Leah Donnella, of NPR’s Code Switch podcast and blog, has written a nice piece on “Podunk” as slang term and place name (with a murky history):

A common implication of Podunk is that it’s a place so dreary and remote that it’s not even worth situating on a map. One of the most famous people to refer to Podunk was Mark Twain, who in 1869 wrote that a certain fact was known even “in Podunk, wherever that may be.”

But there are a couple of things that people who use the term probably don’t know. First, Podunk is the name of a few real towns. There’s a Podunk in Connecticut, one in New York, Vermont, Massachusetts. The Connecticut Podunk is well-known (OK, not that well-known) for an annual bluegrass festival. […]

The other thing people likely don’t know? Podunk was a place name long before it became a punchline. Podunk is an Algonquian word. Quick explanatory comma: Algonquian languages are a family of indigenous languages spoken from New England to Saskatchewan to the Great Plains. Those languages include Fox, Cree and Ojibwe. There are a bunch of words in English that have Algonquian roots: skunk, moose, caribou. […] But beyond its Algonquian roots, much of the linguistic history of Podunk is kind of murky. “We have no idea what the word means,” says Ives Goddard, senior linguist emeritus at the Smithsonian Institution and a leading expert on Algonquian languages. “You’ll be able to find guesses in the sources if you look around. Don’t believe any of it.” (I did, in fact, find some definitions — the most plausible being from the Nipmuc Indian Association of Connecticut’s quarterly newsletter: “Podunk or Pautunke, means ‘where you sink in mire’, a boggy place, in the Nipmuc dialect. But the Podunk called their homeplace Nowashe, ‘between’ rivers.”)

But according to Goddard, when it comes to Native American place names in the Eastern United States, a lot of what we think we know is actually misinformation. He says the standard source for these definitions is a man named William Bright, a linguist who in 2004 wrote a book called Native American Placenames of the United States. “He was a good linguist, a smart guy,” Goddard says of his colleague, who died in 2006. “But when he got to Eastern areas, there wasn’t any information.”

Rather than saying he didn’t know what certain place names meant, Goddard says, Bright cited a man named John C. Huden, who in 1962 published a book called Indian Place Names of New England. But Huden, Goddard adds, didn’t exactly have indisputable definitions himself. Huden “would look through all this amateur literature and find a [place] name, find a translation, and pick the one he liked,” Goddard explains. “And this book was considered authoritative. So if you’re looking at Bright, as I just did, he cites Huden, and then he cites like three or four people after Huden who are just copying Huden, of course, and are equally uninformed.”

That kind of thing drives me nuts. If you don’t know, just say so — don’t pick some amateur guess you happen to like!


  1. /-ənk/ is a locative suffix in Algonquin languages. So it literally means “in Pod” whatever that Pod is.

  2. I discovered last year via this site https://native-land.ca/ that I went to high school in the land of the Podunks (present-day Tolland, CT). Neighbors seem to be heavy on Poquonocks, Pocumtucs, Pequots, Pequawkets, Pawgussets, etc., so I’m guessing that a certain syllable starting with P definitely meant something quite important around there. And is it possible that “Podunk” is some kind of cognate or relative of “Pawtucket”, the town in RI?

  3. Further north, there’s the Piscataqua, flowing between New Hampshire and Maine. The name looks like good Latin for ‘fished water’, but that’s ambiguous: is the water ‘fished out, no longer supplied with fish’ or ‘well-equipped with fish’? Complete opposites.
    There are more P names further south, flowing into Chesapeake Bay. Comparing the adjacent rivers Potomac, Patuxent, and Patapsco, all accented on the second syllable, I’ve always assumed that ‘paht’ or ‘puht’ was a local Indian name meaning ‘river’ or ‘water’. The Pocomoke flows into Chesapeake Bay from the other (east) side, and tiny Poquoson Creek is further south.
    Then again, the next river south of the Potomac is the Rappahannock, followed by the Ni, the Po. and the Matta, which combine to form (what else?) the Mattaponi. Whether the North Anna and South Anna rivers, which flow into the Pamunkey (another P name), were called the Pa and the Munkey by the local Indians, I do not know, but I like to think so.

  4. Explanatory comma? Was she thinking of “parenthesis”?

  5. Stu Clayton says

    In these days of autocorrect, hurry and lack of proofreading, it’s rather extravagant to wonder “what she was thinking of”.

    “Comment” would make more sense than “comma”. Removing “Quick explanatory comma:” entirely would not make less sense.

  6. -pe- in most Algonquian languages means “hither”, encountered frequently because many Algonquian placenames are more like phrases than words.

  7. Jen in Edinburgh says

    comma, n.

    A clause or short member of a treatise or argument. Obsolete. ?

    It seems to have become unobsolete at least in certain contexts

  8. Stu Clayton says

    When “explanatory comma” needs an explanatory comma, I don’t need it. Too precioussss.

    At the link, the depths of guilt are probed:

    # I don’t recall my exact words, but I managed to mumble, “I just want to say that I’m sorry for you needing to explain to me what students of color deal with.” … In what other ways do I unknowingly force students, acquaintances and even friends to be “racial ambassadors”? Or, cause others to feel misunderstood? #

    The megalomania of guilt ! The gender-specific author is apologizing for not knowing everything. Here’s an alternative: just live and learn, and don’t make an issue out of not being Jesus.

  9. January First-of-May says

    followed by the Ni, the Po. and the Matta, which combine to form (what else?) the Mattaponi

    This particular case might well have been deliberate.

    That said, even in Europe – perhaps especially in Europe – despite the millenium (or more) of recorded history, we have no idea what many of the river names mean, and/or only have vague guesses along the lines of “it must have meant ‘river’ in some otherwise unrecorded ancient language”.
    I see no immediate reason why the same should not be the case in North America as well (but even stronger, because the contemporary local native languages aren’t that well known either).

    …Then again, Europe is also unusual for having a lot of personal names that, at least synchronically, do not appear to have any meaning (usually because they were borrowed from other languages a very long time ago). I don’t think this is much of a thing anywhere else in the world (except in areas recently overrun by Europeans).
    Maybe it’s just some kind of cultural tradition thing that just isn’t the case in the rest of the world, and those other places are fine with just using names in their own languages for the rivers as well, without borrowing the names of whoever came there earlier?

    (Fun fact I found out while looking stuff up for the above comment: the name of the Tigris has an internal Sumerian etymology. I knew that the Sumerian name of the river is the origin of all the current names, but didn’t know that it actually meant anything in Sumerian other than “the name of that river”.)

  10. David Marjanović says

    Further north, there’s the Piscataqua, flowing between New Hampshire and Maine. The name looks like good Latin for ‘fished water’, but that’s ambiguous: is the water ‘fished out, no longer supplied with fish’ or ‘well-equipped with fish’? Complete opposites.

    Good! Now add the Piscataway.

    I don’t think this is much of a thing anywhere else in the world (except in areas recently overrun by Europeans).

    It is a thing in the Islamic world – for the exact same reason.

    Also, some Mongolian names and parts of names are Tibetan for Buddhist reasons.

  11. J.W. Brewer says

    Here’s wikipedia at least admitting confusion and uncertainty regarding New Jersey toponyms: “The name Piscataway may be derived from the area’s original Native American residents, transplants from near the Piscataqua River defining the coastal border between New Hampshire and Maine, whose name derives from peske (branch) and tegwe (tidal river),[20] or alternatively from pisgeu (meaning “dark night”) and awa (“place of”)[21] or from a Lenape language word meaning “great deer”[22] or from words meaning “place of dark night”.[23] The area was appropriated in 1666 by Quakers and Baptists who had left the Puritan colony in New Hampshire.[22]”

    I think once you mention three (or four? but #4 just seems a restatement of #2 due to editing glitch …) different theories you’ve satisfactorily admitted that no one really knows.

  12. J.W. Brewer says

    Although here’s a great non-Algonquin toponym etymology:

    “The New Market section [of Piscataway Twp., N.J.] historically comprised the Quaker village of Quibbletown. The early name of the village originated from the fact that settlers of different religious denominations quibbled about whether the Sabbath should be observed on Saturday or on Sunday in the village.”

  13. ə de vivre says

    The Tigris, “Idig(i)na,” probably has Sumerian “id” (river, watercourse) as the first element. It’s tempting to identify the second part with “gina” (fixed, established), itself a loan from Akkadian “kīnum,” but I don’t know how plausible that is. (What exactly happens to Sumerian vowels in multi-syllabic words is a question we might not have the data to answer. Is a “šita” bowl the same thing as an “ešda” bowl? Maybe, but if you can figure out how exactly they’re linked, there’s wealth and fame* awaiting you in Assyriology.) And is a landscape of shifting, anastomosing waterways, I’m not sure how semantically plausible “the established/fixed/permanent river” would be.

    Wikipedia is, typically, way off in its etymology. It’s defo not “the swift river.”

    *There is no wealth or fame in Assyriology

  14. John Cowan says

    It seems to have become unobsolete

    It was never obsolete in New Testament (or probably any other Greek) criticism. The Johannine comma is a phrase (unlike Roger Casement’s comma) in John 5:7-8.

    ⁷For there are three that beare record in heaven, the Father, the Word, and the Holy Ghost: and these three are one. ⁸And there are three that beare witnesse in earth, the Spirit, and the Water, and the Blood, and these three agree in one.

    The italicized words are the only reference to the Trinity in the Bible, are not found in the oldest Greek, Latin, or Syriac manuscripts, and have been generally agreed to be a tendentious interpolation since the 19C.

  15. David Marjanović says

    *There is no wealth or fame in Assyriology

    But there might be tenure.

  16. ə de vivre says

    But there might be tenure.

    And even then…

  17. But there might be tenure.

    “A recent recollection of all tenure-track Assyriology positions in North America in the last ten years yielded just over a dozen advertised positions”

    Good luck

  18. John Cowan says

    Fine, but how do you get from Idig(i)na to Elamite Tigra? That’s dropping the initial vowel, /d/ > /t/, /n/ > /r/ in a mere five phonemes.

  19. The Elamite could only write foreign names with the cuneiform signs they had in their inventory.

    They couldn’t write initial /i/, lacked phoneme /d/, so had to use the only available syllable “ti”, etc.

    Ti-ik-ra – that’s how they wrote it.

    Don’t know why use /ra/ instead of /na/. That’s the only mystery.

  20. ə de vivre says

    The Sumerian consonants conventionally transcribed as voiced stops were actually pronounced as unaspirated voiceless stops, probably allophonically voiced intervocalically and before a voiced consonant—but the exact conditions, synchronically or diachronically, aren’t clear. That said, AFAIK voice has never been posited as the contrastive feature in the 2 stop series in Elamite either.

    It wouldn’t surprise me if the Elamite form came from the Akkadian, “Idiqlat.” L~n alternations are well-attested in Sumerian (the negative prefix “nu-” shows up as “lu-” in OB Sumerian if the next syllable starts with a “b,” and the ubiquitous “lu,” man, might be present in words for professions like “nubanda” [which I just noticed is translated as “laputtû” in Akkadian, which looks exactly like you’d expect Sumerian *labanda to show up as a loanword. I swear, there’s something going on with the distribution of “l,” “n,” and “u” in Sumerian that no one’s quite put together yet…]), and an l > r change is pretty easy to accept.

  21. Excellent; I’ve added Sumerian to the list of l~n alternations in this thread.

  22. ə de vivre says

    My pet theory is that the vowel in “Cu” signs is sometimes really the effect of a labialized consonant, which is why an n can be subject to proximity restrictions on bilabial consonants, like in “nu-” vs “labV-” alternations. It would also explain words with alternations between g and b if a there was a proto-form with *gw. Unfortunately, many details get in the way of such a satisfying explanation. If there’s a solution there, it’s much less elegant.

  23. John Cowan says

    The change can’t be due solely to the writing system, because it’s the phonetic shape /tigra/ that got into Old Persian > Greek > Latin > other IE languages. Akkadian and the other Semitic languages preserved something much closer to the Sumerian.

  24. The whole thing looks to me like a pre-existing river name that was perhaps folk-etymologised in one or both of Sumerian and Elamite.

  25. Makes sense.

  26. January First-of-May says

    The whole thing looks to me like a pre-existing river name that was perhaps folk-etymologised in one or both of Sumerian and Elamite.

    I agree that this makes sense; I should have realized that, when it comes to place names, even if it has a vaguely plausible-sounding internal etymology, this does not necessarily mean that the etymology in question is the original one.
    (As witnessed in the infamous case of York, etymologically Jorvik “horse bay”, which is actually Eoforwic “boar town”, which is in turn actually Eboracum “yew place” – and for all we know it might well have had another name even earlier.)

    The problem here is “turtles all the way down” – a priori, anything can be a folk etymology of an older borrowing, and the prevalence of such examples in recorded history means that it’s probably more likely than not that any individual nontrivial place name “etymology” is just a reinterpretation of a borrowing.
    (At least, unless we can be fairly sure that the name is definitely not older than X, which is rarely the case in Eurasia, and only works for European-origin names in the Americas, but can sometimes show up elsewhere; in particular, we can usually take Polynesian name etymologies on face value, because the Polynesians in question were the first people in the relevant place.)

  27. ə de vivre says

    The change can’t be due solely to the writing system

    I don’t think the change has anything to do with the writing system. Neither Sumerian nor Elamite had phonemic voicing, so the voiced consonants in Old Persian probably come from allophonic variation that neither Sumerians nor Elamites—who called the river something like /idikna/ and /tigra/ respectively—represented in their system. Things are made extra confusing because we transliterated cuneiform based on Akkadian values, which had a voicing contrast that seems to be unusual in the ancient mesopotamian language area.

    Another problem is that the Idigna, like most geographical names, was written logographically. I’d be curious to know when the first attestation of the name spelled syllabically comes from. I’d guess it’s from the Old Babylonian period, some 1500ish years after the logographic spelling’s first appearance.

  28. John Cowan says

    Neither Sumerian nor Elamite had phonemic voicing

    Perhaps not, but the alternations in Elamite at least are systematic, not random, so they express some feature of the language. My version of English doesn’t have phonemic voicing either, it has phonemic aspiration.

  29. ə de vivre says

    I guess I’m not sure where your confusion is, then. The Elamite form /tikra/ would have come either from Sumerian /itikna/ or via Assyrian /idik’lat/: l~r~n variation is well attested internally to Sumerian and in loanword passing in or out of the language. Then from /tikra/ to Old Persian /tigrā/. None of those steps seem particularly surprising to me.

    There was only one series of VC signs for stops in Elamite, so a spelling {ti-ik-ra} does tell us that the first consonant was from the “fortis” series (since they could have used {di}), the {ik} sign could be either series, since there is no second series of iC signs for stops. The alternations in signs with stops in Elamite aren’t random, but they’re only systematic in certain words and positions.

  30. John Cowan says

    That’s clear enough: thanks for the explication.

  31. David Marjanović says

    Eboracum “yew place”

    Rather “rowan place”, as discussed in two threads here earlier.

Speak Your Mind