And the full survey results are in …

Following up on yesterday’s post, this morning, when few others were taking the survey so the server was fast enough so as not to be aggravating, I took the complete, 140-question version of Joshua Katz’s Dialect Quiz and Survey. And I am just as surprised by today’s result as I was by yesterday’s. Apparently I speak most like people in …

Sacramento, California.

Hey, at least this is a city I’ve actually been to. (Also mostly just to check off the capitol, as was the case with Lansing.)

But my results are completely different from yesterday:

Similarity map, based on full survey


Most similar cities (all in California)

  1. Sacramento (52.7)
  2. Santa Rosa (52.6)
  3. Fresno (52.6)
  4. Roseville (52.6)
  5. Visalia (52.5)

Least similar cities (all but one in the South)

  1. Baton Rouge, Louisiana (47.7)
  2. Pittsburgh, Pennsylvania (47.9)
  3. New Orleans, Louisiana (47.9)
  4. Metairie, Louisiana (47.9)
  5. Jackson, Mississippi (47.9)

The numbers in parentheses are “estimates of the probability that a randomly-selected person in that city would respond to a randomly-selected survey question the same way that [I] did.”

At first glance, I apparently totally talk like I’m from California, and not at all like I’m from the South. The map even has a field of “less-similar” blue stretching across the crook in North Carolina’s southern border where Charlotte, the city I consider my hometown, is found. My initial thought was remembering that, yes, I did consciously work as I was growing up not to develop a Southern accent. Nonetheless, I still thought I said some things like a North Carolinian.

But look a little closer. Yesterday, in my results to the shorter, 25-question version, the difference between my most similar city, Flint, Michigan, and my least similar city, Plymouth, Massachusetts, was pretty stark: there was an estimated 31.5% chance that a resident of Flint would respond the same way I did, whereas a resident of Plymouth was half as likely, 16.2%, to do so.

Today, the difference between my most similar city, Sacramento, and my least similar, Baton Rouge, is not nearly so stark: 52.7% versus 47.7%, respectively, a difference of a mere 5 percentage points.

Now, I’m not an expert on what these results mean, but here’s a guess: they reflect my across-the-country upbringing, and my adulthood in three different, very distinct cities, as well as my tendency to listen to language, to focus on the way people speak, and then adopt what I like the most into my own speech. I used to call a shopping cart a “buggy”, and I used to use double modals (my favorite was “might ought”). I continue to pay attention to and refine the way I speak, so give me a few years and this same survey may give me yet another, even more balanced result.

By the way, these results also reflect an incorrect answer I gave: I clicked on “traffic circus” and clicked “submit”, and then realized too late that I have never called one of those things a “traffic circus”. I was going for “traffic circle”.

I speak most like people in …

A heat map of my results in the 25-question Dialect Quiz and Survey by North Carolina State University Ph.D. candidate Joshua Katz.
A heat map of my results in the 25-question Dialect Quiz and Survey by North Carolina State University Ph.D. candidate Joshua Katz.

Flint, Michigan.

Do you know how many times I’ve been to Flint? Zero. I have never been to Flint, Michigan.

And yet, apparently, Flint is where people live who speak most like me. Runners-up include Michigan’s capital, Lansing; Fort Wayne and South Bend, Indiana; and Grand Rapids, also in Michigan.

I have been to Michigan twice: I once spent the night in Detroit for a meeting, and another time I flew in early in the morning for a meeting, drove up to Lansing to check the capitol off my list, and then headed back to Detroit to catch a train to Chicago. I believe I’ve been to Indiana three times: twice to Indianapolis and once to South Bend. My brother and I were driving across the country and we both like visiting college campuses, so an overnight sojourn in South Bend allowed us to visit the University of Notre Dame (which is a beautiful campus, by the way). South Bend also happens to be exactly halfway, as the car drives, between the Hampton Inn that used to be by Westroads Mall in Omaha, Nebraska, where we had slept the previous night, and our parents’ house in Maryland, which was our final destination.

Where do people speak least like me? Massachusetts and Florida, of course. (So does that mean people in New England and Florida speak the same way? If so, that gives Southerners who claim Florida is not actually in the South some extra ammunition.)

Conspicuously absent from any of this: Charlotte and Greensboro, North Carolina; Gilbert and Mesa, Arizona; and Tulsa, Oklahoma, the cities I actually lived in growing up. Or Salt Lake City, Washington, D.C., or New York City, the other U.S. cities I’ve lived in.

I wonder, too, what my results would look like if the survey data included both the United States and Canada. My results are awfully close to our northern border, and people frequently ask me if I’m Canadian.

So, where did I find all this out? The Dialect Quiz and Survey that’s been making its way around social media over the past few weeks. It was created by Joshua Katz, a candidate for a Ph.D. in statistics at North Carolina State University, based upon data from the Harvard Dialect Survey by Bert Vaux and Scott Golder. These results are based on the abbreviated 25-question survey; I’ll take the full 140-question survey soon and let you know if my results change. (I’ll just have to wait until there are fewer internet users taking the survey. Apparently, being a Ph.D. student, Mr. Katz could only afford what appears to the slowest server on the planet.)

How did a North Carolina born-and-raised guy like me end up like this? I don’t know. But I have this much to say: the people in Michigan and Indiana clearly speak very, very well.

Where they speak most like me

  1. Flint, Michigan (31.5)
  2. Lansing, Michigan (31.4)
  3. Fort Wayne, Indiana (30.9)
  4. South Bend, Indiana (30.8)
  5. Grand Rapids, Michigan (30.6)

And where, apparently, our speech would be mutually unintelligible

  1. Plymouth, Massachusetts (16.2)
  2. Port Saint Lucie, Florida (16.8)
  3. Boston, Massachusetts (16.9)
  4. Fort Lauderdale, Florida (17.1)
  5. Coral Springs, Florida (17.1)

The numbers in parentheses are “estimates of the probability that a randomly-selected person in that city would respond to a randomly-selected survey question the same way that [I] did.”