accurate slope or site aspect analysis, no usable 3D modelling, no intervisibility analysis, no accurate positioning in relationship to water sources or geological setting.  There are a number of databases about Greece which are online which could be fruitfully integrated to a Bronze Age database such as the M.A.P. but which rely on accurate locations in order to be useful.   Garbage-in, garbage-out is the inflexible rule

A good example of this (nearly mis-fired) promise is given in an article by Wong  [2020] which demonstrates the types of analysis possible - in this case for the island of Naxos.  Naxos was explored very early in the twentieth century by, among others, Stephanos.  He discovered a number of cemeteries whose locations were not always adequately recorded.  Despite the best efforts of later researchers, notably Doumas and Renfrew, there is still a great deal of imprecision concerning the exact locations of some of these sites.   Consequent to the surveys of the Stelida Naxos Archaeological Project (SNAP) Wong wrote a series of invaluable articles describing the various types of analysis that were available along with very suggestive results.  Unfortunately many of Wong's sites were incorrectly placed (this was not his fault) and so his analysis was somewhat weakened.  Nonetheless I strongly recommend these articles for their valuable array of analytical techniques.

The great advantage of the digital approach is that errors (inevitable) can be corrected rapidly and cheaply.  The cost gradient for digital work is so low and the work  time so rapid that it seems obvious that the day of the printed book or article (along with their infuriating pay-walls) is finished.  Already researchers have more rapid access to a far greater array of material online than they would have even in all of our research libraries with, perhaps, the exception of the very greatest.

As fine as the prospect of powerful analysis may be the plain fact is that BA and other historical studies are imperceptibly moving from the data-gathering phase into the data input phase.  This will last for another half century at least.   I emphasize that this is my personal opinion only.


In light of this the Mycenaean Atlas Project is dedicated to carrying out a Baby Step. 
  • The goal of the MAP is simply to specify, as accurately as possible, the location of as many Bronze Age sites as are discoverable.  The emphasis is on the culture of the Mycenaeans but the actual extent of the Atlas is further than that.
  • To provide additional material to illustrate why each of the sites was included.  To that end nearly all sites have additional information that tries to say what they were or how they were used as well as the period of time in which they were built or used. 
  • To provide analysis tools that allows further investigation of the sites in context.
  • To make all of this information available to whoever might wish to use it or refer to it.  To that end the MAP takes the form of a relational database of BA sites.  It has online presence because it was wrapped in PHP software.  From a usability point of view this makes it exactly the same as the vast majority of sites on the internet. 


The entire DB is now available to anyone who wants it either in .sql form or through the online interface.  Its URL is or

Given that this goal of inclusive and accurate placement can be achieved it is crucial to try to show what kinds of analysis can be carried out on such a database. 

First what is the granularity at which these sites can be examined?  Sites may be analyzed and examined singly or in groups.  Single sites are identified by their place key which takes the form Cnnn or Cnnnn.  Typing the  place key into the search box (on any page) will take you to the Place Key Report page for that site.


<More about single site analysis: Types of analysis available are intervisibility, site aspect, nearest neighbors, and three-dimensional models>

  • Intervisibility shows what sites are visible from your chosen single site within circle of 13 km in diameter. It returns a map with your central site in blue and intervisible sites in green.
  • Site Aspect allows the examination of the slope and aspect of the ground around any site. It computes the slope at 150 and 300 m. from the site in the four cardinal directions. Based on these eight measurements it identifies the aspect of the site (which way it faces).
  • Nearest Neighbors is a multi-way display of the distance and bearing from any site to its neighbors within a 1.5 km radius. This returns a map with the nearest neighbors in a circle of 3 km. radius. It provides graphs of distance and direction
  • Three-dimensional models are available for about three-quarters of the sites. I was assisted in this by Xavier Fischer of He created the models which are hosted on And although this is a different site the model experience is seamless to the user.

The Mycenaean Atlas Project controls page allows the user to define groups of sites in order to analyze them as a group.   The simplest means of defining a group  is by region; you may, for example,  be interested in all the sites on Aegina or in Messenia.  You may do that from the Control Page either by pull-down menus or by using the thumbnail maps at the bottom of the page.  When you click on the thumbnail it expands into a page-size map from which you can choose your region of interest.
You can also choose groups of sites by Ceramic Horizon (Period) or by type or by some combination of the three criteria (Region,  Period, Type).

Once selected you can choose to get a map of these sites or a series of detailed reports.

There are five different reports available for each group search:

  • Aspect and Slope -This page provides an analysis of all the slopes and aspects of the sites in your group
  • Gazetteer - Simple but useful.  An alphabetical list of the sites in your group complete with  links to the sites.
  • Elevation - This provides charts and analyses of the elevations of the sites in your chosen group.
  • Chronology - This is a chart which shows all the ceramic horizons attested for your site groups
  • Bibliography - This consists of a bibliography of all the works used in investigating the sites in your chosen group.

Lessons Learned

  • There are many databases online that cover important topics such as hydrography and geology.  These can be integrated into a product like the Mycenaean Atlas Project in the same way that the API for elevations was integrated (  It is good to plan for this from the beginning.
  • It might be time to redevelop this product to use  a robust GIS such as ArcGIS or QGIS.  The M.A.P. was developed using a rudimentary GIS which consisted of Google Earth  Professional along with database support in the form of MySQL.
The Greek toponymical name space is relatively sparse.  The ancient Greeks were the sort of  people who took along a whole 'culture package' when they migrated to new areas.  This resulted in their naming new features with old names that had been applied to similar things. 

Counter to this re-use of names each site has a long history of renaming, duplications, abandonments, and resettlements consequent to dialectical development, conquest, reconquest, population movement, and changes of language.  Much archaeology in this world was done a long time ago and as a result obsolete names are often embedded in the archaeological literature.  This constitutes a problem of its own.

And yet it is this that gives BA toponymy its subtlety and charm.

  • As more databases  with differing time  periods are integrated the problem of mutations of names becomes more acute.  It is not be enough any more to content ourselves with a simple name or ID mapping to a lat/lon pair.  We need a more dynamic toponymical model that is suitable for historians as well as geographers.  This should be a model that links 'places' diachronically and not just in space.  It might be better to think of replacing Greek toponymical names with an accompanying story of name change (if I could think of a way to represent this).  This in turn starts with a completely redesigned database required to support such a presentation style.

For the Future

  • The future will undoubtedly see the further integration of databases that are now separate.  In order to facilitate this task the international community should define a standard relational database format to which individual contributors can adhere.
  • Compendia of site and  place locations will be of little use unless each of the sites is curated.  Databases that match site name with a lat/lon  pair will never be of much use.

Thanks go to:

The primary contributor to the Mycenaean Atlas Project besides myself is Dr. Sarah Murray of the University of Toronto in Canada.    Several years ago I read her article (Murray [2014]) and learned that she had gathered a database of some 4000 LB and early IA sites.  She generously agreed to share this data with me.  From that data I culled about 2900 sites and from these was able to enrich those I already had as well as add about 1600 more that I did not yet have.

I have very generous friends and collaborators in Greece who have been kind enough to send me pictures and current observations  of several sites.  Among these I must mention 'Pete'  for working with me enthusiastically and selflessly for the last five years.  He has been instrumental in correcting errors and enhancing my understanding of a number of places in Messenia.  This product would be much poorer without his involvement.  Also Dr. Hajo Becker from Germany who has contributed photographs and whose numerous trips to Greece as well as thoughtful observations have helped me to precisely place several sites.

Dr. Michael Boyd (Oxford) kindly shared photographs and information about Kopanaki in Messenia.

Dr. Alex Knodell (Carleton College) worked with me on finding specific locations for many sites in Sterea Ellada.  I thank him warmly for affording me this opportunity.

Dr. Raphael Orgeolet (Aix-Marseille University) generously contributed location corrections for Kouophovouno (C988).

Dr. Dimitri Nakassis (University of Colorado at Boulder) contributed corrections for C6679, Chosti on Naxos.

Many scholars have been kind enough to provide copies of their papers.  Among these is Dr. Rainer Felsch who  kindly provided his paper, ‘'Das Kastro Souvalas bei Kalapodi'’, which allowed me to correct the position of C6887 (when I had no other means of doing so).

I would be very remiss in not thanking Dr. Daniel Libatique and the professionals of the Ancient MakerSpaces organization who afforded me the opportunity to describe this project.

And, of course, special thanks to Michael P. Speidel, Professor Emeritus, University of Hawaii at Manoa who, all those years ago, set my feet on 'the right path.'

Any  errors that remain in these pages are mine. Any opinions that I state here or in the M.A.P. are also mine and mine alone and cannot be attributed to any of the people that I have named.


[1] The SCS website is here. The AIA website is here. The Ancient MakerSpaces page for the 2022 session in San Francisco is here.


Felsch, Rainer [2018]:  Felsch, Rainer , 'Das Kastro Souvalas bei Kalapodi' in Papakonstantinou et al. [2018] 73-82.  Online here.

Goodison and Guarita [2005]:  Goodison, Lucy and Carlos Guarita. ‘A New Catalogue of the Mesara-Type Tombs’,  Studi Micenei ed Egeo-Anatolici [47], pp. 171-212, 2005.  Online here.

Murray [2014]:  Murray, Sarah, "Qualitative Data, Hypothesis Testing and Archaeological Narratives: Was there ever a Greek Dark Age?" in C. Papadopoulos, E. Paliou, A. Chrysanthi, E. Kotoula, and A. Sarris (eds.) Archaeological Research in the Digital Age: Proceedings of the 1st Conference on Computer Applications and Quantitative Methods in Archaeology. Greek Chapter (CAA-GR), Rethymno, Crete, 6-8 March 2014. (IMS-FORTH): 64–69. Online here.

Papakonstantinou et al. [2018]:  Papakonstantinou, Maria-Foteni and Charalampos Kritzas, Ioannes P. Touratsoglou (editors). Πύρρα Melétes gia tin archaiología stin Kentrikí Elláda pros timín tis Fanourías Dakorónia. Athens, SEMA Ekdotike.

Wong [2020]:    Wong, Todd.  'GIS Summary of SNAP 2019 (Part 3): Analyzing Extracted Results', esri Canada, Centres of Excellence.  Published in 2020 in the  'McMaster Blog'.  This article (part 3) is online here. Part 1 is here. Part 2 is here. Part 4 is here.

Thursday, July 22, 2021

Kamari Gouva (C173) Reconsidered

In my last post I discussed the location of a site called 'Gouva', near Kamari (UMME 236).  Gouva was a large tholos tomb from the LHIII A.  And just to recap some of the main points:

a. It has been robbed and destroyed. [1]

b. It had been so large that when the central dome collapsed (or was dug out) it left two little hills. [2]

c. These two 'bumps' on the landscape could still be seen from a nearby highway as late as the 1960's by investigators such as McDonald and Simpson.[3]

In that post I looked over all the constraints and decided on a location at  37.292614° N, 21.78284° E and I gave various reasons for why I thought that was the right location.

But now I have reason to think that I was wrong and that the real location is just a bit to the north.  A correspondent of mine in Greece, Pete, has actually been to this ridge and walked the whole length of it looking for features that might fit our tholos.  He reports:

"I took a general walk around the area either side of the ridge. Going much further south got difficult as the ridge was overgrown and I was weary of snakes in the undergrowth. Of course low scrub and trees make judging any details difficult but I can usually spot a point of interest or identify an area that might fit the bill. So none of this is conclusive, but I offer a suggestion for consideration.

I note that the report stated that the mounds were on the skyline and looked like two small hills. Or some such. So I took a walk along the skyline ridge route. I started a little further to the south of the area you had indicated and walked north, along the top of the little valley and up to the highest point some tens of meters away. Just beyond that high point (mostly covered by bushes) there was a cleared stony/earth field with a somewhat obvious mound near the higher, west side. The mound had a shallow dip in the middle."[4]

What does he mean?

This next image shows the area that he's talking about.  The various camera icons indicate where Pete was standing when he took these photos; the color matching rays indicate the direction the camera was facing.  (Photo 12 is a panorama.)  His suggested tholos location is the white circle and its lat/lon position is 37.293955° N, 21.782660° E.  North is at the top.

And in order to see his suggested mound clearly let's look at Photo 10:[5]

Proposed Kamari: Gouva tholos as seen from E.
For credit see fn. 5.

And from the SE facing NW here is photo 7:[5]

Proposed Kamari: Gouva tholos as seen from SE.
For credit see fn. 5.

Photo 8 shows the mound from the S:[5]

Gouva Mound(?) from the SE.
For credit see fn. 5.

Photo 9 shows the mound from the SW.[5]

Pete has this to say about Photo 9:

"Image 9 is taken from 'behind' the mound in question I walked though the gap in the bush to the SW of the mound as I wanted to see what the other side was like. The mound was not so obvious at this point. It was "lost" in a long linear patch of shrub and undergrowth. In fact if anything there seemed to be a short, steep bank (which the shrubs followed) instead of a smooth sloped side to the mound.

In the photo one can see that the nearest olive trees are smaller and so I suppose this immediate area has been bulldozed in the last 50 years (probably 20 or so). That might explain the missing west slope of the suggested site. It had been removed by bulldozer and  the "cut-edge" left exposed and left to weather. I could not get too close but when I tried to see if there were any stones showing I was not able to see any clearly. I did not see any regular pattern of placement of stone or anything proving or strongly supporting this as a site for a tholos. The exposed parts of the bank were composed of a gritty red earth with a lot of stones."[6]

Photo 12 is a panorama which looks at the site from the NE.[5]

Gouva Mound(?) from the NE.
For credit see fn. 5.

So.  Is this site correct?  Or at least does it fit the description in Messenia III?

Here's what McDonald and Simpson report:

"As one travels w on the road from Kopanaki to Kamari past the spring of Sanova, he sees on the w skyline what appear to be two artificial mounds quite close together.  The N-S ridge on which they lie is at least 1 km. from any road.

   On close inspection it turns out that there was originally one large mound and that its center was deeply dug out.  From a distance the heightened banks of excavated earth create the impression of two mounds.   ...    The shrubs that have since grown up in the crater suggest that the excavation may have been as much as half a century ago.  Perhaps the crater explains the local toponym of Gouva ("hole")."[7]

Can we find the road?  Can we find the spot from which McDonald/Simpson were when they saw these two mounds?

I think we can.  The stretch of road along which they were travelling is shown here:

The yellow line is the road of travel.  In Google Maps the road is labelled 'Epar. Od. Chanion-Naou Epikouriou Apollonos'.  Kopanaki is in the lower right under the red arrow and Kamari is on the upper left center under another red arrow.  The stretch of road from which McDonald/Simpson could have seen the twin mounds is somewhere along the yellow line and inside the i km. circle which is centered on the proposed new location.  This stretch of road leaves Nea Agrilia going NW and the traveler drops from the plateau on which Nea Agrilia is situated into a little valley (some 15 m. or so vertical distance) through which flows a stream (Topoguide says that it's called the 'Kopitsa') whose origin is on the N side of this road.  It turns out that the proposed stretch of the ridge on which the proposed location is positioned is only visible from the very edge of this plateau just before the drop into the valley.  This position is just to the right of the 1 km. circle on the line which represent the road.

Here's a viewshed of these valleys and ridges; it has its origin on the plateau edge as I just suggested

And I have another way of demonstrating this.  Incredibly Google Street View is available for this entire road.   I've made a photo of what that ridge line looks like from the suggested observer's position on top of the plateau.  Here it is:

O.k.  This requires a little explanation.  The road in the foreground (we're facing NNW) is the road to Kamari.  What's that gray thing?  It turns out that when you draw a line in Google Earth and the extend it down to the ground you can often see that extended line in Google Street View.  This scene places us on the right road looking more or less west to the ridge.  The line here was drawn starting at the  proposed tholos position and extended to the observer's position just ahead of us on  the road.  I then extended it to create this unmissable indicator.  The left vertical edge of this figure is directly over the tholos and the ragged corner here imitates the changing elevation of the ridges between us and the observer.  To see where the line hits the ridge line it is enough to continue the edges with arrows to the point where they meet on the ridge.

If Pete is right then this is the area of the ridge that MacDonald/Simpson would have seen.   

Can we see two little mounds from here given that we know that they're there?  I'm just not sure.  The bumps on this ridge are consistent with what M/S saw but whether they're exactly the same we cannot now know.

Anyway, Pete's suggested position (which I will be adopting for C173 in the atlas) is:  37.293955° N, 21.782660° E.

A shout-out to Pete for his extraordinary work on this ridge!


There's just one more mystery to clear up.  Simpson says: 

"As one travels w on the road from Kopanaki to Kamari past the spring of Sanova, ... "[8]

Where is the spring Sanova?  I'll deal with that in my next post.


[1] Messenia III, 137. ' ..., there is practically no doubt that this was a large tholos tomb that was excavated (presumably illicitly) and completely destroyed.  The shrubs that have since grown up in the crater suggest that the excavation may have been as much as half a century ago.'  This article appeared in 1969.  All the authors known to me who have described this tomb are simply repeating McDonald and Simpson.

[2] Messenia III, 137.  ' ... [the traveler] sees on the w skyline what appear to be two artificial mounds quite close together.  The N-S ridge on which they lie is at least 1 km. from any road.'

[3] Idem.

[4] Personal communication on July 18, 2021.

[5] Pete's photos were heavily post-processed by me in Photoshop before publishing.  Typical operations consisted of contrast adjustment (sometimes selective), color-balance and enhancement, and sharpening.   

All of Pete's photographs are licensed under the Creative Commons Licence: Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).  This license means, in brief, that anyone may use, modify, or tweak this data, even for commercial use, as long as credit is given to the original author(s).

[6] Personal communication on July 13, 2021.

[7] Messenia III, 137-8.

[8] Messenia III, 137.


(All of the sources listed here simply copy Messenia III.  Aside from these there are no other reports of Gouva to my knowledge.)

Boyd [2001]:  Boyd, Michael John, Middle Helladic And Early Mycenaean Mortuary Practices In The Southern And Western Peloponnese, 2001   pg. 212   'Unexcavated Tholos Tombs'.  It is online here.

Cavanagh and Mee [1998]: Cavanagh, W., Mee, Ch., A private place: death in prehistoric Greece ((Studies in Mediterranean Archaeology, 125)).1998.  ISBN: 91-7081-178-4

McDonald and Rapp [1972]:  McDonald, William A. and George R. Rapp, Jr.,  The Minnesota Messenia Expedition: Reconstructing a Bronze Age Regional Environment.  University of Minnesota Press, Minneapolis, Minnesota.  USA. 1972.   pg. 298, no. 236, 'Kamari: Gouva'.

Messenia III:   McDonald, William A. and Richard Hope Simpson. ‘Further Explorations in Southwestern Peloponnese: 1964-1968’, American Journal of Archaeology (73:2) (Apr., 1969), pp. 123-177., '23D. Gouva (Kamari)', pg. 137. Online here.

Pelon [1976]:  Pelon, Olivier . Tholoi, tumuli et cercles funéraires; Recherches sur les monuments funéraires de plan circulaire dan l'Égée de l&rsquo;Âge du Bronze (IIIe at IIe millénaires av. J.-C).  Bibliothèques de l'École française d'Athènes et de Rome - Série Athènes, 229. 1976.   pg. 465, 'Gouva'.   It is online here.

Simpson and Dickinson [1979]:  Simpson, Richard Hope and O.T.P.K. Dickinson.  A Gazetteer of Aegean Civilization in the Bronze Age, Vol. I: The Mainland and the Islands. Paul Åströms Förlag, Goteborg. 1979.   pg. 177, 'D 236 Kamari: Gouva.'  It is online here.

Simpson [1981]:  Simpson, Richard Hope.  Mycenaean Greece. Park Ridge, New Jersey: Noyes Press, 1981.   pg. 137   'F 209 Kamari: Gouva'.

Zavadil [2012]:  Zavadil, Michaela  Monumenta: Studien zu mittel- und späthelladischen Gräbern in Messenien.  Wien:Osterreichische Akademie der Wissenschaften Philosophisch-Historische Klasse Denkschriften. 2012.   pg. 411, 'KAMARI/GOUVA (EP. TRIPHYLIAS)'.  It is online here.

Friday, June 4, 2021

Kamari: Gouva (C173)

The area south of Kamari, in Messenia, is a lightly inhabited region of small farms and large olive orchards.  It is characterized by north-south running ridges which run from the outskirts of Kamari itself as far as the east-west running highway 25 which starts at Kalo Nero on  the coast and runs inland through the Soleima valley until it reaches Zeugolateio in the Pamisos river valley.  On the south side of highway 25 lies the wild and jumbled country of central Messenia which was, so far as anyone knows, nearly uninhabited in Mycenaean times.  One of the ridges running directly south of Kamari and stretching towards highway 25 was, in prehistoric times, the site of a great burial mound which is thought to have sheltered a tholos.  The literature of this site knows it as Kamari/Gouva[1].  What is this site?  It appears that, in prehistoric times, at the site location, there had been raised a great mound which (seems to have) concealed a tholos tomb.  At  some early period the mound was destroyed by robbers and their digging left a distinct notch in the ridge-line.  Simpson remarks on this:

"As one travels w on the road from Kopanaki to Kamari past the spring of Sanova, he sees on the w skyline what appear to be two artificial mounds quite close together. ...  On close inspection it turns out that there was originally one large mound and that its center was deeply dug out. From a distance the heightened banks of excavated earth create the impression of two mounds."[2]

We further learn from Simpson that there were many large flat stones lying around.  These were foreign to the place and are supposed to be the remains of the dug-out tholos which had been completely destroyed.   It appears that nothing datable was found in the place and so the significance of this site is somewhat limited.  But the enormous size of the thing and the effort that its destroyers went to suggest that this had been an important depostion site and, perhaps, linked to a hypothetical settlement found a short distance to the north (although its remains, in the form of sherds,  are somewhat exiguous).  So where was this (formerly) important site.  Luckily McDonald and Simpson have left us enough clues.  But first let's look at a map of the area.

The area N of the Soulima Valley and S of Kamari in Messenia.

Here is the area in question.  It is bounded by three highways that form a rough triangle.   Somewhere in the middle of this triangle is our site.

Can we be more precise about the location of site C173?  Luckily, in Messenia III, Simpson and McDonald give us good bearing marks and distances.  They say this:

"From the tomb one gets a fine view to N, s and E. Several ancient sites can be seen: Mesovouni ... 10° and about 2 km. distant; Stilari ... 90° and about 4 km.; Rachi Gourtsia ... 180° and about 4 km."

In the Mycenaean Atlas these  place names correspond to these:

1. Rachi Gourtsia (C171)

2. Stilari (C164, C343)

3. Mesovouni (C172)

I took the trouble of drawing these distance circles and bearing lines in Google Earth and was gratified to see that they all seem to converge on the same place

At the top is a yellow circle centered on Mesovouni (C172) and of 2 km. radius.  The green circle is centered on the sites at Stilari (C164) and is of 4 km. radius.  The dark blue circle is centered on Rachi Gourtsia (C171) and is also of 4 km. radius.  The bearing lines from each of these centers are drawn as Simpson and McDonald would have them.  We see that they all converge on pretty much the same ridge south of Kamari.  Here's a close-up:

The blue circle and line originate in Rachi Gourtsia.  They intersect at the light blue paddle marked 1.  The green line and circle originate at Stilari.  They intersect at the green push-pin marked 2.  The yellow circle and line originate at Mesovouni.  They intersect at the yellow push-pin marked 3.  So far they seem to be in rough agreement.  The distance from 1 to 3 is almost exactly 200 meters.  Is our bowl-shaped site anywhere in the vicinity?

In the view just above we're looking directly west.  The construction lines are the same and the same push-pins are easily visible from far left center to center right, yellow, green, and blue.  Our site should be around here somewhere and, sure enough, just in the center the white push-pin denotes an area that is a dip and still has some remnant of wild bushes not completely overtaken by olive trees.  I suspect that this is our site.  To prove that it is a dip requires some additional work.

I went back into Google Earth and drew a multi-segmented line (the red line) criss-crossing the area of interest in generally a north-south direction.  I then did a 'Show Elevation Profile' which produced the image at the bottom.  The dark red vertical bar is the area just crossing our selected area (the white push-pin).  You can see that travelling from north to south that this is a local minimum.  

When the lines are drawn running east and west a different story emerges.  Here's what that looks like:

From this we see that our location at the white push pin actually sits on a slope that runs from west to east.  So: a) in a dip from north to south, b) on a slope that goes from higher in the west to lower in the east.

I analyzed this site in my Atlas by choosing an 'Aspect' report for site C173.  This is what it shows:

This report (you may have to enlarge it by clicking on it) shows that, at 150 m. from the center the aspect of the site is east and west (red arrow).  This means that the ground falls away from our site to the E and W.  At 300 m. the aspect of the site is also east and west (orange arrow).  But at both distances the slopes to north and south are positive which means that the ground rises to north and south but dips to east and west.  And it dips most significantly to the east.  And that makes sense.  Simpson would have been looking from the east and this notch in the ridge is what he would have seen.  Here's what it looks like in Google street view.

This is the view that Simpson would have had from the road going towards Kamari.  Off to the west the near ridge (some 850 m. distant) is clearly visible as is the dip in its outline.

I've updated the splash page for the Mycenaean Atlas to show some beautiful terra-cotta dancers from the New York Metropolitan Museum of Art.  You can see them here.


[1] Kamari Gouva:  in Zavadil [2012], 'KAMARI/GOUVA (EP. TRIPHYLIAS)', pg. 411;  Simpson [1981], 'F 209 Kamari: Gouva', pg. 137; Messenia III [1969],  '23D. Gouva (Kamari)', pg. 137; McDonald and Rapp [1972], '236 Kamari: Gouva', pg. 298; Boyd [2001], 'Goúva Kamári, Messinía.', 212

[2] Messenia III, 137, '23D. Gouva (Kamari)'.

[3] Ibid. 138.


Boyd [2001]:  Boyd, Michael John. <i>Middle Helladic And Early Mycenaean Mortuary Practices In the Southern And Western Peloponnese</i>. 2001.  Online here.

McDonald and Rapp [1972]:  McDonald, William A. and George R. Rapp, Jr.. <i>The Minnesota Messenia Expedition: Reconstructing a Bronze Age Regional Environment</i>,  University of Minnesota Press, Minneapolis, Minnesota. USA. 1972,

Messenia III:  McDonald, William A. and Richard Hope Simpson. ‘Further Explorations in Southwestern Peloponnese: 1964-1968’, American Journal of Archaeology. Vol. 73, No. 2 (Apr., 1969), pp. 123-177.

Simpson [1981]:  Simpson, Richard Hope. Mycenaean Greece. Park Ridge, New Jersey: Noyes Press, 1981., 

Zavadil [2012]:   Zavadil, Michaela. Monumenta: Studien zu mittel- und späthelladischen Gräbern in Messenien. Wien:Osterreichische Akademie der Wissenschaften Philosophisch-Historische Klasse Denkschriften. 2012., '. Online here.

Sunday, May 16, 2021

Changes to the Chronology Report Page

 The chronology page has been a mess since I originated it.  The central problem is that, because I was exploring how to do it, I committed the programmer's gravest sin: hard-coding things that should be dynamically figured out at run-time.  In this case I hard-coded the column header names - here the ceramic horizon names, LHIII, MMI, etc.  That led to a whole bunch of kludges that severely limited the adaptability of the page.  As an immediate consequence I had to repeat nearly all the code in a whole separate page devoted just to Minoan ceramic horizon names.  And ... endless .. I was going to have to create a new Cycladic Chronology page.  Well, worse and worse.

I have now solved all these problems by rewriting the Chronology page to be self-configurable.  It figures out at run-time what ceramic horizon headings are needed and what order to put them in.  Now a single chronology page works for all ceramic horizon combinations.  It works for Minoan, Cycladic and any other horizons in the database.  And, strangely enough, the resulting code is a lot simpler.

Here's what the new page looks like:

Strangely it looks just like the old page except simpler.  Notice that the annoying 'Minoan' button that was in the upper right of the page is now gone.

This example was generated for 'hab' sites in Achaea.  There are 15 such sites in the DB.  The left-most column lists the names of the sites along with their place keys.  Clicking on one of the place keys will create a Place Key Report page for that site.  Mousing over the column headers (ceramic horizon names) will pop up the start and end dates for that ceramic horizon that are listed for it in the DB.  Some of the charts have very many rows and, as you scroll down, you may lose the info about which ceramic horizon is intended.  Mousing over the 'x' will pop up the ceramic horizon name for that column.  The following illustrations show each of those things:

The ceramic horizon name pops up when the 'x' is moused over.
Here the cursor is simulated by the red arrow.

Here the date-range for the ceramic horizon as reflected in
the DB pops up when you mouse over the ceramic horizon name. 
Here the cursor is simulated by the red arrow.

I should finish up by saying that the new chrono page is already released so you can use it.  You get to the chrono page from the control page by 1) pressing the 'report' button and then 2) pulling down the Detail Reports drop down and selecting 'Chronology'.  

Also don't forget that you can generate .kml or .csv(s) from this page and that those can be inputs into QGIS product (and if you don't have that then you should download it and install it and start learning how to use it).  The download page for QGIS is here.

And that's all there is to it.