Everyone is fascinated by words like disgruntled and unkempt that have no corresponding unprefixed forms; aldiboronti, in a Wordorigins thread on the subject, links to an excellent World Wide Words post that goes into the history of the best-known such words with admirable thoroughness. On unkempt, for instance:

The word unkempt has a complicated history. Kempt comes from the Old English word kemb, “comb”. It seems to have gone out of use about 1600 but to have been reintroduced about 1860. Its usual and literal negative form was unkembed which survived into the middle of the nineteenth century. The form unkempt began to be used about 1580 to mean “language that was inelegant or unrefined”. In the eighteenth century it came to mean specifically “uncombed; dishevelled”, perhaps influenced by the Flemish equivalent ongekempt, and was used alongside the older form for about a century, only taking on a stronger sense of “neglected; not cared for” in the middle of the nineteenth century. Incidentally, the root form of kemb seems to come from a Germanic form which meant “tooth”, so a comb is named for its teeth; the modern form uncombed appeared about 1560.


  1. Doesn’t it seem a bit odd that the metaphoric use of unkempt predates the literal one?

  2. I did a series of Google and concluded that “do not condone” and comparable negatives outnumber positive uses of the word about 10,000 to 1. If you google the phrase “I condone” you will find that it’s almost always in phrases like “nor do I condone” or “they claim that I condone” etc. where the speaker is denying that he condones this or that.
    Perhaps that’s the last taboo: condoning. Perhaps this very minute some despised pariah somewhere is furtively condoning things, feeling all alone in the world and praying he won’t get caught.

  3. Could have sworn I’d seen well-kempt, and seen it often.

  4. It’s all the rage among the couth.

  5. PF: Google has seen “well-kempt” about 1,280 times, and the top results look plausible specimens, but then there are about 11,800 hits for “gruntled” and 686 for “molished”.
    And it turns out that “membered” (cf. dismembered) really is used in the sense of having limbs.
    Incidentally, the usual term in linguistics for a morpheme which looks like it should be able to stand on its own but can’t is a cranberry morph, since un-as-with strawberries, blueberries, raspberries, or gooseberries, there’s no such thing as a cran. (I searched and this doesn’t seem to have been mentioned on Language Hat; apologies if everyone knew all along.)
    It is a pretty name, I think, and a pretty thing to study (although I don’t know the scholarly literature) since there’s interaction of lexification (a compound with a real word out living its real word) and analogy (all those fakey-fake loans, turned into things that merely look like they decompose). And besides, I like morphology.

  6. Des — the folks at Ocean Spray might disagree on the independent existence of “cran”…

  7. Well cran me senseless, so they do!

  8. So does the OED. Not only is it a Scottish word for the swift, it’s:
    A measure of capacity for fresh herrings as caught; fixed by the Fisheries Board at 37 1/2 gallons (about 750 fish). Up to 1815 the cran was measured by heaping full a herring-barrel with the ends taken out, which was then lifted, leaving the heap on the ground or floor. In 1816, the Commissioners for the Herring Fishery fixed the capacity of the ‘cran’ at 42 gallons, Old Wine Measure, which in 1832 was raised to 45 gallons, 42 gallons when ‘pined’ being found insufficient to make a barrel of bung-packed herrings. In 1852 the contents were given in Imperial measure as 37 1/2 gals., making, when pined, a barrel of 30 gals.
    1797 Statist. Acc. Scotl. XIX. 282 (Lewis) They.. bought the herring.. at the great price of from 9s. to 12s. per crane (which is the full of a barrel of green fish as taken out of the net. 1815 Act 55 Geo. III, c. 94 §13 If.. any cran or measure not so marked.. shall be made use of.. in the British herring fishery. 1852 Board of Fisheries Notice (May 15), The Commissioners for the Herring fishery.. do hereby give notice that from and after the date hereof, the Cran shall be of the contents or capacity of Thirty-seven Gallons and One Half Gallon Imperial Standard Measure.. That it shall be made of Oak Staves.. that it shall be bound with Six good Iron hoops, etc. 1870 Pall Mall G. 3 Sept. 4 The catch for the season is now nearly 90,000 crans.
    That seems a perfectly obvious source for cranberry, and I shall henceforth refer to the fruit as the “herringberry.”

  9. Uh, as a New Englander I feel compelled to respond to that, LH. The cranberry is indeed the “craneberry” but not in the sense of a measure of herring. Rather, it’s because the tall flower stem of the cranberry with its hooked blossom looks like the long neck and beaky head of the crane. That’s why “cranberry morpheme” is a poor choice of terms for the phenomenon. Allowing for spelling, “Cranberry” is a compound just like “blueberry” or “gooseberry,” and it’s separable in exactly the same way. So “herringberry” isn’t going to fly, LH, although for a consolation prize may I offer you this nice basket of shadberries?
    A better term than “cranberry morpheme” might be something like “helicopter morpheme” since words for things associated with helicopters (helipad, etc.) get constructed as if the breakdown of the word were “heli-copter” when really it’s “helico-pter” or whirlybird.
    Hmm, that’s still not an ideal example, since it’s mostly the result of the lack of initial “pt” in English, and not by analogy to other “copter” words (as with the berries). Any better suggestions out there?

  10. Bilberry? OED etymology: “[App. of Norse origin; cf. Da. böllebær, f. bölle (used separately for bilberry) + bær BERRY. (The origin of Da. bölle is unknown; the suggestion that it is:ON. bllr BALL is phonetically improbable, since this gives Sw. boll, Da. bold.)] ”
    But “cranberry morph(eme)” has long since stuck, and they’re a pain to get off when that happens…

  11. But… I liked “herringberry”!
    *eats shadberries, feels better*
    Yeah, I don’t think we’re going to change the name. But we can mutter resentfully.

  12. Remember the Naugahyde ads with the cute little naugas running around?
    Eliot uses “maculate” somewhere. Probabaly a good English word, but much rarer than “immaculate”.

  13. So when I ask people “Have you no couth?” or talk about regruntling someone who is cranky, I’m out of line?
    I thought that most people played with words and said things like “gruntled” and “embowelled”
    I guess that explains all those funny I’ve been getting.

  14. Zizka, I don’t recall the ad with the naugas running around, but in Maine we always assumed that they were cultivated on the island of Vinalhaven.

  15. Michelle: There are two kinds of people, those who play with words and those who don’t. The latter give us funny looks.
    zizka: It is a good English word: “Thy rare greene eye… never yet/ Beheld things maculate” (Two Noble Kinsmen v. iii); “Foul walls and maculate table linen” (Stevenson); Elizabeth Bishop (“Song For The Rainy Season”):
    …by the warm touch
    of the warm breath,
    maculate, cherished;
    But the OED says “Now chiefly lit. and poet., in expressed or implied antithesis to immaculate.” So there you have it.

  16. In the eighteenth century [unkempt] came to mean specifically “uncombed; dishevelled
    clever . . .

  17. “selcouth” (seldom-seen) is one obsolete word i have tried to revive.

  18. So as not to become as one selcouth in these pages, I’ll weigh in with a gust of mostly air, an attempt at an explication of the common unkempt and their selcouth unpair.
    One remains unkempt by not doing kempt-ish things; and in a time when combs dragged little nits and other vermin from the head, which were communicable, it was also more urgently socially desirable. People knew intimately why it bothered them to see the bedraggled come their way.
    Once having combed one becomes, not so much combed, as whole, how one should be, and the nominative feels incomplete to the tongue. One could become particularly “kempt” in the sense of overly fastidious I suppose, and I suppose some did, but for the most of the them that were us in those times it was a lot to do merely to get a little clean now and again.

  19. Though it doesn’t have as nice of a ring to it, I would propose “hamburger morpheme” in cranberry’s place. “Hamburger” was originally deconstructed “hamburg-er” and meant “that thing from the German city of Hamburg”. It got reanalyzed as “ham-burger”, and as a result, we have “veggie burger”, “fish burger”, “bacon chili cheesburger”, etc.

  20. I don’t really see any problem with “cranberry”. Yes, it’s etymologically crane-berry, but synchronically the great majority of English speakers don’t recognize “cran” as an allomorph of “crane”. It’s not just a matter of spelling, either – they’re pronounced differently*.
    “Ham-” in “hamburger” can certainly be seen as a kind of cranberry morph, but I don’t think it’s a good example to take as a prototype. Firstly, it’s possible to analyse it in the original sense of “hamburg-er” without arcane** diachronic information. Secondly, “ham-” is homophonous with the common word “ham”, and easily confused with it. Certainly as a child I imagined that hamburgers were somehow made from ham. I don’t know if this disqualifies it as a cranberry morph, but certainly it’s a particular type, and I don’t think it makes a good generic example.
    “-ert” in “inert” is a good one in principle, but obviously we couldn’t go around calling them “inert morphemes” – too much potential for ambiguity. Really, I think cranberry’s the best we’re going to get.
    * By me, and, I assume without any substantial evidence, by most people. Krann vs kreyn (I’d put it in SAMPA, but that tends to make things a little too dialect-specific).
    ** If our host didn’t know it, then I think it qualifies as arcane.

  21. Yeah, “cranberry” is as good a word as any. (And I was kidding about my proposed etymology — I just liked the bizarrerie of those bung-packed herrings.)

  22. I thought “-burger” was the cranberry morph; not “ham-“. Ham is certainly a valid English word, although I am unsure of its productive properties. On the other hand, “-burger” couldn’t exist as a morpheme in English prior to the lexicalization and reanalysis of “hamburger”, so I thought it was a clearer example than “cran”, which turns out to be a variant of “crane”. At any rate, maybe this confusion about “ham-” is a fair indicator that we are better off with “cranberry morph”, as many have stated.

  23. Now I’m not sure whether I’m understanding the concept of the cranberry morph correctly. I thought it was a meaningless morpheme that was assumed to have meaning by analogy to similar-sounding forms — in other words, leaving aside the actual etymology of the word, the key to the cranberry morph is that “there’s no such thing as a cran.”
    But with hamburger, you have a word hamburg-er, that is reconstituted as ham-burger, broken into two halves that already exist in the language (note that the English word “burger” aka “burgher” goes back at least as far as Shakespeare and Marlowe, so the morpheme itself is not new, but only its definition as an edible patty). This seems like a different phenomenon entirely. Many of us wondered, as kids, why hamburgers weren’t made of ham; but while we might have thought “hmm, what does ‘cran’ mean?” we don’t seem to have generated a folk etymology, and might not even have noticed if it weren’t for constructions like cran-raspberry or cran-grape.
    So is the definition of a cranberry morph simply a morpheme which is assumed to have meaning by analogy to other forms (like the cran in cranberry — whose only apparent meaning in present usage is “short for cranberry”), or is it a morpheme which is created by splitting a compound word in the wrong place (like ham-burger or heli-copter) and then ascribed meaning through later usage?

  24. My understanding of the term “cranberry morph” is that it’s a bound morpheme which occurs only in a single morphologically complex lexical item, and doesn’t have any defined meaning outside that context. E.g. “cran-“. Or “-kempt”. This seems to tally with what I’ve found online, although I don’t have access to a linguistics dictionary at this time… no, wait, there’s an entry in the Lexicon of Linguistics.

    Cranberry morpheme
    MORPHOLOGY: a type of bound morpheme that cannot be assigned a meaning nor a grammatical function, but nonetheless serves to distinguish one word from the other. EXAMPLE: the English word cranberry seems morphologically complex, since it must be distinguished from words such as raspberry, blackberry, and gooseberry. Still, cran has no meaning and does not function as an independent word: cranberry is the only word in which cran appears. The existence of cranberry-morphemes plays a role in the discussion whether morphology is word based or morpheme based (e.g. Aronoff 1976).

    “Burger” in the sense of a ground food patty (I’m inclined to see it as a different morpheme to “burgher”, although I guess they’re both homophonous and cognate) clearly isn’t a cranberry morph in this sense – it’s highly productive and has a well defined meaning. If we were to analyse “hamburger” purely in terms of this new morpheme which it has given birth to, and consider the morpheme “ham” (which such an approach would require) to be different from the regular word “ham” – if we were to persue such an analysis, then this “ham” would be a kind of cranberry morph, is what I was getting at above.
    Morphemes formed, like “burger”, by reanalysis of boundaries are themselves rather interesting, and I don’t know if they have a name. Possibly “burger morpheme” or “burger reanalysis” could be pressed into service here. A perhaps even more interesting example of this type is “-oholic”, meaning “addicted to”, which is entirely bound, and has been likened to the lexical suffixes which typify languages of the American northwest.

  25. Ah, that would be what I neglected to consider; “burger” isn’t a cranberry morpheme because it exists as an independent word, is that correct? Thank you for the clarification. That does raise the question: if “cran” somehow managed to get lexicalized as a standalone word, say an acceptable abbreviation of cranberry, would cranberry cease to be a cranberry morpheme?

  26. People misinterpreted the “in-” in “inflammable” on analogy with “inedible”, etc., so the word “flammable” was invented for safety’s sake. Or so I’ve been told.
    Since my youth I’ve wondered why so many Europeans (Hamburgers, Frankfurters, Wieners, Danish, etc.) called themselves by the names of snack foods.

  27. That does raise the question: if “cran” somehow managed to get lexicalized as a standalone word, say an acceptable abbreviation of cranberry, would cranberry cease to be a cranberry morpheme?
    I suppose it would. My intuition is that for “cran” to be seen as a word in its own right, rather than a conscious abbreviation of “cranberry”, it would have to displace it almost entirely, to the extent that “cran-berry” sounded redundant, like “apple-fruit”.
    Incidentally, being unable to stand as an independent word is a necessary, but not sufficient, condition for cranberry-morph-hood. If “cran-” appeared in other words, with some kind of common meaning, then it would be a normal bound morpheme, like “un-” or “pre-” or the recent coinage “-olholic”. What makes it a cranberry morph is that it only occurs in this one compound, and doesn’t mean anything in its own right.

  28. In a bizarre coincidence, the word cran (here spelled crann) is used as an example of the “multiple meanings Gaelic is prone to” in a review of Colin Mark’s new Gaelic-English Dictionary in the TLS of Feb. 13 (the same date, note, that the cran discussion started on this thread). The reviewer, Victor Price, says, “Take the example of the word crann (basically a stake or pole). With an adjective or a noun in the genitive attached, it can transform itself into a bolt, a bar, a vessel, many kinds of trees, a plough, a measure for herrings, even the male member.” Two completely independent cran sightings on the same day? I think I’m going to drop a note to Rupert Sheldrake.

  29. Could “Cranberry Morpheme” be considered a “red herring”, then?

    On a separate note, Would a “tame goose-chase” be the opposite of a “wild” one, in a figurative sense? (“The crook didn’t even bother to cover his tracks- he lead us right to him. It was a real tame goose-chase.”
    But, could you figuratively SEND someone on a “tame goose-chase” and, if so, how would you go about doing so?

  30. David Marjanović says

    A perhaps even more interesting example of this type is “-oholic”, meaning “addicted to”, which is entirely bound, and has been likened to the lexical suffixes which typify languages of the American northwest.

    -gate for scandals and -ghazi for political manufactroversies.

  31. Also -athon for long-duration events.

  32. cranberry


  33. Tim May: A perhaps even more interesting example of this type is “-oholic”, meaning “addicted to”, which is entirely bound, and has been likened to the lexical suffixes which typify languages of the American northwest.

    “Lexical suffixes” do not occur in all the languages of that region, but particularly in the Salishan and perhaps Wakashan languages. They are mostly variants (or sometimes different words) of actual nouns referring to body parts and other common semantic fields, incorporated into the verb. While -oholic appears to be added to such English nouns or parts thereof, with alcoholic or chocoholic it alc- and choc- which are the equivalent of the “lexical suffixes”, not -oholic which behaves as the root or head of the new compound noun. Compare chocoholic with meat-eater: in both compound nouns the head comes second, which is the rule in the vast majority of English compound nouns (a cran-berry is a berry not a cran(e), a meat-eater is an eater not a meat ). It just happens that the formant* -oholic, a result of incorrect analysis of alcoholic, has been interpreted as meaning ‘addict’ and given the head noun role and but has not (yet ?) become totally separate from its origin. Incidentally, alc- itself has also acquired some independence as the root or stem for the noun alkie.

    * A formant is a part of a word that cannot (or not yet) be securely identified in terms of meaning and/or function. This often happens when dealing with very old words or those of foreign origin, or in the early stages of trying to analyze a language. Thus in alcoholic, the last segment -ic is identifiable as an English suffix added to the noun alcohol, but with -oholic, even if one can separates -ic from -ohol- the leftover sequence has no meaning or function on its own.

  34. Oops: with alcoholic or chocoholic it IS alc- and choc- which are…

  35. George Gibbard says

    I used to suppose that those sticks of colored wax for drawing with were “crans”.

  36. David Marjanović says

    Long, long ago in a thread very, very near…

    A better term than “cranberry morpheme” might be something like “helicopter morpheme” since words for things associated with helicopters (helipad, etc.) get constructed as if the breakdown of the word were “heli-copter”

    Much later also roflcopter.

  37. I see no one picked this up:

    “Take the example of the word crann (basically a stake or pole). With an adjective or a noun in the genitive attached, it can transform itself into a bolt, a bar, a vessel, many kinds of trees, a plough, a measure for herrings, even the male member.”

    Since when are those “different” meanings? They’re only slightly extended.

  38. Well, “slightly extended” is not the same as “identical.”

  39. marie-lucie says

    Rodger: crann (basically a stake or pole)

    In what language?

  40. David Marjanović says


    I see what you did there.

  41. Chocoholic is I think a neoclassical compound, though it so happens that its parts don’t originate in Latin or Greek. It normally does not make sense within English to ask which part of a neoclassical compound is the root, as they are made up entirely of bound morphemes, even if those bound morphemes are just pieces of originally monomorphemic words. For example, English marathon was originally monomorphemic < the place name Μαραθών (supposedly < μάραθον ‘fennel’, Mycenean ma-ra-tu-wo — but I have my doubts: there is a vowel quality difference and a stress shift to explain somehow). In the early 20C, however, it was segmented as mar-athon, producing walk-athon, skat(e)-athon, talk-athon, and then later as mara-thon, giving rise to tele-thon. So (a)thon is now a bound morpheme meaning ‘something that lasts a long time’, and we might stretch a point and say mar(a)- is a bound morpheme for ‘run’. Neither is identifiably the root, and though walk, talk, skate can be used independently, they are obviously not the heads of the compounds formed from them.

  42. marie-lucie says

    JC: I agree with you. I was disagreeing with “-oholic” being compared with the Salishan, etc “lexical suffixes”.

    In “new” -(a)thon words, the first formant (indeed not the “head”) is a verb, not a noun, therefore different from the choc- in chocoholic.

  43. Well, in telethon it’s either the neoclassical prefix tele- or a shortening of telephone.

  44. A small correction re unkempt. It comes from the Old English umlauted verb cemban ‘to comb’ (hence the e, corresponding to the noun camb ~ comb). The verb and the noun eventually fell together, leaving this isolated relic unpaired.

    We also have uncanny and uncouth, the latter containing the lost past participle of what has become the modal verb can. When it was still an ordinary non-defective preterite-present, its primary meaning was ‘to know’. Chaucer still had couth ‘known, famous’:

    Thanne longen folk to goon on pilgrimages
    And palmeres for to seken straunge strondes
    To ferne halwes, kowthe in sondry londes…

  45. Please, Steve, could you correct my messy markups and it’s > its? I was SURE everything was OK when I was pressing “Post Comment” 🙁

  46. Don’t feel bad, JC screwed up his tags too.

  47. Talking of unpaired, I also have one unpaired bracket. 😉

  48. When reciting the beginning of the General Prologue, I usually omit that couplet: it has just too much that’s too remote from Modern English. The transition without it goes well enough:

    Whan that Aprille with his shoures soote
    The droghte of Marche hath perced to the roote,
    And bathed every veyne in swich licour,
    Of which vertu engendred is the flour;
    Whan Zephirus eek with his swete breeth
    Inspired hath in every holt and heeth
    The tendre croppes, and the yonge sonne
    Hath in the Ram his halfe cours y-ronne,
    And smale fowles maken melodye,
    That slepen al the night with open ye
    (So priketh hem nature in hir corages):
    Than longen folk to goon on pilgrimages,
    And specially, from every shires ende
    Of Engelond, to Caunterbury they wende,
    The holy blisful martir for to seke,
    That hem hath holpen, whan that they were seke.

    (Of course, I read “hem” as “’em”, which is perfect Modern English.)

  49. Talking of unpaired, I also have one unpaired bracket.


  50. I almost left out the right parenthesis in my latest. Fortunately, “almost” only counts in horseshoes and hand grenades.

  51. PG: uncanny : Canny does exist, although its modern meaning is not quite the opposite of that of uncanny.

  52. shortening of telephone

    television, don’t you think? Like teleplay (or telenovela).

    Or Robin Day’s telechair, which broadcasts “Mid-Century Modern” in a way that his polyside, which Hille still makes, somehow doesn’t.

  53. George Gibbard says

    Why doesn’t uncouth rhyme with mouth?

  54. Probably because it is not a direct descendant of OE (un)cúð but a borrowing from Scots, where the Great Vowel Shift did not affect back vowels. Couth in its modern incarnation is apparently a back-formation from uncouth, though in Scots they both have a continuous history back to OE. The related name Cuthbert ‘famous-bright’ went through irregular shortening and the FOOT-STRUT split.

  55. the place name Μαραθών (supposedly < μάραθον ‘fennel’, Mycenean ma-ra-tu-wo — but I have my doubts: there is a vowel quality difference and a stress shift to explain somehow)

    Not a problem as -ων is a highly productive noun formant in Greek: μάραθον is ‘fennel’, Μαραθών ‘the (place) of fennel’.

  56. (The most interesting place the -ων suffix turns up in, by the way, is the name of Poseidon: potei Dāōn, apparently ‘husband of Dā’, that is, Demeter. Of course in the mythology as we know it Poseidon and Demeter are not married, which is why it’s interesting.)

  57. David Marjanović says

    Cognates of couth in German:

    kundig “knowledgeable” (literary)
    Kunde f. “news, tiding, knowledge” (archaic)
    Kunde m. “customer” (< “insider”)

  58. David Marjanović says

    …and while I’m at it, unkempt has always been transparent to me: Kammkämmen.

  59. Gregious Error
    Many a new little life is begot
    By the hibited man with the promptu plot.

    There never is trouble in finding a spouse
    For the ebriated man with the lapidated house.

    Fine Old Professor
    The students who had gnored him
    Universally adored him
    And he died beknownst and famous:
    A gnominious gnoramus.

    The above and many others are by Felicia Lamport, from Light Metres and Scrap Irony, both illustrated by Gorey.

  60. In what language?

    Scots Gaelic, though the Irish form is identical. (I wouldn’t swear that the Irish word has the same range of meanings.)

  61. Marathon = Hinojosa.

Speak Your Mind