by Dan Tilque
Word Ways, 2001
Articles about American placenames in Word Ways have generally reflected their great diversity. But what about their commonality? There's lots of duplication among the names of American places. Isn't it time that we had an article appreciating that fact?
Two of the most often asked questions about placenames in the US are "What is the most com-mon name for towns in the United States?" and "What placename occurs in the most number of states?" But before answering those questions, let's do a little Q&A to dispel a few myths and misunderstandings about this subject.
A Little Q&A
Q1. Aren't the answers to those two questions the same? After all, having two towns in the same state with the same name is not allowed, right?
A1. Not allowed by whom? The Postal Service? Places with post offices only. The states? For incorporated towns, otherwise not. Other than that, no one seems to care. So unless you just want to count post offices or incorporated places, there's no such restriction. Most named places are unincorporated and post officeless (they get their mail from another town) and duplications of these names within a state is as common as grass.
Q2. It's Springfield. Everyone knows it's Springfield. Why are you wasting our time answering a question everyone knows the answer to?
A2. As it all too often turns out, what everyone knows is wrong. Springfield is not even in the top ten most common names, either for total names or by state. Just count the places with more than 5,000 people and you'll find that there are 13 Clintons and only 11 Springfields.
Q3. But TV uses Springfield as the generic Anytown USA, as in The Simpsons. Surely that must mean something?
A3. It means that TV producers have bought into the myth of Springfield. That is, to the extent that it's used, which is not very much. I could only find three shows that are located in Springfield. Besides The Simpsons, there was Father Knows Best and the soap opera Guiding Light. But three is quite a few for an anytown; I doubt any other generic name is used in more than one show.
Q4. OK, it's not Springfield, so it must be Washington. Doesn't every state have a place named in honor of the first president?
A4. Yeah, like they put that requirement in the Constitution or something. Sorry, but while it is more common that Springfield, Washington is still not top 10 material.
Q5. But what if you check multiple references? Could you then find a Washington in every state?
A5. Been there, done that, have the Word Ways article to prove it. See my statename chain article in the May 1999 issue, where I found a Washington in 42 states in several references. Some of those Washingtons no longer exist, of course. For example, the one in Oregon was a post office that closed 140 years ago. Counting these old names certainly moves it up the list, but even that doesn't put it at the top. And to be fair, we would also have to count old occurrences of other contenders, which I'll let someone else do if they have the inclination.
Search for Candidates
So what are the answers? Obviously, I was going to have to get out and do some counting to find out. And naturally I turned to the largest gazetteer available. It turns out that that is our good friend GNIS, which is more verbosely known as the US Geologic Survey's Geographic Names Information Service. It's on the Internet at http://mapping.usgs.gov/www/gnis/gnisform.html.
But what names should I count? Unfortunately, GNIS cannot be queried for the name with the most entries. Nor can it be browsed like a book. I had to browse some dead tree gazetteers such as Rand McNally's Commercial Atlas and Marketing Guide to get the candidates. I generated some 50 or 60 fairly common names and when I arrived at the point where new candidates were consistently getting smaller counts, I figured that I'd probably got all the most common names.
Setting the Criteria
GNIS has several kinds of names. For some places, it has one or more variant names (usually former names) that can be optionally included in the query. Plus some names have parenthesized labels indicating that they are subdivisions or historical (i.e., no longer exist). And then there's the variant spellings issue. For example, names that are pronounced the same as Greenville could be spelled several different ways. Should all of them be counted as a Greenville? There are people who would consider some, none, or all of these variations as being the same as the unmodified name. What should I do?
After much ruminating, I decided to count the names two ways, a basic count and an extended count. The basic count would have just those that have the same current name. Spelling variations, alternate names, subdivisions, and historical places would be right out.
The extended count would include alternate names, subdivisions and historical places. Only one kind of spelling variation would be counted: where the name was a compound or two-word name, an internal space or lack thereof would not prejudice me against it. So a Greeneville would not be counted as a Greenville, but a Green Ville would. Also, an Oakgrove would count as an Oak Grove. Since I was counting both alternate names and spelling variations, I had to be careful not to count a place twice (e.g., when a Greenville had the alternate name of Green Ville or vice versa).
Multiplicity and a Potential Problem
Before getting to the actual results, I want to note some extreme cases of multiple names within a state. Here are all those where the extended count was at least 20:
Five Points PA 32
Midway TN 25
Mount Pleasant PA 23
Fairview TN 29
Midway TX 24
Centerville PA 22
Oak Grove TN 29
Five Points GA 23
Fairview AL 22
But even more amazing is that 6 of those 32 Pennsylvania Five Points are in Mercer County! It's quite possible that some of those 6 entries are bogus. There are some indications, which are too involved to go into here, that there are actually only 2 or 3 actual places represented by those 6 entries in the database.
While that was the most egregious potential problem of that sort, it wasn't the only one. But I didn't keep track of them and I've already put more time into this article than I want to, so I'm going to leave this issue alone. For this article, every entry in GNIS will be considered accurate and eligible to be counted.
So what did I find? Was some name the runaway winner in both categories for both counts? That'd be nice. Then we could put this question to bed for all time.
So much for wishful thinking; the reality is much more ambiguous: The first number gives the basic count and the second the extended count.
Oak Grove 160,196
Five Points 147,163
Mount Pleasant 112,153
Pleasant Hill 120,151
New Hope 106,128
Pleasant Valley 90,126
Well, this is a fine fix we find ourselves in. The basic and extended counts don't agree. Is the answer Fairview or Midway? I suppose it depend on your personal preferences. Are you a no-nonsense, meat-and-potatoes kind of person? Do you want the ultra-certainty of solidly, unquestionably named places? Then you're squarely in the Midway camp. But those who are willing to accept the less certain, to live on the ragged edge, to go for the gusto, will find them-selves backing Fairview. (But don't ask me; I'm quite willing to live with ambiguity.)
By the way, the extended count for Springfield was 85 places in 35 states, which puts it quite a ways below Salem on the list.
Now for the top placenames by number of states:
Riverside 46,46 (no AK HI LA OK)
Centerville 43,45 (no AK HI NH ND WY)
Fairview 39,43 (no AK CT HI ME MN RI VT)
Franklin 39,42 (no AK CO DE HI NV ND RI WY)
Midway 39,40 (no AK HI ME MA MI NH NJ RI VT WY)
Fairfield 36,39 (no AK AZ CO HI KS NV NH RI SC SD WY)
Pleasant Valley 35,39 (no AZ CO DE FL HI ME MS NH RI SD WY)
Troy 32,39 (no AK CO CT DE FL HI MA RI UT WA WY)
Liberty 35,38 (no AK CT HIU MA MN MT ND NH NJ SD VT WY)
Union 36,38 (no AK AZ HI ID KS MD MA NM RI SD VT WY)
No doubts about this one. Riverside has the lead, fair and square, and Centerville is in a solid second place. But Hawaii has none of these names, and Alaska only Pleasant Valley, making a 50-state sweep very unlikely. (I'll cheer from the sidelines while you search old atlases.)
Back to Word Ways articles
Back to Word Ways home