I thought it would be fitting to mirror this interesting article as it was inaccessible at its original home at the blog Pfly. This blog was last updated in March 2008 and its links were not working when I tried them today. If the author of this blog wishes to contact me to take down this article, I will do so immediately. I just hope that the author puts his article at a more reliable site on the internet. (I write "the author" because his name and contact information is difficult to find through his blog.)
If you would like to view the article, you can do so at the Wayback Machine via this link: http://web.archive.org/web/20060708194402/pfly.net/?p=26
The mirror of the article "Mapping Toponomy" appears below with links to each of the maps.
Note: The only link that does not work from the original article is the link about animals entitled "GNIS animals." This image was not available at the Wayback Machine. However, I believe an image related to this discussion is contained in my next post where I mirror the author's other post about "Mapping Toponomy."
Mapping Toponomy
I’ve been making some maps depicting regionalized toponymy in the United States. The maps are below, but first, a bit of background.
About a month ago I found a book in a local library on the history of placenaming in the United States. I browsed it and liked it, and later at home looked it up and found it was out of print. But I described it to Tara, who found a lovely pristine used one and got it for me, supposedly for my birthday, although it arrived so quickly, it made for a very early birthday present.
The book is by George R. Stewart, and titled Names on the Land. I’ve since seen other people refer to it as a “classic”, and even its subtitle says The famous classic of placenaming in the United States. Although the author writes in the preface that the content was “the work of a lifetime”, the bulk of the text was put together in 1944, and the book was published in 1945. There are occassional references to World War Two as a current event. Second and third editions were published in 1958 and 1967, with additional chapters on Alaska, Hawaii, and a chapter called Current Affairs 1944-1958.
Today, it is out of print, but as far as I can tell, remains the best, and maybe only book on the history of placenaming in the United States. The subject of placenames is sometimes called “toponymy”, and until finding this book, I’d never heard of such a thing. There are books that list placenames in dictionary-like format, with brief descriptions of how places came to have the names they do. But this book by Stewart is more of a history of toponymy, and not a dictionary at all. It is also delightfully written, and fun to read. The chatpers are in a general chronological historical order, and have names like VI. How the Massachusetts General Court dealt with names, XIII. How the names became more English and less English, XXIV. How they took over the French names, XXVIII. Of patterns for street-names, XXXV. How Congress took over, XXXVII. “Change the name of Arkansas — Never!”.
A topic that comes up many times in this book is how regional differences in placenaming developed. I already had the vague sense that terms like “hollow” and “cove” for mountain valleys was more common in the Appalachians and the Ozarks, while “gulch” seemed more of a western term. This book had a few maps of the northeast US, with dots for streams named brook, creek, run, prong, and branch, as well as towns named -burg or -ville. There was also a map showing how the term “coulee” occurs almost only in Louisiana, central Washington, Montana, and along the Mississippi River in southern Wisconsin and Minnesota. In each of these regions, coulee means something slightly different.
These maps were interesting, and it occured to me that it would be easy to make similar maps. Back when this book was published, it probably took a lot of work to make the dot maps, but with today, with GIS, it would be easy. So I set to it.
The USGS manages a database of placenames called GNIS (Geographic Names Information System). At work I had seen the GNIS data, clipped to just Pierce County, working as a regular GIS map layer. I went to the USGS GNIS website and found the data downloadable by state. It was a simple thing to download all the states (I decided not to use the Alaska or Hawaii data), and merge them into a single dataset that could be queried and mapped in GIS. Actually, I ended up merging it into a few regional datasets rather than a single one because of the size of the database. The full GNIS database for the 48 states adds up to many gigabytes, and I think the largest GIS shapefile you can make in MS Windows is 2 gigabytes. But in any case, the process of downloading and merging the data was simple, but slow due to the sheer size of the datasets. I ran the processes while doing other work over the course of a week or so.
When it was done, I began to make maps similar to the dot maps I saw in the book. I made minimal efforts to make the maps pretty, and they are not always self-explanatory. Maybe someday I’ll polish them up.
The first one I made shows streams with names that end either in “Brook” or “Creek”. I was surprised to read in this book that “brook” is used in New England in the way “creek” is used almost everywhere else. The regionalization between brook and creek is the most pronounced of any of the maps I’ve made yet (click on these maps to see full-sized versions):
brook creekThe GNIS placenames are categorized into types like basin, lake, locale, ridge, stream, summit, etc. This brook-creek map shows only stream placenames. So a town named Oakbrook, for example, wasn’t counted. I figured there are probably quite a few non-stream placenames that end in Brook or Creek, and they would just be noise in this case. Also, streams that end in other words, like River or Bayou, are not shown. That is why southern Louisiana is mostly blank. In that area, streams are not called Brook or Creek, but rather Bayou, Coulee, River, and other things.
I like how this map shows the use of Brook in New England so clearly. Note, in all these maps, if colors overlap, they tend to show as black. So, in this map, a black dot in an orange “creek” area is a brook, and vice versa. So you can see there are a few brooks scattered across the country, and a few creeks in New England. The transition is fairly abrupt. Northern New Jersey is mostly brooks, but southern NJ is mostly creeks. In New York, the Catskills and Adirondacks are mostly brooks, but the Mohawk and Hudson valleys in between are mostly creeks. The western part of New York and northern Pennsylvania show a mixture, although creeks seem the norm.
In Names on the Land, this pattern is explained. I won’t go into details here, except to say that neither brook nor creek is the standard term for “small river” in all of England. Brook is apparently common in southeast England, and was brought to New England by the early colonists. A creek in England typically refers not to a small river at all, but rather a small tidal wash or mud flat. The early colonists of Virginia, who first encountered a vast “tidewater” region, used the word river for large tidal inlets, and creek for smaller tidal inlets. As they explored inland, the terms stuck and were applied to streams. Many other terms were used, and I might get around to mapping some of them, like run, branch, prong, fork, race, and stream.
Another New England regionalism I read about is the use of Lake and Pond. In most of the country, the word pond is reserved for bodies of water so small as to hardly require a name at all. But in New England, ponds are often fairly large. The example I am most familar with is the comparison of Boston’s Jamaica Pond to Seattle’s Green Lake. The two are about the same size. Jamaica Pond may be even be larger. Both are a mile or two around, and used for walks and jogging. So here is the map. It shows only placenames of the “lake” type, whose names end either in Lake or Pond (and to be clear, the GNIS names put the ‘generic’ part last, so something like Lake Tahoe is recorded as Tahoe, Lake):
The regionalism is not nearly as obvious for lake and pond as it is for creek and brook. In part I think this is because the word pond is used fairly frequently in much of the country, but mainly for truly small lakes that nonetheless are named. Also, New England is not nearly as solid in using pond as it is in using brook. I wonder if the word Lake became more common in New England over time. Many “lakes” today are actually reservoirs created by dams built recently. Maybe recently made reservoirs in New England tend to be named Lakes?
There is a curious strip on the northeastern border of Maine where the use of Lake dominates over Pond. I don’t know why that is. Perhaps it is due to French influence? I think the word “lake” was a common term among the French colonists from early times, while the English brought “pond” with them. The Great Lakes were first explored and named by the French, in the French style of putting the generic first, as in Lake Ontatio and Lake Michigan. The English colonists, as they explored deeper into the continent, encountered the French and their placenames, and often used the French placenames, spelling the French “Lac” as “Lake”. But the English mostly used the normal English word order, with the generic term second, as in Green Lake. But they kept the French order for the Great Lakes, and in time the custom arose of naming large or significant lakes in the French style. So in Seattle we have Green Lake for a smaller lake, but Lake Washington for a larger one. Minnesota is full of lakes called Round Lake, Long Lake, Clear Lake, and Muddy Lake. But in California you find Lake Tahoe. And the large reservoir behind Hoover Dam is called Lake Mead.
There are many names for “small valley”, often with distinctively regional flavors, as with hollow, cove, gulch, hole, dale, etc. I wanted to map “hollows”, and thinking of them as a generally eastern term, decided to also map the typically western term “gulch”. In this case, I mapped all types of placenames, since hollows and gulches might be typed as “basins” or “valleys” or a number of other GNIS categories. I figured not too many non-valleys would end with Hollow or Gulch, although some do, like the town of Sleepy Hollow. So this map is not quite as “pure” as the previous ones, but it shows some distinct and interesting regional patterns:
The first thing that jumped out at me was that my guess of hollow being an eastern term and gulch being a western one was basically correct. Except for Utah, where there is a large swath of hollows. I’m not sure why that is, although the Utah hollows area seems to conform quite well with the main old Mormon settlement region, which extended from southwestern Utah up through the center of the state and north into eastern Idaho. So I wonder whether the Mormons brought the term “hollow” with them and named things accordingly. A similar thing might explain the smaller region of hollows in Oregon. I’m not sure where the term gulch came from, but it definitely dominates in areas I think of as originally mining-oriented: Colorado, northern California, western Montana, northern Idaho, and the Black Hills.
Another term I was interested in was Coulee. The term is used in central Washington state to refer to canyons that don’t have streams in them, or just tiny trickles. Much of central Washington was scoured by huge floods at the end of the last Ice Age. Large lakes repeatedly formed behind ice dams, only to break out catastrophically, flooding over the land and actually carving out huge canyons, which today are often called coulees. The scoured lands in general are called “The Channeled Scabland”.
So when I saw that the word Coulee is used in Louisiana as well, I found it odd. Obviously it meant something different down there. So I made a map of placenames (of any type) that end in Coulee. Since I was making maps showing two types of placenames at once, I also mapped placenames that end in Bayou:
You can see the four regions where Coulee is used (with the Louisiana usage a bit obscured by the Bayou blue).
There are many different endings for town names, and I’ve been thinking about trying to map some of them. The only one I’ve tried so far is placenames of type “populated place” (that is, cities, towns, villages, hamlets, etc.) that end either in -ville or -burg:
This map is not as interesting as the others. There is some regionalized patterns, especially with -burg in the eastern “middle state” area of Pennsylvania and the Ohio River area. But generally, the distribution seems pretty mixed.
Finally, I thought of animals that occur only in certain parts of the country, and decided to make a map showing any placename that includes one of the four words: Salmon, Possum, Alligator, and Moose. The result is fairly predictable, and surprisingly sparse:
That is all for now. There are many terms that might make more maps along these lines. A list I made includes things like: Slough, -borough, -kill, Santa-, San-, Cove, Dale, Mesa, Butte, Delaware, Seneca, Cherokee, Apple, Yucca, Whiskey, Glade, Bottom, Lick, swamp-marsh-bog-fen-pocosin, and others.
Quick update
(Oct 27, 2005)
Thanks to Nathaniel Vaughn Kelso’s comment, I made a couple new versions of maps showing streams called Run, Branch, and Brook. The regionalization of these terms is strikingly distinct. When I get a chance I’ll polish some of these maps up, but for now, here’s two maps on Run, Branch, and Brook (click on them for fullsize):
This entry was posted on Sunday, October 23rd, 2005 at 11:59 am and is filed under General, Geography. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
No comments:
Post a Comment