The NRCS Web Soil Survey

A couple years ago, this blog published a post about geology and shovel testing. The purpose of that post was to explain what every field technician should know about geology and soil science while shovel testing during an archaeological survey. If you haven’t read that post, I would recommend reading it here; otherwise this new post will be difficult to follow. My academic sources are listed at the end of that blog post, in case you would like to do additional reading.

To briefly summarize my original post, I explained that soils can either be residual or depositional. Residual soils have formed in situ from weathered bedrock. The horizons that you see in the profile of a residual soil are not chronological layers of sediment; they formed within the existing geological material due to biological and chemical processes. Artifacts left on the surface of residual soil are often buried, but not by the deposition of new sediment. They are buried by bioturbation and incorporated into the A horizon (topsoil). Meanwhile, depositional soils have been transported from their original location and deposited elsewhere as sediment. Soil can be transported by wind, water, glaciers, and gravity. Archaeological sites found in depositional environments, such as floodplains, have the potential to be deeply buried, or possibly stratified by time period.

What I didn’t address in my original post is that it can sometimes be tricky for field techs to determine how the soil beneath their feet was formed or transported, even if they have a good academic understanding of geomorphology. It’s not always easy to tell how old a landform is or how it came to be.

For example, it’s essential for archaeologists to be able to distinguish floodplains from upland landforms, because floodplains have the potential to contain deeply buried sites. In some places, the upland areas grade so imperceptibly into the floodplains that it’s very difficult to notice the elevation change at all, and the upland landforms are sometimes almost as perfectly flat as the floodplains, because they formed from some kind of alluvial activity sometime in the deep past. In the Sam Houston National Forest in east Texas, many of the “upland” areas belong to the Lissie Formation, which formed from fluviomarine sediment during the middle Pleistocene. The Lissie Formation is extremely flat because it is basically a series of deltas that formed along the coast when the sea level was much higher than it is today, during one of the Pleistocene interglacial periods (during the Pleistocene, the climate was generally cooler than it is today, which would cause sea levels to drop, but there were interglacial periods during which the climate was hotter, the sea was higher, and the coastal deltas formed at a higher elevation than they could today). So the Lissie Formation consists of a flat, ancient terrace that no longer accumulates any sediment (and has not done so since the middle Pleistocene). But the bayous flowing through the Lissie Formation are depositing Holocene alluvial sediment along their banks, and often, this alluvial sediment is barely any lower than the upland terrain. While walking through this national forest, sometimes the only way that I can distinguish the upland areas from the Holocene floodplains is that the floodplains have standing water and the upland areas have slightly less standing water.

My point is, understanding the history and formation of a landform can be difficult. There’s an entire academic discipline, known as geomorphology, which is devoted to this endeavor. I don’t think it’s necessary for every archaeologist to formally study geomorphology. But I do think archaeologists should be willing and able to use and (at least partially) understand the data that have been collected by real geomorphologists and soil scientists who have dedicated their careers to understanding the earth.

Luckily, soil scientists have made this fairly easy for us, at least in the United States, thanks to the United States Department of Agriculture Natural Resources Conservation Service (USDA NRCS). And that brings us to the topic of this post—the NRCS Web Soil Survey.

The NRCS is a federal agency dedicated to the study and conservation of soil, particularly in the interest of sustainable agriculture. Soil scientists employed by the NRCS have surveyed most of the land within the contiguous 48 states, classifying the soil into hundreds of different “series” and documenting the vertical profiles of each series. A “soil series” is a taxonomic unit consisting of soils that have been grouped together due to similarities in their formation, chemistry, and physical properties, such as color, texture, structure, pH, consistence, and mineral and chemical composition. The information collected by the NRCS is mainly intended to help farmers, but it is a valuable asset for archaeologists as well.

Archaeologists usually refer to NRCS data in their reports, but almost always in a very superficial way that doesn’t explain why the information actually matters. We list off the different soil series that have been mapped in our project areas and tuck that information into the front matter of our reports, along with other background information about the environment or geology. We seldom give the soil data any context, and I suspect this is because most report authors don’t really understand how to interpret the data in a way that is meaningful for archaeologists. We’ll say something like, “The soils within this project include Houston Black clay, 0 to 1 percent slopes; and Eddy gravelly loam, 3 to 5 percent slopes”—but what does that even mean? Why is it relevant for archaeology, and why bother putting it in a report?

Whether most report authors realize it or not, this information is extremely relevant. So let’s take a look at the soil data that have been published by the NRCS Web Soil Survey, which is free and open to the public. It can be accessed here.

If you open the website and navigate to the map page, you can zoom in to your project area on the map, then select your AOI (Area of Interest). After creating (or uploading) your AOI, you can click on the Soils Map tab, and you will see all the different soil units mapped out. Each soil unit will be labeled with some kind of code (Map Unit Symbol), and on the left side of the screen, that code will be paired with its corresponding soil type (Map Unit Name). It should look something kind of like this:

 

Figure 1: Screenshot of NRCS webpage

The Map Unit Name will include the soil series, the texture of the soil, the slope, and sometimes some other relevant information, such as how frequently the soil floods. One example would be “Houston Black clay, 0 to 1 percent slopes.” “Houston Black” is the soil series. “Clay” is the texture. Slope can vary within most soil series, so the soil scientists divide up each series into multiple units based on the slopes in each area.

If you can determine the soil series within your project area, that can usually tell you the age and/or formation of the soil—and that can tell you how deep you need to dig your shovel test, or whether there is any potential for stratified or deeply buried archaeological sites. The NRCS has another website with information about each soil series. That website can be accessed here.

If you look up the page for the Houston Black series, the very first sentence tells you how Houston Black soil formed: “in clayey residuum derived from calcareous mudstone of Cretaceous Age.” In other words, Houston Black is a residual soil that formed in situ from weathered bedrock. In this case, that bedrock is mudstone. Basically, the mudstone weathered into dark clay over time. That clay was not deposited from elsewhere (the sediment that initially formed the mudstone was transported from elsewhere, but that happened over 65 million years ago). This means, if you are shovel testing or excavating in soil that has been mapped as Houston Black, you can expect all the artifacts to be located on the surface or in the topsoil. There is little to no possibility for deeply buried artifacts, because artifacts would have been buried by bioturbation, not by the introduction of new sediment. There can still be quite a high possibility for buried artifacts if there’s a lot of vegetation, but those artifacts won’t be deeply buried. I would expect to find them in the top 20 cm.

 

Figure 2: NRCS webpage for Houston Black series

On the other hand, if your project area is located in an alluvial environment, there is a very real possibility for deeply buried artifacts. But not all alluvium is the same age, and artifacts are not likely to be deeply buried in ancient alluvium, because there has not been any deposition during the era of human occupation. If you look up the “Kian” soil series, you will see that it consists of Holocene alluvium, meaning that it consists of sediment that has been deposited in the past 11,700 years or so, and possibly continues to accumulate sediment today. But if you look up the “Burleson” series, you will find that this series consists of Pleistocene alluvium. Burleson soil can be found on upland terraces that used to be floodplains during the Ice Age, but as the rivers cut deeper and shifted course, these old floodplains stopped flooding, and new floodplains formed below them during the Holocene. Understanding the difference is key. Some archaeologists assume that all alluvium is recent in origin, but that simply isn’t true. There are still alluvial landforms left over from the Pliocene and Pleistocene, but they do not experience any deposition today.

I don’t think it’s impossible to find a deeply buried site on a Pleistocene alluvial terrace, but the site would have to be very old—as in, pre-Clovis old. I don’t think you should have to shovel test deeply into a Pleistocene terrace during a CRM survey simply out of the unlikely chance you might hit a pre-Clovis site (I certainly don’t want to, but maybe that’s out of laziness). On the other hand, you won’t find anything like that if you don’t look.

Often, the NRCS webpage for any soil series will explain that soil’s origin in the first sentence, but if it doesn’t, just scroll down to the “Geographic Setting:” section. This will tell you the “parent material” of the soil, and that will help you understand when and how the soil formed.

To help you visualize this a little better, I’m going to show you the downloaded soil data for a hypothetical survey area in Collin County, Texas. Below, you can see the satellite imagery of the project area—a patchwork of farmland and pasture between the towns of Melissa and Blue Ridge:

 

Figure 3: Satellite imagery of Area of Interest (AOI) in Collin County, TX

Now, let’s take a look at the soil units within the AOI:

 

Figure 4: NRCS Soil Units mapped within AOI in Collin County, TX

I don’t have them labeled because there are too many units, but these units are a combination of the following soil series: Altoga, Austin, Leson, Eddy, Ferris, Heiden, Houston Black, Stephen, Burleson, Lewisville, Wilson, Frio, Tinn, and Trinity.

Austin, Leson, Eddy, Ferris, Heiden, Houston Black, and Stephen soils are residual soils derived from bedrock that weathered in place. They formed from a variety of shales, mudstones, limestones, and chalks, most of which are probably Cretaceous in age, at least in this location.

Burleson, Lewisville, and Wilson soils consist of ancient alluvium. The NRCS explicitly describes Burleson and Wilson soils as being of Pleistocene age. The NRCS does not say during which epoch the Lewisville series was deposited, only that it is “ancient,” so it is probably Pleistocene as well. The Altoga soil series is a little trickier to place, because the NRCS does not say whether it is modern or ancient. What the NRCS does say is that "these gently to strongly sloping soils are on risers on stream terraces." Stream terraces are generally older than the current floodplains beneath them; often they used to be floodplains themselves, but now no longer accumulate sediment. If the Altoga series has formed on stream terraces, this likely means that the parent material is of significant age. So I have classified the Altoga series as "ancient alluvium" here, with the caveat that there may be some Holocene deposition.

Frio, Tinn, and Trinity soils consist of Holocene alluvium. These are places where you might expect to find deeply buried sites. The NRCS does not explicitly say that these series are Holocene in origin, but the Map Unit Name will often say something like “occasionally flooded” or “frequently flooded” after the soil series name. If these soils are still flooding, they are still accumulating new sediment. And we live in the Holocene, so this is obviously Holocene sediment. The NRCS doesn’t have to explicitly say that this is Holocene alluvium, you just have to use your common sense.

The map of all the soil units within the project area probably wouldn’t convey a lot of information to most people, even if I did label it. But a lot of report authors will add a soil map like that without any additional context.

Hopefully, this map will make the picture clearer:

 

Figure 5: Map of Soil Units within AOI in Collin County, TX, classified by soil formation and history

Figure 6: Legend

In the above map, I’ve classified each soil unit in the project area according to its formation. The soils here consist of residual soils, ancient alluvium, Holocene alluvium, and standing water. The gray units consist of residual soils. The light brown units represent ancient alluvial terraces, now heavily weathered and eroded. The red units represent floodplains consisting of Holocene alluvium. It might be difficult to tell from the map, but the residual soils are found on the upland areas. The Holocene floodplains are located along existing creeks. And the ancient alluvial terraces are generally located roughly alongside the current floodplains.

If you’re a field tech who just wants to know how deep to dig a shovel test, this map should make it fairly simple. In the residual soils, just dig until you reach the B horizon. Or until you hit bedrock, whichever comes first. On the ancient alluvial terraces, it should also be safe to terminate your test at the B horizon, unless you’re dead-set on finding a pre-Clovis site (honestly, I think it could happen). On the Holocene floodplains, just dig as deep as you can with a shovel.

That’s one kind of landscape you might have to survey, especially if you work in north Texas. But North America has a diverse array of landscapes, so I wanted to show another example, containing soils with very different formations and histories.

Below, you can see a satellite image of some farmland in Peoria County, Illinois, not far from where I grew up.

 

Figure 7: Satellite imagery of AOI in Peoria County, IL

And here is an unlabeled map of the soil units within that area: 

Figure 8: Map of Soil Units within AOI in Peoria County, IL

The soils within this AOI include the Rushville, Keomah, Sylvan, Dodge, Sable, Oscro, Virgil, Starks, Camden, Strawn, Rozetta, St. Charles, Fayette, Hennepin, Cresent, Radford, Jules, Beaucoup, Huntsville, Sarpy, Landes, Dorchester, Lawson, and Worthen series.

The Strawn and Hennepin series consist of glacial till. If you’ve read my initial post on geology and shovel testing, you will know that glacial till is the term for geological material that was dragged along and deposited by advancing glaciers. You will also know that finely ground silt was washed out of the melting glaciers and then picked up by the wind, forming upland loess deposits that bury the glacial till in many places. These loess deposits have not accumulated much new wind-blown sediment since the end of the Pleistocene, so most artifacts would be located at the surface or in the topsoil. The Rushville, Keomah, Sylvan, Sable, Oscro, Starks, Camden, Rozetta, St. Charles, and Fayette series soils consist of loess. The Dodge and Virgil series consist of a combination of glacial till and loess. The Cresent series consists of glacial outwash. It is basically alluvium, but very old.

And the Radford, Jules, Beaucoup, Huntsville, Sarpy, Landes, Dorchester, Lawson, and Worthen series all consist of Holocene alluvium, which can be found on the modern floodplains alongside the creeks.

Here is a map that puts this all into perspective:

 

Figure 9: Map of Soil Units within AOI in Peoria County, IL, classified by soil formation and history

Figure 10: Map Legend

The light brown units represent the upland areas where periglacial loess is lying directly on the surface. The orange units represent the slopy areas where glacial till is still exposed on the surface, unburied by loess, and the pink units represent the transitional areas that consist of both loess and glacial till. There are a couple of purple units that represent pockets of glacial outwash. The red units represent a Holocene floodplain along the banks of Senachwine Creek and one of its tributaries.

You’ll notice there is no residual soil, at least not on the surface. The bedrock is too deeply buried beneath thick layers of loess and glacial till in the upland areas, and beneath layers of alluvium in the lowland areas. There are very few places in this area in which the bedrock is naturally exposed. Unlike Texas, this landscape was transformed by the Pleistocene glaciers.

If you just want to know how deep to dig a shovel test, it’s safe to terminate your test at the B horizon if digging in ancient loess or glacial till. If digging in a Holocene floodplain, again, it’s wise to dig as deep as you can.

Fortunately, all this data is available in the field if you have a smart device with an internet connection. You can download the Soil Web app on your phone or tablet, and this can tell you whichever soil unit you’re standing on. The app won’t tell you how that soil series formed though, so you may have to look up more information on the soil series with a web browser.

 

How Not to Use the NRCS Data

Mistakes should always be a good opportunity to learn, so it stands to reason that one way to become more proficient at using and understanding the NRCS data would be to examine the work of people who are not using the data correctly, and seeing where they went wrong. There are probably many examples from which to choose, but a topic like this requires a deep dive into some minute details, and I only have the time or energy to take a deep dive into one body of work. I'm going to take a deep dive into the Texas Department of Transportation Potential Archeological Liability Maps (TxDOT PALM). By the end of this section, it will probably seem like I'm trying to cyberbully TxDOT, but that is not my intention (if I were going to cyberbully TxDOT, it would be for designing roads where I have to yield to people who are approaching from behind, but that has nothing to do with archaeology). I don't think the archaeologists or other specialists at TxDOT are any more incompetent than their counterparts at any other government agency, private company, or academic institution. I think misinformation is probably rampant throughout almost all these organizations, but I don't have the time or insider knowledge necessary to thoroughly critique all of them, so I'm just going to focus my efforts on scrutinizing the TxDOT PALM model for now.

The TxDOT PALM maps constitute a probability model designed to assist archaeologists with planning their surveys, by demarcating those areas that have a high, moderate, or low probability for surface sites or deeply buried sites, or any combination thereof (a deeply buried site would be any site in which artifacts are buried more than one meter deep). If you want to look at the model yourself, you can download the various files by district here. I can tell a lot of work went into this. The people at TxDOT who developed this model aggregated multiple datasets into a single set of maps, and though I was not involved in the model’s production, I’m pretty sure the NRCS Web Soil Survey was one of the datasets employed (along with extant water features), because the NRCS soil units seem to match up too well with some of the units in this model. Having reviewed this model carefully, I’m now going to demonstrate that its recommendations are very much at odds with the original NRCS data.

Let me start by explaining how I think the model was made. Again, I didn’t make it, so I can only speculate, based on my knowledge of GIS. But I’m pretty sure the developers used a polyline shapefile showing all the bodies of water in Texas, even small creeks and rivers. And I think they made buffers around those water features, under the assumption that most prehistoric sites would be found near water. That makes sense, I’ll give them that. Those buffers were probably amalgamated with other datasets, including the NRCS data, to create units with varying degrees of probability for locating deeply buried archaeological sites and/or surface sites.

Those map units that are closest to the bodies of running water are classified as having a high probability for deeply buried sites. That sort of makes sense. Running water deposits alluvial sediment. Alluvial sediment buries artifacts. So far, so good. But the problem is that creeks and rivers often do not have floodplains along their entire length. In many places, they cut through ancient formations without depositing any new sediment along their banks. This model is claiming a high probability for finding deeply buried archaeological sites in places where that is not physically possible—in places where the bedrock is so shallow that no artifact could be buried more than a couple feet deep. Let me show you.

Here is an example of a section of the TxDOT PALM data for TxDOT’s Dallas District. It falls within a portion of the section of Collin County that I showed you earlier:

 

Figure 11: TxDOT PALM Data Closeup in Collin County, TX

Figure 12: Map Legend

Here's a rundown of the numbers in the map legend and what they are supposed to represent:

0. Water

1. Low shallow potential, low deep potential

2. Low shallow potential, moderate deep potential

3. Low shallow potential, high deep potential

4. Moderate shallow potential, low deep potential

5. Moderate shallow potential, moderate deep potential

6. Moderate shallow potential, high deep potential

7. High shallow potential, low deep potential

8. High shallow potential, moderate deep potential

9. High shallow potential, high deep potential

The red units (Map Unit 9) represent areas that supposedly have a high probability for both deeply buried sites and surface sites. Now take a look at a map with one of those red high-probability units selected. Below the TxDOT PALM units, I’ve added the NRCS soil units:

 

Figure 13: TxDOT PALM high-probability selection overlaid above soil units

Now let’s look at a map of those NRCS soil units, with the TxDOT PALM data removed. I’ve selected one of the soil map units that was beneath the selected high-probability area. If you look to the right of the map, you’ll see that the Soil Map Symbol is HoB2, one of the symbols for “Houston Black clay.”

Figure 14: Houston Black selection

So this probability model is claiming that there is a high probability for both deeply buried sites and surface sites in Houston Black clay. Now, I think it’s right to claim there is a high probability for surface sites here. This is an upland area overlooking a body of running water. There are probably plenty of sites on the surface, or shallowly buried by biotic activity. But I can’t agree that there is a high probability for deeply buried sites here, because that is almost physically impossible. Houston Black clay does not contain any deposits of Holocene alluvium. It formed in residuum derived from weathered bedrock.

Maybe that’s a fluke. Let’s take a look at one of the other soil units beneath TxDOT’s high-probability unit:

 

Figure 15: Austin silty clay selection

Well, the Map Unit Symbol is AuB, which is code for “Austin silty clay, 1 to 3 percent slopes.” This belongs to the Austin soil series, another residual soil. According to the NRCS, in a typical profile of Austin series soil, the Cr horizon begins at about 74 cm. below surface. The Cr horizon is just a term for bedrock that is soft enough to break with a shovel. But the TxDOT PALM data claims that there is a high probability for finding sites buried deeper than one meter here. If you try to dig deeper than a meter here, you’ll be chiseling through chalky limestone, not finding an archaeological site. What about the next one:

 

Figure 16: Austin silty clay selection

That soil also belongs to the Austin soil series. The Map Unit symbol is AuC2, or “Austin silty clay, 2 to 5 percent slopes.”

Let’s take a look at another probability unit. I’ve selected a unit that is supposed to have a high probability for deeply buried sites and a moderate probability for surface sites (Map Unit 6):

 

Figure 17: TxDOT PALM model selection for area with high probability for deeply buried sites and moderate probability for surface sites

And now let’s see which soil unit lies beneath it:

 

Figure 18: Houston Black selection

Okay, this probability unit overlies Houston Black clay, which, again, cannot contain deeply buried sites.

Now let’s look at a third probability unit that supposedly has moderate probability for deeply buried sites and low probability for surface sites (Map Unit 2):

 

Figure 19: TxDOT PALM model selection for area with moderate probability for deep sites and low probability for surface sites

It matches up quite nicely with an NRCS soil unit labeled as LeC2, or "Lewisville silty clay, 3 to 5 percent slopes," which belongs to the Lewisville series (the fact that it matches up so well leads me to believe that the NRCS Web Soil Survey provided one of the datasets used in this model):

 

Figure 20: Selection of Lewisville soil series

The Lewisville series consists of alluvium, so at least technically it’s physically possible for it to contain deeply buried sites. However, The Lewisville series consists of ancient alluvium, not Holocene alluvium, so this landform has not accumulated sediment in a very long time. It would be very unlikely for anyone to find a deeply buried site here. Maybe a pre-Clovis site. In fact, all the sites here would most likely be close to the surface, yet the model claims there is a higher potential for deep sites than surface sites. In fact, I think there’s a pretty high potential for surface sites, but almost no potential for deep sites.

The TxDOT PALM model clearly contradicts the NRCS data in some big ways. Of course, who’s to say that the NRCS data are not wrong, at least in this spot? The NRCS Web Soil Survey is not perfect. It does contain inaccuracies. After all, the whole point of this soil survey is to help farmers, not to tell archaeologists about the formation and history of every square inch of land in the United States. So maybe TxDOT knows something the NRCS doesn’t.

I would consider that, except for two reasons. First of all, I’m pretty sure TxDOT used the NRCS data as one of the datasets that were amalgamated into this model. If you’re willing to use the data, the implication is that you think it’s correct, or at least good enough. If you’re using the data but your end product directly contradicts it in some fundamental ways, the implication is that you never really understood the data to begin with.

Second of all, I’ve personally been to the spot that I’m examining with these maps. I’ve walked over the creek exactly where that high-probability unit is supposed to be. There is no floodplain where a site could be deeply buried. The creek cuts through exposed bedrock. On either side of the creek, there is a very shallow layer of residual soil above the bedrock. There is a little bit of fluvial sand in some spots on the creek bed, and I guess that sand could bury a coin if you dropped it in the water. But even that sand isn’t very deep, and the water flows directly over slabs of stone in most places. My observations in the field have confirmed that the NRCS data are correct, at least in this location. Which means that the TxDOT PALM recommendations are at odds with actual scientific data and have no basis in reality.

I wouldn't go so far as to say that the people who created this model are bad archaeologists or pseudoscientists. I've heard, secondhand, that they don't really expect anyone to take the model as gospel. And I suspect that at least some of the people who built this model know that it contains a fair amount of physical impossibilities. So it may have been harsh of me to imply that the TxDOT employees who made the model never understood the NRCS data to begin with. The issue lies probably not with human ignorance but with the way that probability models are generated in ArcGIS. The process by which datasets are aggregated into a probability model is automated in ArcGIS, and the way that soil types are organized in the NRCS Web Soil Survey does not really lend itself to automated workflows. If you wanted to create a new probability model that actually takes into account whether a soil type consists of residuum or Holocene alluvium or something else entirely, you would need to make a list of each soil series that fits into each category and then select each map unit that falls within said category; there is probably no good way to automate that process. Maybe the people who built this model simply never had the time or resources necessary to make one that's actually accurate. So I can't really blame them for the product they made, but I still have to acknowledge that this product is detached from reality in some fundamental ways.

I have to wonder, how far removed can our work be from reality before we have to call it pseudoscience? In my opinion, this model has already crossed that line. I'm not going to call the TxDOT employees who made this model "pseudoscientists," but I'm going to be less charitable with those archaeologists who use this model uncritically to make decisions for them. That, I believe, is tantamount to pseudoscience. And I know that there are archaeologists who use this model uncritically, because I was a contractor for TxDOT, and some of those employees wanted me to conduct deep trenching via backhoe in places where it was not physically possible for a deeply buried site to exist, simply because that was what the model recommended. Aside from the fact that it was a waste of taxpayer money, it was also a waste of my time, and I could have used that time more effectively searching for sites in places where they could actually exist.

Correct (or at least, more correct) scientific information is readily available, it’s free, and at least some of it is fairly easy for laypeople to understand, yet this model largely contradicts it. In any other scientific discipline, if you repeatedly rely on a model that frequently advocates for probabilities that are actually physically impossible, when the data that could help you rectify those mistakes are easy to access, your peers would call you a pseudoscientist. But when archaeologists do it, other archaeologists don’t really notice.

Why don’t we notice? Archaeologists are usually all too eager to lambast laypeople as pseudoscientists if they diverge from the general archaeological consensus (and to be fair, such criticism is often deserved, especially when it comes to frauds such as Graham Hancock or Scott Wolter). But professional archaeologists frequently produce work that contradicts actual science, as in the case of the TxDOT PALM model, and rank and file archaeologists within cultural resources management seldom pay attention. Could it be that many of us don’t know nearly as much about science as we think we do, and our accusations of “pseudoscience” are simply due to deviations from accepted dogma by the accused, rather than an understanding of what actually makes those deviations pseudoscientific? Thus allowing the flimsy work of our peers and supervisors to slip through the cracks? After over a decade of conversing with fellow archaeologists and reading comments online, that’s the conclusion I’ve reached.

To be clear, that last part is just my opinion, and you don’t have to share that opinion. But over the years, it has become increasingly clear to me that cultural resources management is rife with misinformation, all the way to the top. I don’t have any particular ill will towards TxDOT (though I expect some ill will from TxDOT if anyone there ever happens to see this obscure blog post). TxDOT has competent specialists who have called me out on my own careless mistakes, and I might talk about those mistakes in this blog if they become relevant. I just wanted to use the TxDOT model as an example, to illustrate the misinformation and lack of comprehension so pervasive throughout all levels of cultural resources management—in every state, every corporation, every government agency that employs archaeologists. If no one corrects you, how will you learn from your mistakes? And how can the young and inexperienced be corrected if the people at the top don’t know nearly as much as they think they do?

Correction: An earlier version of this blog post mistakenly identified the Altoga soil series as having formed within residuum rather than alluvium. The map has been updated to reflect this correction.

Updated May 22, 2024