top of page

An Exercise in Middle Woodland Geometry VI

5 days ago38 min read

Landscape


“Our life is a faint tracing on the surface of mystery, like the idle curved tunnels of leaf miners on the face of a leaf. We must somehow take a wider view, look at the whole landscape, really see it, and describe what's going on here. Then we can at least wail the right question into the swaddling band of darkness, or, if it comes to that, choir the proper praise.”

Annie Dillard [Dillard 1974]


"In the work of art, the truth of beings (des Seienden) has set itself to work (sich ins Werk gesetzt)."


"To grasp space, to know how to see it, is the key to the understanding of building" [Zevi, 1957, p. 23]


"Philosophically, the phenomenon external to an area of interest affects what goes on in the inside; a sufficiently common occurrence as to warrant being called the second law of geography." [Tobler 1999, p. 87]


Introduction

The Middle Woodland period is an archaeological designation of a time in Eastern North America spanning roughly 200 B.C.E. to 500 C.E. This saw the florescence of complex hunter-gatherer societies, the beginning of agriculture, minimal warfare, and increased mobility up and down the region's rivers. There was a great expansion of an ancient mound-building tradition along with the building of diverse enclosures out of earth and other materials. In what is now central and southern Ohio and along the Ohio and Mississippi Rivers from Marietta to western Kentucky, plus one outlier in southern Florida, these enclosures took on increasing complexity and size using interrelated shapes based on circles and squares. These 'geometric' earthworks are some of the most beautiful and complex attainments of this cultural flowering. [Newark, the Core, and Beyond]


Analogies are an important part of human thinking. Some even believe them to be the basic cognitive structure of the human mind. [Hofstadter & Sander 2013] To say that a culture reaches fluorescence is to say that a culture acts like a flower, it builds up to some monumental or beautiful peak and then fades away. A flower continues in the form of seeds. Cultures are 'lost' but the human beings that participate in them continue in cultures of less complex form or abandon a region for better places. What is continued is memories. It is unknown if anything that happened in Ohio was remembered by the Mississippian polities like Cahokia some 700 years later. European contact erased so much modern memory that it is hard to find many later ones. [Memory]


The geometric earthworks lend themselves to analogy. Ohio has many small colleges with concert halls and stadiums. So they are like a college campus in architecture and in that the societies that built these both prized intellectual achievement. These ceremonial centers are believed to celebrate 'world renewal', [Byers 2005] especially places such as Newark with its alignments to the sun and moon. Brad Lepper eloquently calls Newark:


"... a gigantic machine or factory in which energies from the three levels of the Eastern Woodland Indian’s cosmos ... were drawn together and circulated through conduits of ritual to accomplish some sacred purpose." [Lepper 2004]


and compares it to our modern attempts to understand the universe in places like CERN. John Hancock compares Newark to 'works' of architecture and designates three interpretations:


  1. " ... 'as' archaeological sites that hold knowledge about distant cultures or

  2. 'as' monumental works of architecture or

  3. 'as' places of Native American meaning." [Hancock 2016]


In this article, I will be looking at interpretations 1 and 2, the powerful concept of place I reserve for later. [Memory] Hancock sees the archeological analogies and the analogy of a work of art as being at odds. I think this is changing and I see a resolution through his third interpretation, indigenous meaning.


A. Martin Byers uses the analogy of a court warrant, an object representing a set of beliefs and practices. The person bearing the warrant becomes a representative of these beliefs and practices. Speaking of Newark:


"Newark may be understood as just that, a monumental iconic warrant constructed in order to embody the totality of the cosmos in accordance its Solar/Lunar sacred temporal structuring and its vertical/horizontal Heavens/Underworld sacred spatial structuring."


Archaeology has its analogies, cultural florescence being one. Another is the cultural 'interaction sphere' first proposed by Joseph Caldwell in 1964. [Caldwell 1964] The rise of social complexity in the Middle Woodlands plus relative peace meant that both goods and ideas moved freely. These goods and ideas not only had individual trajectories but were used locally in a diverse manner. The 'Interaction Sphere' analogy tries to formalize this complexity. In 2020 the Midcontinental Journal of Archaeology released an issue themed 'Anarchy and Archaeology.' The lead article by Edward R. Henry and Logan Miller introduced the analogy of 'situations.' [Henry & Miller 2020] The Situationist International was an artistic and political movement in Europe, founded in 1957 and dissolved in 1972. [Wikipedia Situationist International] It represented a critique of consumer capitalism as a form of 'spectacle' [Debord 1967] and suggested the creation of spontaneous 'situations' as a way to release human freedom from the constrictions of spectacle. Henry & Miller see a problem with the Interaction Sphere analogy as all or nothing. One either sees a strictly local manifestation or a global one, and this leads to a confusing terminology of types that says little of what is going on. By looking at the local manifestation as situations, as spontaneous works by individuals, one can better understand the dynamic and diversity between local and regional interactions. A colleague of Henry, Alice Wright, uses the analogy of globalism as used in modern times to describe the Middle Woodlands.

Again the local is very personal and diverse yet still defined by the global. [Wright 2020]


I want to add my analogy for the geometric earthworks:


They are like baskets woven into the landscape.


Basketry is a highly manipulated and diverse 3D form. Basket-making goes back at least 12,000 years, the earliest baskets were found in Egypt. Basket making is probably much older and predates pottery. [Abdalla 2024] The beginnings of geometric knowledge could come from basket weaving. [Gerdes 2003] Roofed structures started as inverted baskets, and post rows could have been tied together with woven mats. Building with earth creates basket-like shapes upon the land. Impressions of baskets used to haul soil were found in earthworks in Indiana. [McCord 2006]


Landscape
Splashed Ink Landscape [Sesshū 1495]
Splashed Ink Landscape [Sesshū 1495]

North America in 1492, when Columbus landed in Cuba, and in 1750, when settlers began to pour over the Appalachian Mountains into Kentucky and Ohio, was a very different place. Disease, conquest, and strife had reduced the population by 70% to 90%, and the vast wilderness that was being 'tamed' by white settlement had only just come into existence. [Denevan 1992] North America was not a vast wilderness to its inhabitants but a 'known' landscape consisting of named 'places' connected by stories and songs. [Memory] Some of these places had been heavily modified by human activity for thousands of years. The landscape consisted of not only land and water but also the sky, a dynamic temporal element. [Sky]


Chinese and Japanese aesthetics represent some of the earliest representations and analyses of landscapes both built and wild and the relationship of humans to nature. [Lorch 2002] Fūkei is a word for landscape in Japanese. The word has its roots in two Chinese characters meaning 'wind' and 'light.' This is about how the wind changes the light of leaves and grasses as they are blown about. [Hatakeyama 2023] There are important elements here that tie into Western notions of landscape as well:


  • Landscape is an individual human experience

  • It is a perceptual experience, what one perceives. Most importantly visibility.

  • It can be a shared experience, that of intervisibility.


The Boyne River Valley is located in Eastern Ireland just north of Dublin. In Neolithic times many structures and mounds were built on the hilltops and in the river terraces including large burial mounds like Newgrange. Newgrange covers over an acre and contains an underground passage 1/3rd of the way to the center. The passage is oriented so the winter solstice lights it up for five days of the year. Radiocarbon puts the mound as over 5,000 years old, predating Stonehenge by 500 years. There are four other passage mounds in the Boyne Valley and the largest, Knowth, forms a complex with 19 other smaller tomb mounds. The passage mounds are believed to be oriented towards each other with passageways having astronomical alignments. [Boyne Valley Tours 2024]


The terraces of the river valleys of Ohio have earthworks that are similar to the Irish Neolithic. The earthworks do not contain burials but sometimes enclose burials in mounds of different types, and are often placed within or near mound complexes. There have been attempts to work out orientation between sites. [DeBoer 2010] For instance, the octagon-circle complexes at Newark and High Banks in Chillicothe are at a 90-degree angle from each other. [Hively & Horn 1984] These are the only two proven octagonal earthworks in the Middle Woodlands and both have been shown to have astronomical alignments. [Hively & Horn 2010]


The number and diversity of earthworks in Ohio alone is astounding. Earthworks do have certain noticeable design features. A. Martin Byers calls these 'traditions' [Byers 2005] and it is important to remember that each tradition has its trajectories in time and space and that individual earthworks can have elements of several of these traditions. Separate traditions may inform the various mound types found inside or near; wooden structures, lines and circles of upright posts, hearths, and middens. Any attempt at categorization is always dangerous and needs to be continually challenged. Here is a slight modification of Byers' traditions:


  • Hilltop enclosures. These enclosures ring a flattened hilltop.


  • Partial or complete non-geometric enclosures that separate some sort of nondomestic space.


  • Enclosures using earthwork and ditch combinations.


  • 'Large Geometric Earthworks',' or LGEs.


I want to limit the discussion for now to just LGEs. Large is a qualitative term although it can be quantified. Large can be in terms of the individual areas of the elements or the total area encompassed by the site. Geometric means that they comprise elements of a circle, a square, or some combination of straight lines and angles. The term geometric also means that circles and squares can somehow be geometrically related. Also, there are possible relationships between elements in terms of size or measure as related to a standard size or measure. Some earthworks are large in total area but are not geometric. Some are geometric but the total area is small although they might be large in the sense of individual size. [Measure] [Geometry]


LGEs are part of a more general class of structures called enclosures. Enclosures define a space much like a building, except an enclosure is open to the sky while a building is roofed. An enclosure divides space into an inside and outside. This division can be accomplished in many ways. It can be just a scratched line or placed stones or shells, piled earth or sod or rocks, A dug ditch or a natural drop-off, or a line of posts. Most LGEs consist of piled earth with or without inner or outer ditches. A. Martin Byers does not like the term 'enclosure.'


"... This is not a neutral term since, of course, it implies that the earthwork was built to close off a space, separating this from the 'outside,' and this further implies exclusion, a select in-group separated from an amorphous out-group, and so on. ..."


Now I agree that 'enclosure' can be a loaded term and I understand Byers' concern. His solution is to define the enclosures by the shape of their vertical cross-section. This is a nice way to encode ditch and embankment variation but says nothing about their relation to space. These forms do enclose something so I think the term works. Here is the most neutral definition I can give:


An enclosure is a delineated space open to the sky in which people do things.


LGEs are convex polygons made up of straight lines and arcs of circles, a full circle being a polygon having only one side. This polygon is not closed but is broken by open spaces. These open spaces form three general types:


  1. Openings that have only one structure, an open space in an embankment. These will be called 'entrances.'

  2. Openings that consist of multiple structures. These include what some authors call 'gateways.' I don't like this term because a gateway implies a gate, and to my knowledge, none have been found. If a gate was found then this would constitute a multiple structure. These will be called 'entranceways.

  3. Situations where an opening may or may not be an opening at all. These will be called 'unknowns' and can include:

    1. Errors in current data. Current can mean maps drawn in the mid-19th century.

    2. An embankment was built and an entrance was later dug out.

    3. An embankment was built and a notch was cut in the embankment. This notch allowed a feature on the horizon to be visible. Because it was lower it had eroded and it looked like an entrance to the map makers.


I have found only one example of the 'unknown' type. Jarrod Burks discovered this in the Circle at High Banks in Chillicothe, Ohio. A geomagnetic sweep found no break in the embankment [Burks 2013] although one exists on the Squier and Davis map. [Squier & Davis 1847, plate XVI]  The missing 'unknown' covered up an important line of sight used to calculate a possible astronomical usage for the Octagon. [Hively & Horn 1984] This suggests that it was perhaps a notch or low point in the embankment, not an entrance.


Entrance 9 is offset from vertex 12 possibly due to astronomical alignment. [Hively & Horn 1984, p. 94, fig. 7]
Entrance 9 is offset from vertex 12 possibly due to astronomical alignment. [Hively & Horn 1984, p. 94, fig. 7]
Entrance 14 was believed to be a deliberate anomaly like entrance 9 but no gap in the embankment was found by a geomagnetic survey. [Hively & Horn 1984, p. 93, fig. 6]
Entrance 14 was believed to be a deliberate anomaly like entrance 9 but no gap in the embankment was found by a geomagnetic survey. [Hively & Horn 1984, p. 93, fig. 6]
GIS

GIS stands for 'Geographic Information System.' Basically, it is just making maps using computers. Whether the analysis and visualization of spatial data make GIS an actual science is debated. To me, GIS is an extension of what is now called data science to spatiality. GIS is a specialized technique of data visualization called a map, data storage, and data structure. GIS data consists of several parts, a data frame of attributes associated with a location, a geometric structure consisting of points or lines or polygons, a model of some 3-dimensional surface like the surface of the earth called a Geoid, a projection of that surface onto a 2D plane, and a 2D system of measurement. This is encapsulated into a computational abstraction called an S4 class. [Chambers 2014] In addition there is a second important structure called a raster, a grid of spatially located attributes.


A GIS is a digital rendering of a landscape projected onto a map. Archaeology uses GIS to test models, to run archaeological experiments. Two popular classes of models associated with GIS are visibility models and mobility models. There are many other types of models, models of legacy mapping errors, [Ullah 2015] [Mainfort et al. 2011] erosion and land use, [Mitasova et al. 2013] [Magnani & Schroder 2015] human hunter/prey interaction, [Kopels & Ullah 2024] exchange, and resource acquisition. [Carr & Sears 2022] Models can be global in that they work upon the whole landscape or local in which autonomous agents interact with the landscape and each other. Complex interactions like this can lead to global patterns called 'emergence." [Janssen et al. 2019] These local models are called 'Agent-Based Models' (ABM.) [Wurzer et al. 2015]


Earthworks studies have used a model for the energy use of construction. The model outputs the number of people and the time needed to construct the work. This model doesn't necessarily include a GIS but if information like where the dirt used in construction comes from is included a GIS can be used. [Bernardini 2004] [Magnani & Schroder 2015] There is a question as to how the geometry of the LGEs was laid out and what technology was involved. For instance, it is easy to lay out a small circle using rope and a post but the size of circles like the Great Circle at Newark would require rope technology of a different order. [Geometry] Perhaps a different method was used. This would require many different models, not necessarily containing a GIS. I have as yet to find much about this.


GIS in archaeology is a data storage technique, a visualization technique, and an experimental modeling technique. I have seen attempts to build narrative products to include GIS, and ESRIs Story Maps is one example. [ESRI 2025] Of course in gaming, maps are an important element. One critique of GIS is that the map is just one representation of space, one element of human spatial thinking. Putting meaningful marks on a surface is as old as modern humans. Why so little of it is left is more a matter of material technology than anything else. Another issue is uncertainty. Errors and assumptions are part of every parameter set but errors in some parameters cause more problems than others. A way to find which parameters are most sensitive to error is called 'sensitivity analysis.' [Brouwer Burg 2016] An example of GIS is the 'Digital Elevation Model" (DEM), a raster image of the landscape where each square of the raster represents the elevation of a square of land. The resolution of a DEM is the size of this square. It has been shown that visibility and mobility modeling are sensitive to the resolution of the raster. [Kantner & Hobgood 2016] Also, the area of the raster has to be large as opposed to the region being studied as many mathematical models assume an infinite surface and if the DEM is too small edge effects will distort the analysis.


Visibility
Fields of Human Vision [Dryfuss 1960]
Fields of Human Vision [Dryfuss 1960]
View of Portsmouth, Ohio from Kentucky at Cropper Square [Greber 2011, p. 83]
View of Portsmouth, Ohio from Kentucky at Cropper Square [Greber 2011, p. 83]

For GIS, the basic element in the study of visibility is called a 'viewshed.' The surveyor and landscape architect Clifford Tandy first used the term in 1967. [Inglis et al. 2022] A viewshed is what can be seen from a particular height at a particular location in the landscape. At its simplest, a viewshed is a binary raster of seen and not seen. The assumption of this is what can be seen by a person. A similar term is 'isovist', also coined by Tandy. [Tandy 1967] These two terms are equivalent but isovist is an attempt to directly relate various measures to the psychology of human perception. [Benedikt 1979] In isovist literature, there are several important concepts. One is a 'region' which is the maximum extent of a smooth 3D surface given the curvature of the surface. 'Environment' is the roughness of this surface brought about by structures and elevation changes in the landscape. This is a set of isovists calculated on some grid placed over the region. Grid points that share similar isovists constitute a 'visibility graph.' In GIS, these like points are termed 'intervisibility.' 'Shape grammar' comes from the sociological aspect of human interaction with architecture. [Hillier & Hanson 1984] This type of thinking is relational resulting in a spatial network. Current shape grammar studies have incorporated isovist visibility graphs into their analysis. A final isovist term is 'object.' An object can be mobile or semi-permanent. An example is the Great Circle at Newark. Standing at the entrance to the Great Circle, what was seen before it was constructed, after, with the addition of the Great House in the center, and after the Great House was decommissioned and covered over. Since we don't know the exact temporal order, what can be learned by studying visibility in time?

Critiques of GIS deal with the inherent limitations of the map as an abstraction of a downward view, 'the God Trick', [Haraway 1990] and especially with limitations to the viewshed approach. Despite isovist theory, the connection between view and human perception is still controversial. One solution is phenomenological, the perceptions of the researchers themselves. This was first proposed by Christopher Tilly in 1994. [Tilley 1994] One problem with earthwork studies is how does one record personal perceptions when the earthworks no longer exist? When they are covered by a town? When at best one has a magnetic shadow on the ground and at the least an old map? There are solutions. One is 3D modeling and virtual reality, not perfect but a start. John Hancock's work at the University of Cincinnati. [Hancock 2014] [Hancock 2019] Current comments on the landscape can be important. In an article about a recent walk of the Great Hopewell Road, the narrator mentions Gerhart Road, the last place he sighted Sugar Loaf Hill by Chillicothe. [Stanzione 2009]


The Visual and Spatial Structure of Landscapes by Tadahiko Higuchi is a seminal work in landscape analysis. [Higuchi 1988]  Higuchi combines the phenomenological empirical insight of Japanese landscape esthetics within a quantifiable framework. A view is made up of flat perceived surfaces that range from vertical to near horizontal. An important concept is that of 'horizon.' Humans perceive the distant horizon as a flat vertical surface. The horizon is the shape of the furthest view of a region. It is a vertical cylindrical raster that can be unrolled onto a planar surface. It can be binary or it can incorporate various horizontal distances and angles from the viewer to various elements on the horizon. Japanese landscape combines built and natural views which manipulate what the viewer perceives. Open compounds and gardens are common elements. What is shown outside the near view, not blocked by walls is termed shakkei in Japanese, translated as 'borrowed scenery.' [Lin 2021] The visible horizon is not static but varies according to atmospheric conditions and weather. It is also limited by the curvature of the earth. Higuchi notes the importance of optical illusion in manipulated environments. This is particularly important regarding intersecting and non-intersecting lines.


Monsters in the Hallway Illusion. Both monsters are the same size [Eye 2 Eye Opticians Harrow 2024]
Monsters in the Hallway Illusion. Both monsters are the same size [Eye 2 Eye Opticians Harrow 2024]

LGEs include a major feature, straight embankments that connect at angles and parallel and non-parallel walled walkways.


Higuchi identifies several important human viewpoints in a landscape. One is from a promontory, looking down from a high point over a landscape. another is before the entrance to a feature, and a third is immersive, just at or after the entrance, and at various points controlled by the design. For any enclosure, the centroid is an important element. For angular enclosures, corners are highly visible elements. LGEs can contain various objects: structures, mounds, middens, hearths, and lines of posts. Sometimes they are centrally located. LGE squares and octagons can have up to eight entrances and small mounds, called 'gateway mounds' in some cases bar the initial view, dividing it into two locations after one enters. It also rounds the angles when looking from the inside. There are also many departures from this pattern.


Astronomical archaeology deals with the relationship between built or marked objects and lines of sight between the immersed viewer and a point on the horizon. These lines of sight designate the places on the horizon marking the rising or setting of celestial objects. Researchers from Earlham College, Ray Hively and John Horn (H & H) from 1982 have been studying the astronomical significance of Newark, High Banks, and other LGEs. [Hively & Horn 1982] [Hively & Horn 1984] [Hively & Horn 2010] [Hively & Horn 2016] Their main theory is that the Great Circle was laid out according to lines of sight between the entrance and a point on the horizon and the Octagon/Observatory Circle complex was laid out in a similar fashion. Indeed, all of Newark Earthworks has been laid out according to a fixed set of horizon points. The line of sight between an observer at the entrance and a horizon point is called a 'foresight.' The point on the horizon viewed from the entrance is called a 'horizon foresight.' The point on the horizon looking back to the viewer is called the 'backsight.' In 2011 Christopher Turner wrote a critique of H & H's work [Turner 2015a] [Turner 2015b] in which he analyzed point H1, a promontory point believed to share a line of sight with both the Great Circle and the Octagon/Observatory Circle. This used GIS and viewsheds were constructed at H1 and the lines of sight at each location. What he showed is the variability of lines of site and the fact that from H1, the Great Circle is not visible. This difference is mainly due to a higher resolution mapping, the resolution of the DEM being the most sensitive parameter in a viewshed analysis. He doesn't reject H & H's theory of shared lines of sight but adjusts the location of H1 based on better information.


Viewshed based on the new location of H1, the old location is the large green dot. [Turner 2015a, fig. 38]
Viewshed based on the new location of H1, the old location is the large green dot. [Turner 2015a, fig. 38]

The Chacoan landscape of New Mexico consisted of 'great houses,' large monumental structures, and hundreds of smaller 'lesser great houses.' These are made of worked stone and consist of many rectilinear rooms, some stacked in multiple stories, and large circular subterranean structures called 'kivas.' Around 1100 B.C.E. new structures called 'tower kivas' started being built. From one to four story high stacked circular rooms. They are called kivas because they are round but their actual usage is speculative. One theory is that they were used as communication towers between distant houses. In 2003 Kantner & Hobgood tested this theory using a DEM and the then-current viewshed software. [Kantner & Hobgood 2003] They concluded that the towers could see each other. In 2016 the same researchers revisited the problem with a larger and higher resolution DEM and viewshed software that compensated for distance. [Kantner & Hobgood 2016] This time they were able to show that the towers were not visible to each other and what the towers did was make the houses more visible to people on the ground. In 2018 a third study was made of the whole region using what was called a 'total viewshed.' [Dungan et al. 2018] Also called 'continuous viewsheds' or 'cumulative viewsheds,' These are more than one viewshed combined to create a view from multiple points in the landscape, in the case of a total viewshed, a grid was placed over the region. Because of the computational time to create such a large viewshed with so many points, a parallel program was run on a graphics processor. In addition, it was determined that these combinational viewsheds were not sensitive to resolution allowing for a coarser DEM. It was determined that the houses were built to be optimally viewed by the local population.


In 2012 Rebecca Rennel did a study using a methodological dialogue for what she calls 'experiential landscape archaeology' encompassing GIS and phenomenological practices. [Rennel 2012]


"The rationale behind experiential landscape archaeology is that human experience informs understanding, knowledge and action and, by implication, contributes to the development of social practices, social relationships and social structures. Therefore, exploring how people experience archaeological places and landscapes can help inform an understanding of the past. From a phenomenological point of view, emphasis is placed more specifically on embodied and multi-sensuous experience." [Rennel 2012, p. 511]


Her phenomenological approach is specifically in terms of field survey practices. The study was in the Outer Hebrides, a group of islands off the west coast of Scotland. These islands contain many Iron Age structures from 400 B.C.E. to 200 C.E. Her study involves two landscape scales, a regional island scale encompassing 196 known site locations and a local scale of each site. Eleven locations were picked for phenomenological surveys, visibility and audibility around each site, and journeys between sites. She looked at sites on small islets in freshwater lochs located within a larger island. A GIS cumulative viewshed model was created. It showed an open view around the islet. The phenomenological categories were quite different. Everything beyond the water was considered distance and the effect of the surrounding water made the site feel enclosed. Rennel sees this difference between these two methods as complimentary, the difference between what is looked at and what is perceived.


Mobility

Humans move about. In North America there were no pack animals except for dogs and dogs are too small to ride but perhaps teams pulled sleds. So people walked or they used boats. Watercraft go way back in time but there are few examples due to their fragility. In the Woodlands, there were two types of boats, dugouts and canoes. Dugouts are made by taking a log and hollowing it out using fire. Canoes consist of a frame covered by a thin outer shell, usually bark or hide. Different types of bark were used but the best was the bark of white birch. Birch canoes were light and could easily be carried over a portage. Birch bark canoes are what come to mind when one thinks about Native American canoes. Modern canoes have the same basic design, just different materials. [Adney 1964] Birch forests retreated quickly north after the last glacial maximum and most birch is found north of the lower Great Lakes except for a stand in the northern Appalachians in Pennsylvania.


The current range of White Birch [Little 1971]
The current range of White Birch [Little 1971]

It is unknown how far back canoes go as they don't age well. In Middle Woodland Central Ohio there may have been no birch trees but canoes or packages of bark could have been traded or made and taken south. Dugout canoes have been found at various locations. Three were found in Ashland County, Ohio in a portage area between the Lake Erie watershed and the Ohio River watersheds near Savanna Lake. [Brose & Greber 1982] These have been dated during the Late Archaic, some one thousand years before the earthworks. Recently, in Wisconsin at Mendota Lake near Madison, a whole series of dugouts have been found dating from Archaic to precontact. [Lehr 2024]


The Canoes of Dejope [Lehr 2024]
The Canoes of Dejope [Lehr 2024]

Along with watercraft, there was an extensive system of foot trails through the Woodlands. How old these trails are is unknown. There seems to have been an uptick in travel and movement of goods from the Early Woodland to the Middle Woodland. It may have been that travel was safer or that religious fervor created pilgrimages. It could also be that technological innovations in canoe building allowed for further travel.


Movement models start with the individual, the agent. To move or not and in what direction and how far is a decision. Thus agents are autonomous, they contain the computational power to make their own decisions. Individual movement can be explored by a class of models called Agent-Based Models (ABM.) [Romanowska et al. 2021] Individual movement has two parameters, heading and distance. Individual movement decisions apply to all biology down to the single cell. One form of movement is searching or foraging. Both heading and distance are considered random. Randomness takes on many forms according to what type of random distribution is used. A uniform distribution creates what is called Brownian motion, the motion of individual molecules. An agent using this type of search goes to the same spot many times and expands the range of the search very slowly. Another distribution would be Gaussian. When a Gaussian distribution is used to model the heading an agent turns mostly -45 to +45 degrees and much less any larger angle. This gives the agent less chances of landing on the same spot.


Probability of a heading direction based on a Gaussian distribution [Romanowska et al. 2021. p. 115]
Probability of a heading direction based on a Gaussian distribution [Romanowska et al. 2021. p. 115]

A third distribution uses a Cauchy distribution on the distance which creates what is called a Levy Flight. [Chechkin et al. 2008] Levy Flight is considered an optimal random search in terms of maximum ground covered with minimal revisiting of the same location. Levy Flight is used in a biological theory called Optimal Foraging Theory which states that organisms evolve their search routines to optimize for Levy Flight. This theory has been very controversial, new statistical tests had to be developed when current tests were found to have failed. Recent research has shown that many anomalies exist that don't fit the theory, agents switch between different distributions, sometimes within a continuous rather than a discrete range. Also, time may be needed as a third parameter in terms of pauses and speed of motion. [Reynolds 2018] Just with two parameters and three possible distributions, there are 9 possible models available. With more parameters and models this expands combinatorically.


Another type of motion is goal-oriented motion. How does one get from point A to point B? In GIS this is modeled using an optimal path. A region is turned into a cost surface that identifies the cost of moving through brush, woods, or other terrain. Another cost surface is derived from the DEM which shows the cost of vertical movement. This function is called Tobler's Hiking Function (THF), created by Waldo Tobler in 1993. {Tobler 1993]



"Where v is walking velocity in km/h, α is a constant that controls the maximum velocity, g is the gradient of the terrain measured as the tangent of the angle of the slope in the direction of travel, β1 controls the rate of decline as the gradient increases, and β2 offsets the gradient by some amount to capture how walking speeds are highest on a slight downward slope." [Higgins 2021]


Comparing the function estimates with fitness tracking apps on cell phones gives a difference with THF of 1.3%. This changes to 9.4% in certain urban situations like stairs. [Higgins 2021] THF is a good start and eventually, there will be a set of calibrations based on actual walking data. A horizontal surface could deal with weather or environmental conditions like wind, rain, temperature, and humidity. Barriers form a set of features such as mountains, canyons, rivers that can only be crossed at certain points, large bodies of water, predators, forbidden places, or hostile groups.


Optimality may not always be the purpose of the journey. Does one wish not to be seen by others? Is getting there quicker more important? Is the journey sacred? Notice that these alternate purposes are directly tied to visibility, which is called a route vision profile.


An alternate means to create a cost surface uses electrical circuit theory. [McLean & Rubio-Campillo 2022] [Howey 2011] This allows for a better visualization of the resistance to travel in the landscape. Also, since a mobility surface is constructed, different journey scenarios can easily be constructed as paths or corridors.


Finally, there is wayfinding. Wayfinding is how humans gather knowledge of a landscape and transmit this knowledge to others. Wayfinding is a way to create, communicate, and use a mental abstraction of a journey. I will talk about this more in [Memory]


A PhD thesis by Nicholas Reseburg at Redlands University specifically targeted mobility in Middle Woodland Ohio. [Reseburg 2013] He emphasizes river travel with only short land hikes. Also, the method of travel was dugouts and not canoes. The thesis was not a research project but the building of a GIS tool for his advisor, the anthropologist Wesley Bernardini [Bernardini et al. 2013] Unfortunately I can find no evidence that the tool was ever used or any trace of Reseburg. Bernardini started working with the Hopi Nation soon after that. One problem is that Reseburg used the commercial software system ESRI which is headquartered in Redlands, California, and provides major funding for Redlands University. Also, mobility analysis is an active area of research with new ESRI implementations [ESRI 2023] and open-source implementations using the GRASS GIS system. [Frank 2024]


Embanked walkways, paths between two parallel embankments, are a feature of many earthworks. Sometimes, they connect two geometric earthworks and other times, they lead to a river. This happens in Portsmouth, Marietta, and Newark. There are three in Newark and one is unusually straight and leads from the Octagon southwest at an angle of 211 to 213 degrees to where it hits Ramp Creek. The Salisbury brothers in 1862 wrote that they followed it another six miles further south until it disappeared into swampland. Caleb Atwater in 1820 first proposed that it extended to Circleville or Chillicothe. Brad Lepper took on this idea in 1993 proposing a straight 60-mile sacred road built as a pilgrimage route between Newark and Chillicothe. [Lepper 2024] Extensive road building exists in Mesoamerica with the Maya, [Shaw 2001] in the Southwest at Chaco Canyon, [Sofaer 2011] and in Florida, the Belle Glades Culture built straight canals from south central Florida west to the Gulf coast. [Johnson 1998] Lepper named this roadway the 'Great Hopewell Road' (GHR.) Various researchers have found evidence of the road existing both before Ramp Creek and beyond [Schwartz 2016] [Romain 2008] with analysis of the Ohio 2023 LIDAR data by Jamie Davis [Davis 2024] showing possible parallel walls up to 8 miles south of the Octagon, close to the Salisbury extent. There is also evidence that the roadbed was lined with white limestone. Here is a cross-sectional drawing of the Marietta 'Sacrae Viae:'


Cross Section of a Parallel Embankment [Schwartz 2016 p. 20, fig 10]
Cross Section of a Parallel Embankment [Schwartz 2016 p. 20, fig 10]

That the GHR extended for eight miles is extraordinary. Whether it goes any further is still an open question. Lepper found a few other locations but he now thinks they may have been false positives. James Marshall looked around Baltimore, Ohio, and found nothing. [Lepper 2024] Rock Mill in Fairfield County is another location as it is a possible crossing point of the Hocking River and there are earthworks overlooking the river. In 1820, Caleb Atwater mentions the road crossing the Hocking River. The old name for the river: HockHocking, supposedly means "Bottle" after the wide area of the pool below the fall at Rock Mill. [Partridge 2016] I have looked at the Rock Mill location and both the terrain and the angle of the road suggest a more eastern location. Rock Mill would make more sense if the road was turning towards Circleville.


"Lepper also cites an account of local historian Samuel Park, who recorded in 1870 the presence of a graded road visible in the timber on the land claim of Jesse Thompson near Walnut Creek in Fairfield County. Lepper also reports that in 1930 Newark businessman Warren Weiant Jr used an airplane to follow parallel walls from the Newark Octagon south-west in a straight line to Millersport, Ohio." [Lynott 2015]


Walnut Creek is near Baltimore, Ohio. Unfortunately, there is as yet no high-resolution LIDAR south of Licking County. Where exactly the road ends at Chillicothe is also open. High Banks Works is a logical choice as it has the only other confirmed octagon and the Octagon-Circle combination is at a 90-degree angle to the Newark Observatory Circle- Octagon. Unfortunately, it is shadowed by a ridge from Newark. Lepper believes that the road leads to an old crossing point of the Scioto River. This would have the road just passing the eastern hills and is a good candidate for a southernmost extent. [Lepper personal communication] The new angle of 2013 degrees based on the new LIDAR data has the road ending at Hopeton [Turner 2011] or Mound City. Mound City is on the western side of the river so there would still have to be a crossing.


Two GIS analyses have been done regarding the Great Hopewell Road. One in 1999 by Jennifer Pederson [Pederson 1999] I cannot find much about it. Brad Lepper describes it thus:


"Jennifer Pederson ... examined the projected route of the Great Hopewell Road across a single USGS quadrangle map (Stoutsville) as a case study. She found that, at least in this limited area, the route corresponded to a remarkably linear geographic boundary between relatively flat, poorly drained soils to the west (Till Plains) and relatively rugged, well-drained soils to the east (Glaciated Appalachian Plateau). The mostly flat and dry ground would therefore have been an optimal route for pedestrian travel" [Lepper 2024, p. 16]


In 2007 Timothy Price [Price 2007] did a least-cost path analysis of the terrain between Newark and Chillicothe. Most of this path falls within a 1/2 mile buffer around the GHR. He also used data from the Ohio Archaeological Inventory to show that the GHR went through a region devoid of many Early and Middle Woodland sites. This makes sense since these were riverine peoples and this was a land route between two watersheds. They had to ford several rivers. Also, if this were a pilgrimage route, it would take between 4 and 8 days to make the journey. Schwartz suggests that stopovers would be good places to look.


Given the importance of water features, what still exists within the GHR corridor can be significant. The fact that the formal embanked road may have ended in a swamp is an important data point. Journeys to the bottom of a watery underworld abound in native woodland stories and a swamp can be crossed with some difficulty walking or "paved" with log planks. [Stuijts et al. 2008] The region was first crossed by Christopher Gist in 1750. [Gist 1750] After the Revolutionary War the Continental Congress gave land to ex-soldiers. A tract called the 'Refugee Tract' was given to Canadians who had fought on the American side. In 1801 Elnathan Schofield surveyed this area including a large swamp where Buckeye Lake was built. The swampland extended westward in patches. [Bonar and Lisska 1991] Did the embanked roadway resume after the swamp ended?


Portion of Scofields' 1801 Refugee Tract Map showing 'The Great Swamp' [Scofield 1801]
Portion of Scofields' 1801 Refugee Tract Map showing 'The Great Swamp' [Scofield 1801]

I got my Masters in GISP in 2014 at the University of Arizona. The one-year intensive program had no time for theory or critique although it was part of the Geography department and geography is rich in GIS theory and critique of the map. Archaeological critique is mostly the work of British archaeologists in the 1990s. I did some research on my own and had a question that I thought needed to be asked; "Is this map just a bad graph?" People in general didn't like me asking this question. My work during the last ten years with Santa Cruz County in Arizona has been with visualization, workflow tools, building digital assets, and now, mainly, the maintenance of these assets. I have not done as much with mobility and visibility. I use the software from the GIS giant ESRI which is very expensive and often overhyped. The company constantly publishes glossy full-color examples of beautiful maps made by its clients with the idea that these maps are driving positive change. Unfortunately, without a shred of evidence. I call these "Nice maps of the Apocalypse."


A recent article by Ullah et al. uses textual analysis to parse out the history and trajectory of GIS in archaeology. [Ullah et al. 2024] This article is a treasure trove of references. They found that although methodologies and acceptance are growing steadily, what GIS means to archaeological theory is spotty, no one journal is dedicated to this, only discussions in special issues or compilations of conferences specifically covering this topic. In the book, Re-Mapping Archaeology, published in 2019 Gillings et al. lay out a set of criteria of what maps could be:


  • "Maps are never stable and we continually need to question what a map is, as well as what the potential consequences are of its creation.

  • Our maps have histories (and genealogies), and we need to understand these in all of their nuanced detail.

  • How we map shapes what it is possible to do with the maps we create.

  • Our maps can act and should be encouraged to do so. We need to accept that our maps can be affective as well as effective and must embrace their performative character.

  • There is nothing wrong with maps that are argumentative, discordant, disruptive, playful, provocative or simply beautiful.

  • There should be no limits on what is deemed mappable."


As a metaphor a landscape is likened to a 'palimpsest', a document that has been erased and written over many times but past markings have never been completely removed. A landscape is not just about space so any representation of a landscape must take into account movement and time. [Alfred & Lucas 2018] Each completed representation instantly becomes another artifact holding its own mistakes. biases, and erasures. The landscape of Ohio has been modified by humans since the last Ice Age ended. The building of the great earthworks was just one episode in a continual process of doing and forgetting. Traces left behind, a sherd of pottery, an old map, a slight change in the shape of a field. Like humans, the land has memory. In a final metaphor replace the word 'map' in the quoted list below with 'landscape' or 'earthwork.'


  • ".. a map is a construction by the labours of the people who make it and use it.

  • It is a discovery in a form that is hidden but revealed through particular skill and understanding.

  • It is an invention, because without the promise of fieldwork or its materialisation in practice, it does not appear.

  • ... it is a convention, encoded and given symbolic value by us which without decoding would be just ‘scribbles’."


 






Part VII - Sky


Part VIII - Memory


Part IX - Quidnunc


 

Commentaires

Drop Me a Line, Let Me Know What You Think

Thanks for submitting!

© 2023 by Train of Thoughts. Proudly created with Wix.com

bottom of page