I grew up warshing clothes and memorizing the U.S. capital as Warshington, D.C. When I moved away from home, I realized others did the wash (you know — how the word is actually spelled) and some fiercely argued that sugary, carbonated drinks are called pop, not soda as I called it. I discovered linguistics at the same time as geography in college and was instantly fascinated — not just with quirky regional differences in English (Rhode Islanders call a milkshake a cabinet?), but also phonetics, borrowing, sociolinguistics, all of it. Bringing my two favorite fields together, I grew interested in the cartography of language maps. How do we capture and symbolize this fluctuating, vibrant, cultural thing?
When I say language maps, I’m referring to the broadest thematic bucket possible: any attempt to map some aspect of language, from pronunciation to proto languages. This map genre has unique difficulties, but it also provides examples of the challenges and responsibilities of mapping anything, from languages to lava flows, bus routes to building footprints, diseases to disasters. (I may have a thing for alliteration.)
We Can’t Get It Right.
Points, lines, polygons, and cells. That’s what we work with. While some things have easy associations (road network – line network), others have no ideal graphic conversion. Mapping language is a “do the best you can” scenario. We can’t get it perfectly right. But if perfectly right is the only acceptable outcome, then our other option is to “throw hands wildly into the air and not map at all.” I admit that is my starting point sometimes, but I eventually move past it.
Any map is a representation of reality. No matter the increasingly amazing, high-accuracy, high-resolution gizmos available, our maps are still translations. And things get lost. Maps can’t include everything exactly how it is. But there is a lot to be learned from mapping things as best we can, from mapping imperfectly. When we map language we gain understanding about the distribution of, relationships between, and change over time of our very own voices — how we communicate in and about our world. I learned that warshing things, with that pesky added R, reveals my geographic and socioeconomic roots, not a speech impediment. It would be great if I could blame my intermittent lisp on geography as well.
Our linguistic landscape is informed by continent-scale language families to neighborhood lingo and everything in between. I want to know the language spoken where I’m going as well as local speech patterns so I can try to say “hello” and “thank you” without sounding like an AI voice reading a dictionary. (They are sounding much better though. It scares me.) A variable like language will never be captured in one map or one type of map. But the same is true for less “messy,” physical environment variables like surface temperature or vegetation type.
Maps capture a specific scale, extent, and point in time. (Even dynamic maps have limited parameters.) Some things are more stable over time than others (geologic strata vs. currency exchange rates), some are more detailed and localized than others (real estate values vs. world climate regions), but rarely does one map moment wholly represent a subject. It represents a part of it. We choose which part to represent for our map’s purpose and audience. There is no one map to rule them all. (Though it is March, so I’ll mention the college basketball imperialism map for applicable fans.)
Boundaries are hard (as in difficult), but often not hard and fast (as in definitively fixed). Where does one language or dialect end or another begin? Lines worked a bit better for language past than language present. Patterns followed then-impassable obstacles like mountain ranges or deserts. But we have planes now. And the internet. Both are quite good at hopping over things (though there are still many remnants and pockets of linguistic isolation that yield unique results).
Some language maps alter labels (size, capitalization, kerning, and placement) to convey extent, prominence, and/or hierarchy (family vs. language vs. dialect). Think of how “R O C K Y M O U N T A I N S” is stretched along the range’s spine on a terrain map. Additional, traditional-boundary-avoiding methods include heat maps or interspersed boundaries with teeth-like transition zones of varying widths. As a naive student, I tried to borrow from physical geography’s ecotone concept to coin “linguatone” to describe these changeover areas. To no one’s surprise, it did not catch on.
While language maps provide an especially tricky case, boundaries appear deceptively simple in many map types. Carefully considering and setting boundaries is an important skill not to be overlooked or done hastily (in mapping or in life).
Power in Presence
I’m guessing that the first language map you saw was in an introductory geography textbook or classroom atlas. It probably showed language families, major languages, or official languages of the world, neatly showing one per place. Generalized, small scale maps are a good starting point, but they are just that — the start. If additional language maps aren’t introduced, this is the only spatial representation provided of the world’s language landscape. And it leaves a lot of voices off the map — anyone speaking something other than what is selected for the global scale — languages implied as unofficial or minor.
Mapping majority languages and/or one language per place is one choice. Another choice is to map language diversity rather than dominance, providing a very different perspective. Explore the languages most frequently spoken besides Spanish and English in New York City, or seek out the smallest speaker groups rather than the largest. Counter the typical monolingual symbology with a linguistic diversity index map quantifying multilingualism. And don’t forget the Indigenous linguistic landscape, like the Smithsonian’s Native Languages and Language Families of North America map.
When you make a map, carefully consider what (or who) you put on it and what (or who) you leave off, and why. It’s powerful to be on the map, to be visible. It’s equally powerful, in a different way, to be left off, invisible. (By the way, the U.S. doesn’t have an official language at the federal level. I’ll pause while you fact check this.)
Uncertainty Is Certain.
I once took a GIS in Natural Resources course that included a unit on uncertainty. (Shout out to Dr. P!) It was liberating to openly discuss and share not the certainty of one’s map but the uncertainty in it. As a field exercise, we attempted to map vegetation boundaries in a local conservation area. It’s not always easy to delineate where the forest ends and the field begins. (I am dangerously close to George Strait lyrics.)
There is uncertainty in all maps. Yes, I said it. Don’t @ me. (Would someone kindly confirm if I used this new linguistic construct correctly?) You can’t avoid uncertainty if you are mapping language. Your starting point is knowing you can’t be certain. But conveying uncertainty doesn’t invalidate your work. It does the opposite. It validates your expertise as the map author, in solidly knowing your data, your methods, and their application limits. Discussing uncertainty also educates the map reader to think critically about any map, particularly those that don’t provide much information about how they were put together. Language maps are inherently uncertain but still full of informational value. The two are not mutually exclusive — for language maps or any map.
We are all linguistic data points. It’s what makes language maps a relatable example for thinking about how we attempt to map things. It’s also what makes me fall down the rabbit hole of accent challenge and dialect coach videos — surely that’s not just me? But for now, I need to go do my laundry here in D.C. Yep, that’s a double warsh workaround of my Midland accent’s intrusive R. After that I’ll get back to my Japanese course on Duolingo. Learning another language does something to my brain, in a good way. And thinking about how to map language does something to my mapmaking, in a good way.